EDXRF accurate detection method for Cu content in Ni and Cu coexisting soil

Abstract

The invention provides an EDXRF accurate detection method for the content of Cu in Ni and Cu coexisting soil, and belongs to the technical field of soil heavy metal detection. Standard soil samples with different Cu contents are prepared from soil with low Ni content, quantitative relation models between the net peak value intensity of a Cu K alpha spectrum peak and 3 / 4, 1 / 2 and 1 / 4 full widths in the standard samples are established respectively, the 3 / 4, 1 / 2 and 1 / 4 full widths and 3 / 4, 1 / 2 and 1 / 4 peak value intensities of the Cu K alpha spectrum peak of the soil to be detected correspond to right-side point position information, interpolation is carried out by combining spectral information before a right-side 3 / 4 peak value point, and the content of Ni in the soil to be detected is calculated. And obtaining complete spectral information of a Cu Kalpha characteristic spectral peak, and realizing accurate inversion of the Cu content in the to-be-detected soil by using a Cu content quantitative inversion model established based on a standard sample. The EDXRF accurate detection of the Cu content in the Ni and Cu coexisting soil can be realized without changing the shape of the Cu spectrum peak.

Description

Technical Field

[0001] The present invention belongs to the technical field of soil heavy metal detection, and particularly relates to an accurate EDXRF detection method for Cu content in soil coexisting with Ni and Cu. Background Art

[0002] With the rapid development of social economy, industry and agriculture, heavy metal pollutants generated by a large number of industrial, agricultural and human activities enter the environment, resulting in soil heavy metal pollution, which has become one of the main environmental problems faced by the world today. Due to the characteristics of being difficult to degrade and having biological toxicity, once heavy metals enter the soil, they have an important impact on the growth, development and reproduction of microorganisms, animals and plants, and even cause the death of organisms, thus destroying the safety of the ecological system. In addition, agricultural land polluted by heavy metals will also affect the quality and yield of crops, hinder the development of agricultural economy, and the accumulation of heavy metals through the food chain will also affect human health. Therefore, realizing the rapid on-site detection and screening of soil heavy metals has very important practical significance for protecting the safety of the ecological system, promoting agricultural production and the sustainable development of social economy, and maintaining human health.

[0003] At present, soil heavy metal detection still mainly adopts the method of on-site sampling - off-line laboratory analysis. The main methods include graphite furnace atomic absorption spectrometry, inductively coupled plasma mass spectrometry, atomic fluorescence spectrometry, etc. Although these methods have high sensitivity and accuracy, due to the need for strong acid digestion for sample pretreatment, the process is cumbersome and complex, the detection speed is slow, the analysis time is long, it is easy to cause secondary environmental pollution, and the detection relies on high-cost precision instrument equipment, so the rapid on-site detection and screening of soil heavy metals cannot be realized.

[0004] In contrast, energy dispersive X-ray fluorescence (EDXRF) spectrometry has many advantages such as non-destructiveness to the sample to be measured, no need for complex sample pretreatment, simple instrument, convenient operation, fast analysis speed, and simultaneous detection of multiple elements. It has been widely used in the fields of geological exploration, industrial manufacturing, metal processing, resource investigation, environmental monitoring, etc., and has become an important technology for the rapid on-site detection and screening of soil heavy metals.

[0005] Currently, affected by human industrial and agricultural activities and various aspects such as nickel smelter dust emissions, copper mine exploitation and open-pit tailing storage, stainless steel production dust emissions, electroplating industry wastewater emissions, fossil fuel combustion, transportation, the application of pesticides and fertilizers in agricultural production, e-waste dismantling, nickel industrial waste landfill, and urban pipeline corrosion, soil often faces the phenomenon of combined pollution with multiple heavy metals such as nickel (Ni) and copper (Cu). In the soil EDXRF spectrum, for the Cu element, compared with other characteristic spectral peaks, its Kα characteristic spectral peak is the optimal characteristic spectral peak for sensitive detection of the heavy metal Cu content due to its highest intensity and best sensitivity. However, for soil samples coexisting with Ni and Cu, when using the EDXRF spectral technology to detect and screen soil heavy metals, due to the extremely close energies of the Cu element Kα characteristic spectral peak (8.04 keV) and the Ni element Kβ characteristic spectral peak (8.27 keV), there is cross-over overlap. Therefore, the Ni Kβ characteristic spectral peak will affect the accurate extraction of the Cu Kα characteristic spectral peak information, thus affecting the accurate quantitative detection and analysis of the Cu content by the EDXRF spectral technology. At present, for the problem of XRF spectral peak overlap, although various overlapping spectral peak resolution methods have been established based on the Gaussian mixture model, such as the overlapping peak resolution method combining the improved crow algorithm and the Gaussian mixture model (Zhu Dandan et al., Acta Metrologica Sinica, 2021, 42(10): 1386-1392), the overlapping peak resolution method combining the sparrow search algorithm and the Gaussian mixture model (Chen Ying et al., Spectroscopy and Spectral Analysis, 2021, 41(7): 2175-2180), etc., due to the fact that the heavy metal characteristic spectral peaks in the actual soil XRF spectrum are not completely symmetric due to various factors such as the matrix effect, if the Gaussian function is forced to fit the spectral peaks, it will cause changes in the peak position and peak shape of the spectral peaks in the overlapping peaks. Therefore, for soil samples coexisting with Ni and Cu, the EDXRF spectral technology still faces the problems of difficult accurate acquisition of the Cu element Kα characteristic spectral peak information and difficult accurate detection of the Cu content in the Ni and Cu overlapping spectral peaks. Summary of the Invention

[0006] Aiming at the problem that it is difficult to accurately detect the Cu content in soil coexisting with Ni by EDXRF spectroscopy, the purpose of the present invention is to make up for the defects of the existing technologies and provide an accurate EDXRF detection method for the Cu content in soil coexisting with Ni. This method uses soil with a low Ni content to prepare standard soil samples with different Cu contents. Since the intensity of the Kβ characteristic spectral peak is weak in the presence of low-content Ni and the interference with the Cu Kα characteristic spectral peak is small, first, according to the spectral information between the peak point of the Cu Kα characteristic spectral peak and the 1 / 4 peak intensity point on the right side in the standard soil sample and the spectral information after setting the fluorescence intensity on the right side of the peak point of the Ni Kβ spectral peak to 0, the Lagrange interpolation method is used to obtain the Cu Kα characteristic spectral peak information after the 1 / 4 peak intensity point on the right side of the spectral peak. According to the interpolated complete Cu Kα characteristic spectral peak information, its net integrated fluorescence intensity value is obtained, so as to establish a quantitative relationship model between the standard Cu content and the net integrated fluorescence intensity of the Cu Kα characteristic spectral peak, which is used for the inversion of the Cu content in the soil to be measured; secondly, quantitative relationship models are respectively established between the peak intensity of the Cu Kα characteristic spectral peak of the standard soil sample and the 3 / 4 full width at half maximum, the full width at half maximum, and the 1 / 4 full width at half maximum. Based on this, according to the peak intensity of the Cu Kα characteristic spectral peak in the soil to be measured coexisting with Ni and Cu, the 3 / 4 full width at half maximum, the full width at half maximum, and the 1 / 4 full width at half maximum of the Cu Kα spectral peak can be obtained without Ni interference. Combining the energy values corresponding to the 3 / 4 full width at half maximum, the full width at half maximum, and the 1 / 4 full width at half maximum on the left side of the Cu Kα spectral peak of the soil to be measured that are not affected by Ni, the 3 position information corresponding to the 3 / 4 full width at half maximum, the full width at half maximum, and the 1 / 4 full width at half maximum on the right side of the Cu Kα spectral peak without Ni influence can be obtained. Further, these 3 position information are combined with the spectral information between the peak point and the 3 / 4 peak intensity point on the right side and the spectral information after setting the fluorescence intensity on the right side of the peak point of the Ni Kβ spectral peak to 0, and Lagrange interpolation is performed on the spectral data after the 3 / 4 peak intensity point on the right side of the Cu Kα spectral peak, so as to obtain the entire Cu Kα characteristic spectral peak information in the soil to be measured coexisting with Ni and Cu. And substituting the net integrated fluorescence intensity value of the entire Cu Kα spectral peak into the quantitative relationship model between the standard Cu content and the net integrated fluorescence intensity of the Cu Kα characteristic spectral peak established based on the standard soil sample, the accurate inversion of the Cu content in the soil sample to be measured coexisting with Ni and Cu can be realized, thus solving the problem that it is difficult to accurately detect the Cu content in soil coexisting with Ni by the existing EDXRF spectroscopy.

[0007] The technical solution of the present invention is described as follows:

[0008] An accurate EDXRF detection method for the Cu content in soil coexisting with Ni and Cu, comprising the following steps:

[0009] Step 1: Prepare a series of standard soil samples with different Cu contents.

[0010] Select soil samples with a low Ni content and a weak Kβ spectral peak, add different amounts of heavy metal Cu to them, stir evenly, and then make a series of flat circular standard soil samples with the same physical properties and different Cu contents through drying, grinding, sieving, and pressing, so as to eliminate the influence of physical property differences on the EDXRF spectral measurement and quantitative analysis of Cu element.

[0011] Step 2: Determine the standard content value of Cu in the standard soil samples.

[0012] Use flame atomic absorption spectrophotometry to measure the standard content values of Cu in the series of standard soil samples prepared in Step 1 to obtain the standard content values of Cu (C1, C2, …, C k ), where k represents the number of standard soil samples.

[0013] Step 3: Measure the EDXRF spectra of the standard soil samples.

[0014] Under the same measurement conditions, perform EDXRF spectral measurements on the series of circular standard soil samples with different heavy metal Cu contents prepared in Step 1 respectively, and obtain the EDXRF spectra corresponding to different standard soil samples.

[0015] Step 4: Denoise and subtract the background from the EDXRF spectra of the standard soil samples to obtain the net EDXRF spectra.

[0016] Use the Savitzky-Golay filtering method to denoise the EDXRF spectra of different standard soil samples obtained in Step 3, and use the extreme value method to identify the peak valleys. Fit the identified series of peak valley points with a cubic smoothing spline curve corrected by a penalty term to obtain the spectral background baseline. By subtracting the original spectrum from the fitted background baseline, the background subtraction of the EDXRF spectra is realized, and the net EDXRF spectra of the standard soil samples are obtained.

[0017] Step 5: Extract the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ in the EDXRF spectra of the standard soil samples.

[0018] For the net EDXRF spectra of the standard soil samples, according to the energy values of the Cu Kα characteristic spectral peak (energy is 8.04 keV) and the Ni Kβ characteristic spectral peak (energy is 8.27 keV), use the extreme value method to determine the left valley point [E1, F1], peak point [E p , F p of the Cu Kα characteristic spectral peak and the peak point [E q , F q of the Ni Kβ characteristic spectral peak, and the right valley point [E n , Fn , extract the spectral information in the interval from the left valley point of the Cu Kα characteristic spectral peak to the right valley point of the Ni Kβ characteristic spectral peak ([E1, F1], [E2, F2], …, [E i , F i , …, [E n , F n ). This spectral information is the complete spectral information of the overlapping spectral peak of Cu Kα and Ni Kβ in the EDXRF spectrum of the standard soil sample, where E i is the i-th energy value in the extracted overlapping spectral peak, and F i is the net fluorescence intensity corresponding to the i-th energy value.

[0019] Step 6: Calculate and obtain the full width at three-fourths maximum, full width at half maximum, and full width at one-fourth maximum of the Cu Kα characteristic spectral peak in the standard soil sample.

[0020] According to the net fluorescence intensity value F p at the peak position of the Cu Kα spectral peak in the net EDXRF spectra of a series of standard soil samples with different heavy metal Cu contents, calculate the fluorescence intensity values of 3 / 4F p , 1 / 2F p , and 1 / 4F p respectively. Since the Ni content in the standard soil sample is low and its Kβ characteristic spectral peak is weak, with little influence on the Cu Kα characteristic spectral peak, based on the measured spectral information between the left valley point [E1, F1] of the Cu Kα characteristic spectral peak in the standard soil sample and the position near the 1 / 4F p fluorescence intensity on the right, use the Lagrange interpolation method to obtain the left position information [E p , 3 / 4 F p , [E p , 1 / 2 F 3 / 4L , [E p , 1 / 4 F 1 / 2L and the right position information [E p , 3 / 4 F 1 / 4L , [E p , 1 / 2 F 3 / 4R , [E p , 1 / 4 F 1 / 2R of the Cu Kα characteristic spectral peak when the fluorescence intensity is 3 / 4F p , 1 / 2F 1 / 4R , 1 / 4F p respectively. Further calculate and obtain the full width at three-fourths maximum W 3 / 4 , full width at half maximum W 1 / 2 , and full width at one-fourth maximum W1 / 4 。

[0021] W 3 / 4 = E 3 / 4R - E 3 / 4L (1)

[0022] W 1 / 2 = E 1 / 2R - E 1 / 2L (2)

[0023] W 1 / 4 = E 1 / 4R - E 1 / 4L (3)

[0024] Step 7: Based on standard soil samples, establish quantitative relationship models between the full width at 3 / 4 maximum, full width at half maximum, full width at 1 / 4 maximum of the Cu Kα characteristic spectral peak and the net peak fluorescence intensity, respectively.

[0025] According to the net peak fluorescence intensity value F of the Cu Kα spectral peak in the EDXRF spectra of a series of standard soil samples with different heavy metal Cu contents p and the full width at 3 / 4 maximum W 3 / 4 , full width at half maximum W 1 / 2 , full width at 1 / 4 maximum W 1 / 4 values, establish quantitative relationship models between W 3 / 4 and F p , W 1 / 2 and F p , W 1 / 4 and F p respectively, as shown in formulas (4)-(6).

[0026] W 3 / 4 = f1(F p ) (4)

[0027] W 1 / 2 = f2(F p ) (5)

[0028] W 1 / 4 = f3(F p ) (6)

[0029] Step 8: Based on the net integrated fluorescence intensity value of the Cu Kα characteristic spectral peak and the Cu standard content value of the standard soil sample, establish a quantitative inversion model for the Cu content of the soil sample.

[0030] Take the peak point (energy is E q ) of the Ni Kβ characteristic spectral peak in the overlapping peak of Cu Kα and Ni Kβ to the right valley point (energy is E nThe fluorescence intensity values of the intermediate spectra are all set to 0. According to a series of standard soil samples with different heavy metal Cu contents, the spectral information between the peak point of the Cu Kα spectral peak in the net EDXRF spectrum and the right point corresponding to the fluorescence intensity of 1 / 4F p ([E p ,F p , …, [E 1 / 4R , 1 / 4 F p ) and the spectral information after setting the fluorescence intensity between the peak point of the Ni Kβ spectral peak and the right valley point to 0 ([E q , 0], …, [E n , 0]) are used. The Lagrange interpolation method is adopted to obtain the intensity information of the Cu Kα spectral peak after the point [E 1 / 4R , 1 / 4 F p . The number of interpolation points and the corresponding energy values of the abscissa are the same as those of the corresponding points in the original measured EDXRF spectrum with energy between E 1 / 4R ~E n . According to the interpolated fluorescence intensity information, the point information [E m , F m corresponding to the minimum value of the fluorescence intensity on the right side of the Cu Kα spectral peak is determined. After interpolation, the spectral data corresponding to the Cu Kα characteristic spectral peak in the standard soil sample is ([E1, F1], [E2, F2], …, [E i , F i , …, [E m , F m ), and m < n. According to ([E1,F1], [E2, F2], …, [E i , F i , …, [E m , F m ), the net integrated fluorescence intensity S values (S1, S2, …, S k ) of the Cu Kα spectral peaks in a series of standard soil samples with different heavy metal Cu contents C (C1, C2, …, C k ) in the EDXRF spectra are calculated. k represents the number of standard soil samples, and the heavy metal Cu standard content C values (C1, C2, …, C k ) are linearly fitted with the corresponding net integrated fluorescence intensity S values (S1, S2, …, S k ) according to the following formula (7) to establish a quantitative inversion model for the heavy metal Cu content.

[0031] C = a S + b (7)

[0032] Among them, a represents a coefficient and b is a constant.

[0033] Step 9: Prepare the test soil sample, perform EDXRF spectral measurement, spectral denoising, and background subtraction on the test soil sample to obtain the net EDXRF spectrum of the test soil sample.

[0034] For a test soil sample with coexisting Cu and Ni and unknown contents, a disc-shaped soil sample with the same physical properties as the standard soil sample and a flat surface is prepared using the same treatment method as in Step 1 above. Under the same EDXRF spectral measurement conditions as in Step 3, the EDXRF spectrum of the test soil sample is measured, and the measured EDXRF spectrum is denoised, peak-valley point identified, background baseline fitted, and background subtracted in the same manner as in Step 4 to obtain the net EDXRF spectrum after background subtraction.

[0035] Step 10: Extract the complete spectral information of the overlapping peaks of Cu Kα and Ni Kβ in the net EDXRF spectrum of the test soil sample.

[0036] For the net EDXRF spectrum of the test soil sample after background subtraction obtained in Step 9, in the same manner as in Step 5, according to the energy values of the Cu Kα characteristic spectral peak (energy: 8.04 keV) and the Ni Kβ characteristic spectral peak (energy: 8.27 keV), the extreme value method is used to determine the left valley point [E1', F1'], peak point [E p ', F p '] of the Cu Kα characteristic spectral peak, and the peak point [E q ', F q '] and right valley point [E n ', F n '] of the Ni Kβ characteristic spectral peak, and then the complete spectral information of the overlapping peaks of Cu Kα and Ni Kβ in the interval from the left valley point of the Cu Kα characteristic spectral peak to the right valley point of the Ni Kβ characteristic spectral peak ([E1', F1'], [E2', F2'], …, [E i ', F i '], …, [E n ', F n ']) is extracted.

[0037] Step 11: Obtain the position information corresponding to 3 / 4 peak fluorescence intensity, 1 / 2 peak fluorescence intensity, and 1 / 4 peak fluorescence intensity on the left side of the Cu Kα characteristic spectral peak of the test soil sample through interpolation.

[0038] According to the complete spectral information of the overlapping peaks of Cu Kα and Ni Kβ in the test soil sample ([E1', F1'], [E2', F2'], …, [E i ', F i'], …, [E n ', F n ']), extract the net fluorescence intensity value F at the peak position of the Cu Kα spectral peak p ', and calculate the fluorescence intensity values of 3 / 4F p ', 1 / 2F p ', and 1 / 4F p ' respectively. According to the spectral information between the valley point [E1', F1'] on the left side of the Kα characteristic spectral peak of the Cu element and the peak point [E p ', F p '], use the Lagrange interpolation method to obtain the fluorescence intensity between [E1', F1'] and [E p ', F p '] on the left side of the Cu Kα characteristic spectral peak is 3 / 4F p ', 1 / 2F p ', 1 / 4F p ' corresponding point position information [E 3 / 4L ', 3 / 4 F p '], [E 1 / 2L ', 1 / 2 F p '], [E 1 / 4L ', 1 / 4 F p '], where E 3 / 4L ', E 1 / 2L ', E 1 / 4L ' are the energy values of the corresponding points of the fluorescence intensity of 3 / 4F p ', 1 / 2F p ', 1 / 4F p ' on the left side of the Cu Kα characteristic spectral peak.

[0039] Step 12: Calculate and obtain the 3 / 4 full width at half maximum, full width at half maximum, and 1 / 4 full width at half maximum of the Cu Kα characteristic spectral peak of the soil sample to be measured.

[0040] Substitute the net fluorescence intensity value F p ' at the peak position of the Cu Kα characteristic spectral peak of the soil sample to be measured extracted in Step 11 into the quantitative relationship models between W 3 / 4 and F p , W 1 / 2 and F p , W 1 / 4 and F p shown in Formulas (4) - (6) in Step 7 respectively, and obtain the 3 / 4 full width at half maximum W 3 / 4 ', full width at half maximum W 1 / 2 ', and 1 / 4 full width at half maximum W 1 / 4 ' values of the Cu Kα characteristic spectral peak of the soil sample to be measured.

[0041] Step 13: Calculate the position information corresponding to the fluorescence intensities of 3 / 4 peak value, 1 / 2 peak value, and 1 / 4 peak value on the right side of the Cu Kα characteristic spectral peak of the soil sample to be measured.

[0042] The energy values E p ' corresponding to the positions where the fluorescence intensities on the left side of the Cu Kα characteristic spectral peak in the soil sample to be measured obtained in Step 11 are 3 / 4F p ', 1 / 2F p ', and 1 / 4F 3 / 4L ' are added to the full width at half maximum W 1 / 2L ' of 3 / 4 of the Cu Kα characteristic spectral peak, the full width at half maximum W 1 / 4L ' of the Cu Kα characteristic spectral peak, and the full width at half maximum W 3 / 4 ' of 1 / 4 of the Cu Kα characteristic spectral peak obtained in Step 12, respectively. According to the following formulas (8)-(10), the energy values E 1 / 2 ' corresponding to the positions where the fluorescence intensities on the right side of the Cu Kα characteristic spectral peak in the soil sample to be measured are 3 / 4F 1 / 4 ', 1 / 2F p ', and 1 / 4F p ' are calculated respectively, so as to obtain three position information on the right side of the Cu Kα characteristic spectral peak without interference from the Ni Kβ characteristic spectral peak: [E p ', 3 / 4 F 3 / 4R '], [E 1 / 2R ', 1 / 2 F 1 / 4R '], [E 3 / 4R ', 1 / 4 F p ']. 1 / 2R ', 1 / 2 F p '], [E 1 / 4R ', 1 / 4 F p '].

[0043] E 3 / 4R ' = E 3 / 4L ' + W 3 / 4 ' (8)

[0044] E 1 / 2R ' = E 1 / 2L ' + W 1 / 2 ' (9)

[0045] E 1 / 4R ' = E 1 / 4L ' + W 1 / 4 ' (10)

[0046] Step 14: Obtain the complete spectral information of the Cu Kα characteristic spectral peak of the soil sample to be measured by interpolation.

[0047] The peak point (energy is Eq ') to the right valley point (energy is E n '), the fluorescence intensity values of the spectrum between them are all set to 0. According to the three point position information on the right side of the Cu Kα characteristic spectral peak without interference from the Ni Kβ characteristic spectral peak obtained in step 13 [E 3 / 4R ', 3 / 4 F p '], [E 1 / 2R ', 1 / 2 F p '], [E 1 / 4R ', 1 / 4 F p '] and the peak point [E p ', F p '] to the spectral information of the Cu Kα characteristic spectral peak before the position corresponding to the intensity of 3 / 4 F p ' on the right side, combined with the spectral information after setting the fluorescence intensity between the peak point of the Ni Kβ spectral peak and the right valley point to 0 ([E q ', 0], …, [E n ', 0]), the Lagrange interpolation method is used to interpolate the spectral information after the position corresponding to the intensity of 3 / 4 F p ' on the right side of the Cu Kα characteristic spectral peak. The number of interpolation points and the energy values of the abscissa are the same as those of the corresponding points in the original measured EDXRF spectrum with energy between E 3 / 4R ' and E n '. According to the fluorescence intensity information obtained by interpolation, determine the position information [E m ', F m '] corresponding to the minimum value of the fluorescence intensity on the right side of the Cu Kα spectral peak. Then, after interpolation, the spectral data of the Cu Kα characteristic spectral peak in the soil sample to be measured is obtained as ([E1', F1'], [E2', F2'], …, [E i ', F i '], …, [E m ', F m ']), and m < n.

[0048] Step 15: Invert to obtain the content value of Cu in the soil sample to be measured.

[0049] According to the complete spectral information of the Cu Kα characteristic spectral peak in the soil sample to be measured obtained in the above step 14, calculate the net integral fluorescence intensity value S' of the CuKα characteristic spectral peak, and substitute S' into the Cu content quantitative inversion model shown in formula (7) in step 8 to invert and obtain the content value of Cu in the soil sample to be measured with coexisting Ni and Cu, realizing the accurate detection of Cu content in the soil with coexisting Ni and Cu by EDXRF.

[0050] Beneficial effects:

[0051] Compared with various overlapping spectral peak analysis methods currently based on the Gaussian mixture model, the method proposed in the present invention does not change the peak shape of the Cu Kα characteristic spectral peak in the overlapping peak of Cu Kα and Ni Kβ, and can well retain the original shape characteristics of the Cu Kα characteristic spectral peak in the soil coexisting with Ni and Cu. Therefore, it can simply, quickly, and accurately obtain the characteristic spectral peak information of Cu Kα in the overlapping peak of Cu Kα and Ni Kβ, realize the accurate analysis of the Cu content in the soil coexisting with Ni and Cu, solve the problem that it is difficult to accurately detect the Cu content in the soil coexisting with Cu and Ni by the current EDXRF spectroscopy technology, and is more conducive to improving the application performance of the EDXRF spectroscopy technology in the on-site rapid detection and screening of soil heavy metals, providing an important method basis for the further development of on-site rapid detection equipment for soil heavy metal EDXRF. Description of the Drawings

[0052] Figure 1 : Flow chart of the EDXRF accurate detection method for the Cu content in the soil coexisting with Ni and Cu.

[0053] Figure 2 : Complete spectral information of the overlapping peaks of Cu Kα and Ni Kβ in the net EDXRF spectra of 6 standard soil samples.

[0054] Figure 3 : Complete spectral information of the overlapping peaks of Cu Kα and Ni Kβ in the net EDXRF spectra of 5 soil samples to be measured. Detailed Embodiments

[0055] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. For the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0056] Embodiment

[0057] As Figure 1 shown, the embodiment of the present invention proposes an EDXRF accurate detection method for the Cu content in the soil coexisting with Ni and Cu, including the following steps:

[0058] Step 1: Prepare a series of standard soil samples with different Cu contents.

[0059] Agricultural soil with low Ni content and weak Kβ spectral peak which was not polluted by heavy metals and was collected from Science Island in Hefei was selected to prepare standard soil samples with different heavy metal Cu contents. Different contents of heavy metal Cu were added to the selected agricultural soil and stirred evenly. Then, a series of standard soil samples with flat surfaces and different Cu contents were made through drying, grinding, sieving and pressing to eliminate the influence of physical property differences on EDXRF spectrum measurement and quantitative analysis of Cu element.

[0060] Different contents of heavy metal Cu were added to the selected agricultural soil. 2+ Copper nitrate (Cu(NO3)2·9H2O) standard solution with a mass concentration of 1000 mg / L.

[0061] The specific preparation method of the standard disc-shaped soil samples with the same physical properties and different Cu contents is as follows: the soil samples with heavy metal Cu added are stirred evenly, placed in a 60℃ constant temperature electric heating blast drying oven for drying to remove moisture, and the soil is fully ground after drying, passed through a 100-mesh sieve, and then weighed 3.000 ± 0.005 g and placed in a mold with an inner diameter of 30 mm. A manual powder tablet press is used to maintain the pressure at a constant pressure of 18 MPa for 3 min to obtain disc-shaped soil samples with a thickness of 2 mm and a diameter of 30 mm. The prepared series of heavy metal Cu standard soil samples have the same physical properties such as particle size, compactness, thickness, and diameter, which effectively avoids the influence of physical property differences on the EDXRF spectrum measurement and quantitative analysis of the Cu element.

[0062] Step 2: Determine the standard content value of Cu in the standard soil sample.

[0063] The standard content values of Cu in a series of standard soil samples prepared in step 1 were measured by flame atomic absorption spectrophotometry to obtain the standard content values of Cu (C1, C2, …, C k ), where k represents the number of standard soil samples.

[0064] In the series of standard soil samples with different heavy metal Cu contents prepared in this embodiment, the number of samples k is 6. The standard content values of Cu are 40.65 mg / kg, 92.84 mg / kg, 173.31 mg / kg, 238.48 mg / kg, 322.86 mg / kg, and 425.16 mg / kg, respectively, as determined by flame atomic absorption spectrophotometry.

[0065] Step 3: Measure the EDXRF spectrum of the standard soil sample.

[0066] Under the same measurement conditions, a series of circular standard soil samples with different heavy metal Cu contents prepared in Step 1 were respectively subjected to EDXRF spectral measurement to obtain the EDXRF spectra corresponding to different standard soil samples.

[0067] In this embodiment, the instrument used for the EDXRF spectral measurement of the standard soil samples is the EDXRF spectrometer of AMPTEK Company, USA. The excitation source is a Min-X-ray tube, the target material is an Ag target, and there are 2 pieces of 10 mils Al primary filters in front of the tube. The detector is an SDD-123 type silicon drift detector. The conditions for the EDXRF spectral measurement of the standard soil samples are as follows: the working voltage of the X-ray tube is 40 kV, the working current is 20 μA, the pulse shaping time is 6.4 μs, and the gain is 47.47; the soil sample is placed flat, and the EDXRF spectral measurement of the soil sample is carried out at a position 1.6 cm away from the detection window of the instrument. The cumulative time of the EDXRF spectral detection signal is 120 s.

[0068] Step 4: Denoise and background subtract the EDXRF spectra of the standard soil samples to obtain the net EDXRF spectra.

[0069] The Savitzky-Golay filtering method is used to denoise the EDXRF spectra of different standard soil samples obtained in Step 3, and the extreme value method is used for peak-valley identification. The identified series of peak-valley points are fitted with a cubic smoothing spline curve corrected by a penalty term to obtain the spectral background baseline. By subtracting the original spectrum from the fitted background baseline, the background subtraction of the EDXRF spectrum is realized, and the net EDXRF spectra of the standard soil samples are obtained.

[0070] Step 5: Extract the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ in the EDXRF spectra of the standard soil samples.

[0071] For the net EDXRF spectra of the standard soil samples, according to the energy values of the Cu Kα characteristic spectral peak (energy is 8.04 keV) and the Ni Kβ characteristic spectral peak (energy is 8.27 keV), the extreme value method is used to determine the left valley point [E1, F1], peak point [E p , F p of the Cu Kα characteristic spectral peak and the peak point [E q , F q and the right valley point [E n , F n of the Ni Kβ characteristic spectral peak, and extract the spectral information in the interval from the left valley point of the Cu Kα characteristic spectral peak to the right valley point of the Ni Kβ characteristic spectral peak ([E1, F1], [E2, F2], …, [E i ,F i , …, [En , F n ), and this spectral information is the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ in the EDXRF spectrum of the standard soil sample, where E i is the i-th energy value in the extracted overlapping spectral peak, and F i is the net fluorescence intensity corresponding to the i-th energy value.

[0072] In this embodiment, the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ in the net EDXRF spectra of 6 standard soil samples is as Figure 2 shown.

[0073] Step 6: Calculate and obtain the full width at three-fourths maximum, full width at half maximum, and full width at one-fourth maximum of the Cu Kα characteristic spectral peak in the standard soil sample.

[0074] According to the net fluorescence intensity values F p at the peak positions of the Cu Kα spectral peaks in the net EDXRF spectra of a series of standard soil samples with different heavy metal Cu contents, the fluorescence intensity values of 3 / 4F p , 1 / 2F p , and 1 / 4F p are calculated respectively. Since the Ni content in the standard soil sample is low and its Kβ characteristic spectral peak is weak, with little influence on the Cu Kα characteristic spectral peak, based on the measured spectral information between the left valley point [E1, F1] and the point near the 1 / 4F p fluorescence intensity on the right side of the Cu Kα characteristic spectral peak in the standard soil sample, the Lagrange interpolation method is used to obtain the left-side point information [E p , 3 / 4 F p , [E p , 1 / 2 F 3 / 4L , [E p , 1 / 4 F 1 / 2L and the right-side point information [E p , 3 / 4 F 1 / 4L , [E p , 1 / 2 F 3 / 4R , [E p , 1 / 4 F 1 / 2R of the Cu Kα characteristic spectral peak when the fluorescence intensity values are 3 / 4F p , 1 / 2F 1 / 4R , and 1 / 4F p respectively. Further, the full width at three-fourths maximum W 3 / 4 , full width at half maximum W 1 / 2 , and full width at one-fourth maximum W 1 / 4 of the Cu Kα characteristic spectral peak in a series of standard soil samples are calculated according to the following formulas (1)-(3).

[0075] W 3 / 4 = E 3 / 4R - E 3 / 4L (1)

[0076] W 1 / 2 = E 1 / 2R - E 1 / 2L (2)

[0077] W 1 / 4 = E 1 / 4R - E 1 / 4L (3)

[0078] For the 6 standard soil samples in this embodiment, the net fluorescence intensity values F of the peak positions of the Cu Kα spectral peaks in the net EDXRF spectrum p are 93.2301, 108.0782, 125.3709, 137.1876, 161.5427, and 192.3228 respectively. The 3 / 4F p , 1 / 2F p , 1 / 4F p fluorescence intensity values of the 6 standard soil samples are shown in Table 1 respectively.

[0079] Table 1 3 / 4F of the 6 standard soil samples p , 1 / 2F p , 1 / 4F p fluorescence intensity values

[0080] According to the measured spectral information between the left valley point [E1, F1] and the point near the 1 / 4F p fluorescence intensity on the right side of the Cu Kα characteristic spectral peak in the 6 standard soil samples, the fluorescence intensity values obtained by using the Lagrange interpolation method are 3 / 4F p , 1 / 2F p , 1 / 4F p respectively. The left point information [E 3 / 4L , 3 / 4 F p , [E 1 / 2L , 1 / 2 F p , [E 1 / 4L , 1 / 4 F p in the Cu Kα characteristic spectral peak, and the E 3 / 4L , E 1 / 2L , E 1 / 4L values and the right point information [E 3 / 4R , 3 / 4 F p , [E 1 / 2R , 1 / 2 F p , [E 1 / 4R , 1 / 4 Fp E in 3 / 4R 、E 1 / 2R 、E 1 / 4R The values of E are shown in Table 2. Further, the full widths at three - fourths maximum (W), full widths at half maximum (W), and full widths at one - fourth maximum (W) of the Cu Kα characteristic spectral peaks in 6 standard soil samples are calculated according to Formulas (1)-(3). 3 / 4 、full widths at half maximum W 1 / 2 、full widths at one - fourth maximum W 1 / 4 are shown in Table 3.

[0081] Table 2 The energy values of the left and right positions of the Cu Kα characteristic spectral peak when the fluorescence intensity values of 3 / 4F p 、1 / 2F p 、1 / 4F p in 6 standard soil samples obtained by interpolation

[0082] Table 3 The full widths at three - fourths maximum (W), full widths at half maximum (W), and full widths at one - fourth maximum (W) of the Cu Kα characteristic spectral peaks in 6 standard soil samples 3 / 4 、full widths at half maximum W 1 / 2 、full widths at one - fourth maximum W 1 / 4 values

[0083] Step 7: Based on the standard soil samples, establish quantitative relationship models between the full widths at three - fourths maximum, full widths at half maximum, full widths at one - fourth maximum of the Cu Kα characteristic spectral peak and the net peak fluorescence intensity respectively.

[0084] According to the net peak fluorescence intensity values F p and the full widths at three - fourths maximum W 3 / 4 、full widths at half maximum W 1 / 2 、full widths at one - fourth maximum W 1 / 4 values in the EDXRF spectra of a series of standard soil samples with different heavy metal Cu contents, establish quantitative relationship models between W 3 / 4 and F p 、W 1 / 2 and F p 、W 1 / 4 and F p respectively. Then, the quantitative relationship models between W 3 / 4 and F p 、W 1 / 2 and F p 、W 1 / 4 and F p established based on 6 standard soil samples are shown in Formulas (4)-(6) respectively.

[0085] When F p <125.3709, W 3 / 4=0.1715 - 0.002 F p +1.13503 10 -5 F p 2

[0086] When F p ≥125.3709, W 3 / 4 =0.12802 - 3.28927 10 -4 F p +7.8116 10 -7 F p 2 (4)

[0087] When F p <125.3709, W 1 / 2 =0.04013 + 0.00167 F p -5.42624 10 -6 F p 2

[0088] When F p ≥125.3709, W 1 / 2 =6571.05218 exp(-F p / 9.30544)+0.15514 (5)

[0089] When F p <125.3709, W 1 / 4 =0.24136 - 8.21375 10 -4 F p +5.35149 10 -6 F p 2

[0090] When F p ≥125.3709, W 1 / 4 =4.73243 exp(-F p / 20.28833)+0.21268 (6)

[0091] Step 8: Based on the net integrated fluorescence intensity value of the Cu Kα characteristic spectral peak and the Cu standard content value of the standard soil sample, establish a quantitative inversion model for the Cu content in the soil sample.

[0092] Set the fluorescence intensity values of the spectrum between the peak point (energy is E q ) of the Ni Kβ characteristic spectral peak in the overlapping peak of Cu Kα and Ni Kβ to the right valley point (energy is E n ) to 0. According to the spectral information between the peak point of the Cu Kα spectral peak and the right position corresponding to the fluorescence intensity of 1 / 4F p in the net EDXRF spectra of a series of standard soil samples with different heavy metal Cu contents ([E p , F p , …, [E 1 / 4R , 1 / 4 F p ) and the spectral information after setting the fluorescence intensity between the peak point of the Ni Kβ spectral peak and the right valley point to 0 ([E q , 0], …, [E n , 0]), use the Lagrange interpolation method to obtain the intensity information of the Cu Kα spectral peak after the point [E 1 / 4R , 1 / 4 F p . The number of interpolation points and the corresponding energy values of the abscissa are the same as the number and energy values of the corresponding points in the original measured EDXRF spectrum with energy between E 1 / 4R ~E n . According to the interpolated fluorescence intensity information, determine the position information [E m , F m corresponding to the minimum value when the fluorescence intensity on the right side of the Cu Kα spectral peak decreases. After interpolation, the spectral data corresponding to the Cu Kα characteristic spectral peak in the standard soil sample is ([E1, F1], [E2, F2], …, [E i , F i , …, [E m , F m ), and m < n. According to ([E1, F1], [E2, F2], …, [E i , F i , …, [E m , F m ) calculate the net integrated fluorescence intensity S values (S1, S2, …, S k ) of the Cu Kα spectral peak in the EDXRF spectra of a series of standard soil samples with different heavy metal Cu contents C (C1, C2, …, C k ). And the heavy metal Cu standard content C values (C1, C2, …, C k), and perform linear fitting with the corresponding net integrated fluorescence intensity S values (S1, S2, …, S k ) to establish a quantitative inversion model for the heavy metal Cu content.

[0093] For the 6 standard soil samples, the energy range of the fluorescence data of the Cu Kα characteristic spectral peaks obtained by interpolation is 7.8746 keV to 8.2258 keV, with a total of 53 spectral data, that is, m = 53. The net integrated fluorescence intensity S values of the Cu Kα spectral peaks in the EDXRF spectra of the 6 standard soil samples calculated from the 53 spectral data are 2088.7188, 2410.4081, 2902.0850, 3239.0799, 3722.6154, and 4419.9801 respectively. Perform linear fitting between the standard content C values (40.65 mg / kg, 92.84 mg / kg, 173.31 mg / kg, 238.48 mg / kg, 322.86 mg / kg, 425.16 mg / kg) of Cu in the 6 standard soil samples and the corresponding net integrated fluorescence intensity S values (2088.7188, 2410.4081, 2902.0850, 3239.0799, 3722.6154, 4419.9801) of the Cu Kα spectral peaks to establish a quantitative inversion model for the heavy metal Cu content, as shown in formula (7).

[0094] C = -308.75545 + 0.16748 S (7)

[0095] Step 9: Prepare the sample of the soil sample to be measured, measure the EDXRF spectrum, perform spectral denoising and background subtraction to obtain the net EDXRF spectrum of the soil sample to be measured.

[0096] Take 5 soil samples with unknown coexisting Cu and Ni contents as the soil samples to be measured, numbered 1#, 2#, 3#, 4#, and 5# respectively. Use the same treatment method as in Step 1 above to prepare disc-shaped soil samples with the same physical properties as the standard soil samples and a flat surface. Under the same EDXRF spectrum measurement conditions as in Step 3, measure the EDXRF spectrum of the soil sample, and use the same method as in Step 4 to perform denoising, peak-valley point identification, background baseline fitting, and background subtraction on the measured EDXRF spectrum to obtain the net EDXRF spectrum after background subtraction.

[0097] Step 10: Extract the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ in the net EDXRF spectrum of the soil sample to be measured.

[0098] For the net EDXRF spectra of the 5 soil samples to be measured after background subtraction obtained in step 9, in the same manner as in step 5, according to the energy values of the Cu Kα characteristic spectral peak (energy: 8.04 keV) and the Ni Kβ characteristic spectral peak (energy: 8.27 keV), the extreme value method is used to determine the left valley point [E1', F1'], peak point [E p ', F p '] of the Cu Kα characteristic spectral peak and the peak point [E q ', F q '] of the Ni Kβ characteristic spectral peak, as well as the right valley point [E n ', F n ']. Furthermore, the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ in the interval from the left valley point of the Cu Kα characteristic spectral peak to the right valley point of the Ni Kβ characteristic spectral peak is extracted ([E1', F1'], [E2', F2'], …, [E i ', F i '], …, [E n ', F n ']).

[0099] The complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ in the net EDXRF spectra of the 5 soil samples to be measured is as Figure 3 shown.

[0100] Step 11: Obtain the position information corresponding to 3 / 4 peak fluorescence intensity, 1 / 2 peak fluorescence intensity, and 1 / 4 peak fluorescence intensity on the left side of the Cu Kα characteristic spectral peak of the soil samples to be measured through interpolation.

[0101] According to the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ of the 5 soil samples to be measured in step 10 ([E1', F1'], [E2', F2'], …, [E i ', F i '], …, [E n ', F n ']), the net fluorescence intensity value F p ' at the peak position of the Cu Kα spectral peak is extracted, and the fluorescence intensity values of 3 / 4F p ', 1 / 2F p ', and 1 / 4F p ' are calculated respectively. According to the spectral information between the left valley point [E1', F1'] and the peak point [E p ', F p '] of the Kα characteristic spectral peak of the Cu element, the Lagrange interpolation method is used to obtain the left side [E1', F1'] and [E p ', F p' The fluorescence intensity at this point is 3 / 4F p ', 1 / 2F p ', 1 / 4F p ' corresponding point information [E 3 / 4L ', 3 / 4 F p '], [E 1 / 2L ', 1 / 2 F p '], [E 1 / 4L ', 1 / 4 F p '], where E 3 / 4L ', E 1 / 2L ', E 1 / 4L ' are the energy values of the points corresponding to the fluorescence intensities of 3 / 4F p ', 1 / 2F p ', 1 / 4F p ' on the left side of the Cu Kα characteristic spectral peak.

[0102] Then, based on the complete spectral information of the overlapping peaks of Cu Kα and Ni Kβ in 5 soil samples to be measured, the net fluorescence intensity values F p ' and 3 / 4F p ', 1 / 2F p ', 1 / 4F p ' of the peak positions of the Cu Kα spectral peaks in the 5 soil samples to be measured are shown in Table 4 respectively. The energy values E p ', E p ', E p ' corresponding to the fluorescence intensities of 3 / 4F 3 / 4L ', 1 / 2F 1 / 2L ', 1 / 4F 1 / 4L ' on the left side of the Cu Kα characteristic spectral peak obtained by Lagrange interpolation for the 5 soil samples to be measured are shown in Table 5.

[0103] Table 4 F p ' values and 3 / 4F p ', 1 / 2F p ', 1 / 4F p ' values of the Cu Kα spectral peaks in 5 soil samples to be measured

[0104] Table 5 Energy values of the points corresponding to the fluorescence intensities of 3 / 4F p ', 1 / 2F p ', 1 / 4F p ' on the left side of the Cu Kα characteristic spectral peak in 5 soil samples to be measured

[0105] Step 12: Calculate and obtain the full width at three-fourths maximum (FWHM3 / 4), full width at half maximum (FWHM), and full width at one-fourth maximum (FWHM1 / 4) of the Cu Kα characteristic spectral peak of the soil sample to be measured.

[0106] Substitute the net fluorescence intensity values F p ' of the peak positions of the Cu Kα characteristic spectral peaks in the 5 soil samples to be measured extracted in Step 11 into the quantitative relationship models between W 3 / 4 and F p , W 1 / 2 and F p , W 1 / 4 and F p shown in Formulas (4)-(6) in Step 7 to obtain the FWHM3 / 4 W 3 / 4 ', FWHM W 1 / 2 ', and FWHM1 / 4 W 1 / 4 ' values of the Cu Kα characteristic spectral peaks in the 5 soil samples to be measured, as shown in Table 6 specifically.

[0107] Table 6 FWHM3 / 4 W 3 / 4 ', FWHM W 1 / 2 ', and FWHM1 / 4 W 1 / 4 ' values of the Cu Kα characteristic spectral peaks in the 5 soil samples to be measured

[0108] Step 13: Calculate and obtain the position information corresponding to the fluorescence intensities of three-fourths peak value, one-half peak value, and one-fourth peak value on the right side of the Cu Kα characteristic spectral peak of the soil sample to be measured.

[0109] Add the energy values E p ', E p ', and E p ' corresponding to the positions where the left fluorescence intensities of the Cu Kα characteristic spectral peaks in the 5 soil samples obtained in Step 11 are 3 / 4F 3 / 4L ', 1 / 2F 1 / 2L ', and 1 / 4F 1 / 4L ' (Table 5) to the FWHM3 / 4 W 3 / 4 ', FWHM W 1 / 2 ', and FWHM1 / 4 W 1 / 4 ' values (Table 6) of the Cu Kα characteristic spectral peaks of the soil sample to be measured obtained in Step 12, and calculate the energy values E p ', E p ', and E p ' corresponding to the positions where the fluorescence intensities on the right side of the Cu Kα characteristic spectral peak of the soil sample to be measured are 3 / 4F 3 / 4R ', 1 / 2F 1 / 2R ', and 1 / 4F 1 / 4R', so as to obtain the information of 3 positions on the right side of the Cu Kα characteristic spectral peak without being interfered by the Ni Kβ characteristic spectral peak [E 3 / 4R ', 3 / 4 F p '], [E 1 / 2R ', 1 / 2 F p '], [E 1 / 4R ', 1 / 4 F p '].

[0110] E 3 / 4R ' = E 3 / 4L ' + W 3 / 4 ' (8)

[0111] E 1 / 2R ' = E 1 / 2L ' + W 1 / 2 ' (9)

[0112] E 1 / 4R ' = E 1 / 4L ' + W 1 / 4 ' (10)

[0113] By calculating from formula (8) to formula (10), the fluorescence intensities on the right side of the Cu Kα characteristic spectral peak in 5 soil samples to be measured are 3 / 4F p ', 1 / 2F p ', 1 / 4F p ' corresponding to the energy values E 3 / 4R ', E 1 / 2R ', E 1 / 4R ' as shown in Table 7. The information of 3 positions on the right side of the Cu Kα characteristic spectral peak without being interfered by the Ni Kβ characteristic spectral peak [E 3 / 4R ', 3 / 4 F p '], [E 1 / 2R ', 1 / 2 F p '], [E 1 / 4R ', 1 / 4 F p '] are shown in Table 8.

[0114] Table 7 Energy values E 3 / 4R ', E 1 / 2R ', E 1 / 4R ' on the right side of the Cu Kα characteristic spectral peak in 5 soil samples to be measured

[0115] Table 8 [E 3 / 4R ', 3 / 4 F p '], [E 1 / 2R ', 1 / 2F p '], [E1 / 4R ', 1 / 4 F p '] Three point position information

[0116] Step 14: Obtain the complete spectral information of the Cu Kα characteristic spectral peak in the soil sample to be measured through interpolation.

[0117] Set the fluorescence intensity values of the spectral between the peak point (energy is E q ') of the Ni Kβ characteristic spectral peak and the right valley point (energy is E n ') in the overlapping peak of Cu Kα and Ni Kβ to 0. According to the three point position information [E 3 / 4R ', 3 / 4 F p ']、[E 1 / 2R ', 1 / 2 F p ']、[E 1 / 4R ', 1 / 4 F p '] and the peak point [E p ', F p '] of the Cu Kα characteristic spectral peak on the right side without interference from the Ni Kβ characteristic spectral peak obtained in Step 13, and the spectral information between the peak point of the Ni Kβ spectral peak and the right valley point with the fluorescence intensity set to 0 ([E p ', 0],…, [E q ', 0]), use the Lagrange interpolation method to interpolate the spectral information after the position corresponding to 3 / 4 F n ' intensity on the right side of the Cu Kα characteristic spectral peak. The number of interpolation points and the energy values of the abscissa are the same as those of the corresponding points with energies between E p '~E 3 / 4R ' in the original measured EDXRF spectrum. According to the fluorescence intensity information obtained by interpolation, determine the position information [E n ', F m '] corresponding to the minimum fluorescence intensity on the right side of the Cu Kα spectral peak. Then, after interpolation, the spectral data of the Cu Kα characteristic spectral peak in the soil sample to be measured is ([E1', F1'], [E2', F2'], …, [E m '], …, [E i ', F i '], …, [E m ', F m ']), and m < n.

[0118] For 5 soil samples to be measured, the energy range where the fluorescence data of the Cu Kα characteristic spectral peak obtained by interpolation is located is 7.8746 keV to 8.2258 keV, with a total of 53 spectral data, that is, m = 53.

[0119] Step 15: Invert to obtain the Cu content value in the soil samples to be measured.

[0120] According to the complete spectral information of the Cu Kα characteristic spectral peak in the 5 soil samples to be measured obtained in the above Step 14, calculate the net integrated fluorescence intensity value S' of the Cu Kα characteristic spectral peak, and substitute S' into the Cu content quantitative inversion model shown in formula (7) in Step 8 to invert and obtain the Cu content value in the soil samples to be measured where Ni and Cu coexist.

[0121] For the 5 soil samples to be measured, the net integrated fluorescence intensity S' values of the Cu Kα spectral peak in the EDXRF spectrum calculated from 53 spectral data are 1990.2462, 2529.6138, 3702.6553, 4108.8587, and 5004.0560 respectively. Substitute the S' values of the 5 soil samples to be measured into the Cu content quantitative inversion model shown in formula (7) in Step 8. Through inversion, the Cu content values in the 5 soil samples to be measured where Ni and Cu coexist are 24.57 mg / kg, 114.90 mg / kg, 311.37 mg / kg, 379.40 mg / kg, and 529.32 mg / kg respectively.

[0122] The flame atomic absorption spectrophotometry was used to measure the true content of Cu in the 5 soil samples to be measured, and the measured results are shown in Table 9. At the same time, the traditional Gaussian mixture model was used to analyze the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ in the EDXRF spectra of the 6 standard soil samples in Step 5. Figure 2 Re-establish the Cu content quantitative inversion model with the net integrated fluorescence intensity of the obtained Cu Kα characteristic spectral peak and the standard content value of Cu, and use the traditional Gaussian mixture model for Step 10. Figure 3The complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ in the EDXRF spectra of 5 soil samples to be measured was analyzed. The net integrated fluorescence intensity of the Cu Kα characteristic spectral peak obtained was brought into the Cu content quantitative inversion model to invert the Cu content values of the 5 soil samples to be measured. The results are shown in Table 9. It can be seen that for the 5 soil samples to be measured, when using the method of this patent to perform EDXRF detection of Cu in soil coexisting with Ni, the relative error is between 5.55% and 18.48%. Compared with the relative error of 7.22% - 22.83% for Cu content inversion after overlapping peak analysis using the traditional Gaussian mixture model method, both are reduced, by 12.44% - 26.34% respectively, with an average reduction of 20.64%. Therefore, the method proposed in this patent can perform EDXRF detection of the Cu content in soil coexisting with Ni more accurately compared with the traditional Gaussian mixture model overlapping peak analysis method.

[0123] Table 9 Comparison of Cu content detection errors between the method of the present invention and the Gaussian mixture model method

[0124] The above are only specific embodiments of the present application, enabling those skilled in the art to understand or implement the present application. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but rather will conform to the broadest scope consistent with the principles and novel features claimed herein.

Claims

1. An accurate EDXRF detection method for Cu content in soil coexisting with Ni and Cu, characterized in that, It includes the following steps: (a)Prepare a series of standard soil samples with different Cu contents using soil with low Ni content; (b)Determine the standard content value of Cu in the standard soil samples by flame atomic absorption spectrophotometry; (c)Measure the EDXRF spectra of the standard soil samples under the same conditions, and perform denoising and background subtraction to obtain the net EDXRF spectra; (d)Extract the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ; (e)Based on the spectral data of the standard soil samples, calculate the full width at three-fourths height, full width at half maximum, and full width at one-fourth height of the Cu Kα spectral peak by Lagrange interpolation; (f)Establish a quantitative relationship model between the full width at three-fourths height, full width at half maximum, and full width at one-fourth height of the Cu Kα spectral peak and the net peak fluorescence intensity; (g)Based on the net integrated fluorescence intensity of the Cu Kα spectral peak and the Cu standard content value of the standard soil samples, establish a quantitative inversion model for the Cu content in the soil samples; (h)Prepare the sample of the soil sample to be measured, measure the EDXRF spectrum, perform spectral denoising and background subtraction, and extract the complete spectral information of the overlapping spectral peaks of Cu Kα and Ni Kβ; (i)Determine the corresponding positions of the three-fourths, one-half, and one-fourth peak intensities on the right side of the Cu Kα spectral peak of the soil sample to be measured by interpolation and calculation; (j)Obtain the complete spectral information of the Cu Kα of the soil sample to be measured by interpolation; (k)Invert the Cu content in the soil sample to be measured by combining the quantitative model.

2. The method according to claim 1, wherein The preparation of the standard soil samples in step (a) includes: adding copper nitrate solutions with different concentrations to the soil with low Ni content, drying, grinding, sieving, and pressing into round samples with a flat surface to ensure consistent physical properties.

3. The method according to claim 1, wherein In step (c), Savitzky-Golay filtering is used for spectral denoising, and background subtraction is achieved by fitting a cubic smoothing spline curve with peak valley recognition and penalty term correction.

4. The method according to claim 1, characterized in that, In step (e), the Lagrange interpolation method is used to obtain the spectral data of the three-fourths, one-half, and one-fourth peak intensity points on the left and right sides of the Cu Kα spectral peak, and based on this, the full width at three-fourths height, full width at half maximum, and full width at one-fourth height of the Cu Kα spectral peak are calculated.

5. The method according to claim 1, wherein The formula of the quantitative relationship model in step (f) is as follows: When F p <125.3709, W 3 / 4 =0.1715 - 0.002 F p + 1.13503 10 -5 F p 2 ; When F p ≥ 125.3709, W 3 / 4 = 0.12802 - 3.28927 × 10 -4 F p + 7.8116 × 10 -7 F p 2 ; It should be noted that the original text seems to have some unclear or incomplete expressions in terms of mathematical relationships. The translation is done as accurately as possible based on the provided text. When F p < 125.3709, W 1 / 2 = 0.04013 + 0.00167 F p - 5.42624 10 -6 F p 2 ; When F p ≥ 125.3709, W 1 / 2 = 6571.05218 exp( - F p / 9.30544) + 0.15514; When F p <125.3709, W 1 / 4 =0.24136 - 8.21375 ×10 -4 F p +5.35149 ×10 -6 F p 2 ; When F p ≥ 125.3709, W 1 / 4 = 4.73243 exp( - F p / 20.28833)+0.21268.

6. The method according to claim 1, characterized in that In step (g), the quantitative inversion model is a linear equation C = a S + b, where S is the net integrated fluorescence intensity of the Cu Kα spectral peak, and a and b are determined by fitting with standard samples.

7. The method according to claim 1, characterized in that, In step (i), the energy values corresponding to the peak intensities at 3 / 4, 1 / 2, and 1 / 4 on the right side of the Cu Kα spectral peak are calculated by the formulas E 3 / 4R ' = E 3 / 4L ' + W 3 / 4 ', E 1 / 2R ' = E 1 / 2L ' + W 1 / 2 ', E 1 / 4R ' = E 1 / 4L ' + W 1 / 4 ', where W 3 / 4 ', W 1 / 2 ', and W 1 / 4 ' are obtained from the quantitative model in step (f).

8. The method according to claim 1, characterized in that, The relative error of inverting the Cu content in step (k) is 5.55% - 18.48%, which is on average 20.64% lower than that of the traditional Gaussian mixture model.

9. The method according to claim 1, characterized in that, The EDXRF spectrum measurement conditions include: an Ag target X-ray tube, a voltage of 40 kV, a current of 20 μA, and an accumulation time of 120 s.

10. The method according to claim 1, wherein The method is applicable to the detection of Cu in soil with a content range of 24.57 mg / kg to 529.32 mg / kg.

Images