SU1691724A1 - X-ray fluorescent analysis of multicomponent sample containing n elements to be determined - Google Patents

Abstract

The invention relates to X-ray spectral methods for analyzing the composition of a substance. The purpose of the invention is to improve the accuracy of the analysis. Three groups of calibration samples are prepared for each detectable element A such that in all samples of the first and second groups a mass attenuation coefficient of the characteristic radiation, the corresponding analysis. The invention relates to X-ray spectral methods for analyzing the composition of the substance. The purpose of the invention is to improve the accuracy of the analysis. The drawing shows a graph of the calibration dependencies. To implement the method, calibration samples are prepared, they are irradiated and the sample under examination with x-rays, the secondary radiation of the samples is recorded, and the concentrations of the detected elements are calculated. The tic line of element A is equal to the clock one in element A, while in samples of the first group the content of element A varies in the interval (0.100%) and there are no elements whose characteristic radiation excites the analytical line of element A. The samples of the second group contain a constant number element A, and the content of element B, whose characteristic radiation excites the radiation of the analytical line of element A, changes in them from zero to the maximum value satisfying the above condition from the expected content of element B in the analyzed samples. Samples of the third group do not contain element A and consist of element C, which is the element with the maximum expected content in the analyzed sample, and one of the determined elements X. The calibration and the analyzed sample are irradiated with X-rays and the secondary radiation is recorded. According to the obtained data, the concentrations of the determined elements are calculated. 1 ill., 1 tab. Three groups of synthetic reference materials are prepared for each definable element A, and the mass attenuation coefficient of the analytical line i of element A in all samples and the second group is equal to the clock one in element A (i 1N) From samples of the first group in which the content of element A varies from zero to 100% (mass fractions) and there are no elements, the characteristic radiation of which excites the analytical line i of element A, measure the intensity of this EO O VI th

Description

of the analytical line I of element A. Based on the known contents of element A in the synthetic standard samples of this group and measured intensities, a calibration characteristic is created — the content as a function of the intensity of the analytical line i of element A, For convenience, these contents are called the conditional contents Gi, and the resulting dependence - analytical chart. An analytical graph, for example, can be written in the form of a second-order polynomial

Gi aiRi2 + biRi + d, (1)

where RI is the relative intensity of the characteristic radiation of the analytical line i, found as the ratio of the measured intensity (after subtracting the background) to the same intensity from a standard sample of the same group containing 10% of element A. Because of the limitations imposed on the composition of these samples the intensity of the background from them is constant and equal to the intensity measured from a standard sample of the same group with zero content of element A; ai, bi, and di are the approximation polynomial parameters calculated from this experiment.

To improve the accuracy of the approximation, the experimental curve G f (Ri) is divided into segments, which are then approximated either by a polynomial, or by an exponential, or by any other function.

The conditional Gi content is related to the Ci content in the sample by the ratio

 mi

WITH;

(2)

mi

where is the mass attenuation coefficient of the line I radiation in the sample x, or in the atoms of the detected element A.

From samples of the second group containing a constant amount of element A, the contents of element B, the characteristic radiation of which excites the radiation of the analytical line i of element A, changes from zero to the maximum number of Cm from the range of expected maximum contents of element B in the analyzed samples RI.X intensity of analytical line i of element A and R "

-analytical line K of element B and build a second calibration characteristic

-the change in the relative intensity RI / RK as a function of the relative intensity RK of the characteristic radiation of element B, where RI is the intensity of the standard with zero content B. Such a construction of dependence allows

0

50

five

0

to take into account the effect of selective excitation of secondary radiation of element A by the radiation of atoms of element B and the effect of selective absorption of primary radiation by atoms of element B for any variations in the contents of element A and B and the mass attenuation coefficient. Such a calibration characteristic can be written, for example, analytically as the relative intensity corrected for these effects.

Ri Ri.x (aKRK2 + bkr + 1); (3)

where RI.X is the relative intensity in which correction has not yet been introduced;

ak, bk are the approximating polynomial parameters found from this experiment.

The background is determined by measuring the intensities from standard samples of the same group with zero content of the corresponding element. To ensure the requirements for the composition of the samples of both groups, chemical reagents D and F (chemical elements or chemical compounds that do not contain elements A and B) are added to each sample, selected so that the mass absorption coefficients in them / rR 1.0, a / nf 1.0. The mass fractions of Co and CF at given contents of SA and C are found by solving a system of equations

CdSA + SvSv + Co + Cf 1.0; (4)

five

0

five

0

five

KSL

ft A

L to with

J 1 & 6

PR g JM- + &.

.

ji / rJMK, j0

D J i Fs

where Kd is the coefficient inversely proportional to the magnitude of the numerical value of the mass fraction of element A in the chemical reagent used;

ij are the mass absorption coefficients;

RA and PB - indices, meaning absorption in a reagent containing element A or B.

Denote by Ri / Rix; the constant V with the index denoting the element, the maximum expected content of this element in the analyzed samples; the main element is an element for which the content of V exceeds the same values for all other elements of the sample.

Analytical plots are plotted for (0.100)% intervals, or for (O, VA)% content of element A.

The graphs for plotting (3) are plotted for the interval (0, Cm), where Cm is defined as follows.

The system of equations (4) has positive solutions for the interval (O, WB)%

the contents of element B. Therefore, two situations are possible

1) L / in VB, then Cm - VB;

2) WB VB, then Cm WB, and to provide corrections in the whole range of contents (O, VB), the obtained relationship (3) is extrapolated to the RK value corresponding to the VB value.

The drawing shows the relation (3) under consideration for the case of steel analysis. In this case, the main element is iron, VB 100% (RB 5.63). Analytical line СГК (the greatest selective effect).

Reagents: chromium oxide, iron powder, sulfur, boric acid.

When Csg 1% content Wpe 80% (Rpe 5.01). In the drawing, the crosses are plotted with experimental values for the interval (0; 5.01). The solid curve is the approximation of the experiment, the points are the extrapolatingly established dependence (3).

The greatest selective effect is observed for neighboring elements, for which the mass absorption coefficients

always / rG1 / rR, and this guarantees a WB content of 80%. With a significant difference in the atomic numbers ZA and ZB, the value of HA almost approaches unity. Consequently, the error of extrapolation in all cases does not affect significantly the results of the analysis.

If the chemical element called the main one is designated Nm, then its content CN can be determined by the formula

N -1

 Cj (5)

j 1

This makes it possible to refuse to measure the intensity of its analytical line in the analyzed samples. For samples consisting of N chemical elements, the emission intensities of N-1 chemical elements are measured,

From samples of the third group, not containing elements A and consisting of a main element and one of the determined elements X, whose contents vary from zero to Vx, measure the intensities Ri, x - analytical line i (background) and RJ.X - analytical line j of the element X and constructing a third calibration characteristic — measuring the intensity of the analytical line i (background) of the element A as a function of the intensity of the analytical line j of the element X. This calibration characteristic provides an introduction of a correction to the background change of the analyte cal line, due both to the imposition of

characteristic radiation of element X and variations in the chemical composition of the sample. As an example, such dependence can be represented by the formula

Ri.x RI.O + 5ts Rjx. (6)

where RI.O is the background from the main element on the analytical line i;

d - parameter approximating polynomial.

The analytical relationship (6) is presented as an example to illustrate the method. In other cases, the dependence RI.X f (Ri, o; Rj.x) can be approximated by any other analytical function.

Calibration using samples of the second group RI f (Ri, x; RK) opened up the possibility of using relation (2) for each of the N-1 analytical lines of the analyzed sample, ensuring that Ci is calculated by the value

GI and coefficient // G is the mass attenuation coefficient of the analytical line i in the analyzed sample.

To determine the coefficient / and use the law of the adequacy of the attenuation of x-rays

o..Qiicjju;

mL. J w 1 jj

.

J4

J PP

(7)

where N is the number of all sample elements;

// Ini — mass attenuation coefficient of the analytical line i (subscript) in atoms of only one element j (superscript).

The coefficients are // L-constant values, therefore the ratio of these coefficients / l is constant. The calculation of the contents is performed according to formula (2) after substituting relations (5) and (7) into it. This formula (2), written for each of the N-1 definable elements, allows

get a system of algebraic equations

IM

/V(w-i)JVo.o, 2 L); GK,

T -

where i 1, 2, 3N-1; С is the content of the element being determined, mass fractions,%.

The system of equations is solved, for example, by the Gauss method.

To account for mode fluctuations, you can

use in the process of mass analysis reference standard samples. In the process of mass analysis of samples, the calculation of the contents is carried out in the following order. The calculation of the GI conditional content is performed as shown below

the formulas for the sequential processing and the measured intensity li of the analytical line I containing the background component. A correction for errors caused by the dead detector time is preliminarily introduced into the intensity li.

The relative intensity Rix of the useful signal, independent of the operating mode of the device, is calculated by the formulas

1st li - IIN; (9)

N

Rix lioDi + 2 5ij RJ.X, (10)

J 1

n

where IIN is the measured intensity due to the main element;

DI - calibration parameter of the instrument operation mode, calculated from the intensity of the reference samples in one of the described methods;

5ji is the same as in formula (6).

Corrections for the effects of selective excitation of secondary radiation and the effect of selective absorption of primary radiation are introduced by the formula

Ri (aKRKx2 + bKRKx) +1, (11)

where K is the summation index over all elements that create effects, the remaining symbols are similar to the clock ones for formula (3).

The correction to the filtering of the primary radiation by the atoms of the element being detected is the last step in calculating the conditional GI content using formula (1). After that, by solving the system of equations (8), the content of all N-1 elements of the sample is found, and by the formula (5) - the content of the N-ro element.

The formula (11) includes the relative intensity of RNX. In the first cycle of calculations, it is assumed to be equal to an arbitrarily chosen value. After calculating the content of directly determined elements by the formula (5), Cm is calculated, the mass attenuation coefficient of the analytical lines N in the sample and, finally, the conditional content GN by the formula (2). This further allows us to find RNX using formula (1) and (11). The calculation cycle is repeated several times with the new RNX until the minimum of the trace is reached. See Practically the number of cycles does not exceed three, four.

The DI parameter for calibrating the instrument's mode of operation is equal to the reciprocal of the lip intensity from the sample that is used to construct the analytical graph (1), and containing 10% element A. This parameter

depends on the same variables as the coefficients, and, moreover, on the tuning state of the photon detection channel, and therefore is characteristic of a particular instrument and at the same time independent of the chemical composition of the samples and reference standard samples. With modern equipment, the intensity remains stable within the statistical scatter of the data over many months. This determines the frequency of lip measurement. In most cases, it is practical to obtain samples for directly measuring lip or

5 is very difficult, or almost impossible. Therefore, to calculate the parameter, it was proposed to use certified reference standards with an arbitrary chemical composition belonging to the class of substances being analyzed. One reference standard sample can provide intensity determination for all analytical lines. It is practically more convenient to use several benchmark reference samples in order to ensure the content of element A in the range of 5-20%, or the highest expected content of element A in the analyzed samples. For selected reference points

0 standard samples according to the content of Ci and according to the calculated mass attenuation coefficient of the line I radiation in them calculate the conditional content of GI by the formula (2) and further by the formulas (1) and (11) determine

5 relative intensity rjx. By measuring the intensity h from the reference standard sample on the instrument, the parameter DI is calculated using formulas (9) and (10). The calculation process Di envisages the content calculation algorithm using a specific method. Thus, reference standard samples serve only to assign the operating mode of a specific instrument.

Calibration characteristics (1), (3),

5 (6), (8) are not related to the quantitative composition of the sample being analyzed and therefore are applicable for arbitrary variations from zero to 100% of the chemical composition of the sample. Correct calculation of the mass attenuation coefficient in combination with the equations of communication, the coefficients of which are not related to the chemical composition of the sample and the mode of operation of the device, ensures the determination of the content of all chemical elements

5 samples in the range of contents that can fundamentally provide a direct X-ray spectral analysis method.

Improving the accuracy of the analysis results is achieved by directly calculating the mass attenuation coefficient.

radiation of each analytical line directly in the analyzed sample, introduction of corrections to the effects of selective excitation of the secondary and selective absorption of the primary radiation, filtration of the primary radiation by the atoms of the element being determined in accordance with the full specific chemical composition of the analyzed sample, calculation of the background component and tax other lines based on the specific composition of the sample being analyzed.

The method was tested on the example of the analysis of steel. The analysis of steels containing N 16 chemical elements, including: molybdenum, niobium, zirconium, tungsten, copper, nickel, cobalt, manganese, chromium, vanadium, titanium, sulfur, phosphorus, silicon, aluminum, and the basic element iron. In the experiment, a device with a x-ray tube made of rhodium (anode) was used. The voltage on the tube is 50 kV. The analytical lines of 15 elements (including the main one) are Ka and WLa radiation. Eight standard images were used as reference points. Three groups of synthetic reference materials were prepared for each element and the parameters ai, bi, di, ak, bk, and d were calculated from them.

Consider preparing samples for manganese. Analytical line MpKa. For the preparation of the first and second groups of synthetic reference materials, chemical reagents were used: manganese dioxide, copper oxide, sulfur and boric acid.

The first group of samples was prepared in the following way.

For a given manganese content in the sample, the content of sulfur and boric acid was calculated using formula (4). The number of such samples (set) required to construct one analytical graph (1) was chosen based on the condition of providing a statistically reliable determination of the coefficients ai, bi, and di. The proposed method for introducing corrections makes it possible to effectively use the method of interpolation and extrapolation of data when determining these coefficients. As a result, the actually required number of samples of all three groups is reduced by 4-6 times. This can be illustrated by this example. The first measurements were carried out to build analytical MoKa.MpKa graphs and AlKct. The data on these three graphs were plotted on one graph G f (R) and from these data the parameters were uniform (for these three

graphs) approximating curve and the corresponding approximation error. This error did not exceed the approximation error when calculating the parameters of the curves, for each individual analytical line. This gave a complete basis for all the analytical lines of Ka to be considered as an analytical graph of a single approximating curve. In this case, to construct 15 analytical plots (1), only three sets of samples of the first group, rather than 15, and a fourth set for the WLa analytical line were required for the emission of a CD.

The second group of samples prepared as follows.

For a given manganese content, for example, Cmp 0.20, samples were prepared with different specified copper contents, and the sulfur and boric acid contents were calculated using formula (4). The manganese content was chosen differently and then a new set of samples with different copper contents was calculated again.

The intensity of the characteristic radiation MpKa from the first and second groups of synthetic standard samples was obtained by subtracting the background intensity, measured from the same manganese-free samples, from the line intensity measured on the instrument. Similarly, the intensity of the characteristic radiation of CuKa from the second group of samples was determined using the same samples that did not contain copper to determine the background.

To calculate the RK ratio, an additional two synthetic standard samples were used: one containing 10% copper and the other not containing copper (to measure the background intensity from the first sample). These samples belong to the first group of synthetic reference materials for constructing an analytical graph of CuCa.

When constructing a functional relationship (3), two graphs were originally constructed for the elements of the type: B, C, and Co. The approximation curves plotted on one graph were a family of curves formed by rotating one parabola around the point (0,1). This gave reason to believe that a similar curve for the element B - nickel is located between these curves. By interpolation, parameters were found for this dependency. Thus, to build all the functional dependencies (3), it took about 30 sets of samples of the second group.

The third group of samples is binary mixtures of iron with one of the determined elements. Powdered materials have been used for these purposes .5

In general, it took about three hundred samples to determine the parameters of the entire system.

51 State Standard Samples (GSO) of steels 10 on a quantometer were analyzed. When comparing the results of the analysis with the certified data of the State Standard Samples, it was established that there is no statistically significant correlation between the error of the results of the analysis and the chemical composition. The average value of the six parallel determinations and the certified data for the two samples are presented in the table.

Claims (1)

Formula 20 X-ray fluorescence analysis method for a multicomponent sample containing N detectable elements, including irradiating the test and calibration samples with x-rays, determining in the secondary radiation spectra of the test sample the intensities of the analytical lines of N + 1 detectable elements, determining the intensities of the characteristic radiation elements in the secondary radiation spectra of calibration samples and the determination of the concentrations N of the elements being determined, characterized in that To improve the accuracy of analysis, as a calibration, three groups of samples are prepared for each definable elei of the second group; the mass coefficient of fading of the characteristic radiation of the corresponding analytical element A is equal to the clock element in the element; in samples of the first group, the element A changes in the inter from 0 to 100% and there are no elements, the characteristic radiation of which produces an analytical line element; samples of the second group contain a fixed amount of element A, and contain element B, which is characteristic The emittance of which excites the emission of the analytic line of element A changes to zero to satisfy the specified condition imposed on the values of the absorption coefficients, the maximum values from the range of expected contents of element B in the analyzed samples of the third group do not contain ment A and m from element C, in the quality of which one selects the element with the maximum expected content in the analyzed sample, and one of the defined element X, whose content varies from zero to the maximum expected content of this element in ana in the test, in the process of measurements, the intensity of the analytical element A in the spectra of the secondary radiation of the samples of the first group, the intensity of the analytical lines of the elements A and the spectra of the secondary radiation of the second group, the intensities of the analytical lines of the elements C and X in the spectra of the samples are determined three groups, and on the composition of the sample judged by the totality of the data. ment A such that in all samples of the first Results of X-ray analysis (top line) and certified GSO data (bottom line) and the second group, the mass attenuation coefficient of the characteristic radiation corresponding to the analytical line of element A is equal to the clock one in element A, while in the samples of the first group the content of element A varies from 0 to 100% and there are no elements whose characteristic radiation excites the analytical line of element A, samples of the second group contain a constant amount of element A, and the content of element B, the characteristic radiation of which excites the radiation of the analytical line of element A, is measured In them, from zero to satisfying the specified condition imposed on the values of the absorption coefficients, the maximum value from the range of expected contents of element B in the analyzed samples, the samples of the third group do not contain element A and consist of element C, which is chosen as element C the maximum expected content in the analyzed sample, and one of the determined elements X, whose content varies from zero to the maximum expected content of this element in the analyzed sample, in the process measurements determine the intensity of the analytical line of element A in the spectra of secondary radiation of samples of the first group, the intensity of analytical lines of elements A and B in the spectra of secondary radiation of samples of the second group, the intensity of the analytical lines of elements C and X in the spectra of secondary radiation of samples of the third group, and on the sample composition judge t on the totality of the data. approximate (uncertified) GSO data