JPH11352080A - Fluorescent x-ray analyzer - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enhance the efficiency of processing when automatically separating approximate spectrum lines and analytical accuracy and to shorten analysis time in a fluorescent X-ray analyzer for detecting a peak from a measured spectrum waveform to identify the spectrum line of an element. SOLUTION: A deciding means (step n8) for deciding whether the interval D between left and right peaks approaching each other is smaller than a preset threshold η1 of the trough between both peaks is larger than a preset threshold η2 and a peak separation means (steps n12-n14) separating both peaks by the calculation of a function fitting, only when both conditions (D<=η1 as well as I>=η2 ) are decided to be satisfied by the deciding means are provided.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an X-ray fluorescence spectrometer, and more particularly, to a technique for separating overlap of adjacent peaks.

[0002]

2. Description of the Related Art Generally, a fluorescent X-ray analyzer converts a fluorescent X-ray generated by irradiating a sample surface with primary X-rays into a spectrum having a wavelength component corresponding to each element by using a plurality of spectral crystals. The X-rays are separated and detected by an X-ray detector to obtain a spectrum waveform. Alternatively, the fluorescent X-rays from the sample are detected by a semiconductor detector, and energy selection (wave height analysis) is performed by an energy selector, and the X-ray intensity at each energy is counted to obtain a spectrum waveform.

The spectral waveform thus obtained is subjected to a smoothing process for removing a statistical error, a background removal for removing background intensity contained in measured data, and a peak position / peak detected by detecting a peak in the measured data. A peak detection process for calculating the intensity / background intensity is performed.

[0004] Further, with respect to the detected peaks, the names of the elements and the names of the spectral lines are usually identified in the descending order of peak intensity using a wavelength table stored in advance, and the other spectral lines of the element are identified. Is also identified.

At this time, the intensity of the spectrum line occupying the peak is estimated and subtracted by using the line intensity ratio table of each element stored in advance, and if there is a residue, the spectral lines of the other elements are overlapped. Assuming that there is a peak, the identification is advanced again for the residual peak, and the identification process is repeated for all the peaks to obtain the element name, spectrum line name and intensity value for each peak.

In the above-described peak detection processing, there is a state where the spectral lines of a plurality of elements are close to each other at a specific position and a part of the peaks overlap with each other. Spectral lines affect each other and increase in intensity over the actual intensity.

On the other hand, the line intensity ratio of the line intensity ratio table used when calculating the intensity of the spectral line of each element is obtained in a state where there is no influence of spectral lines adjacent to each other.

For this reason, when the intensity of the spectral line of each element is calculated using the line intensity ratio of the line intensity ratio table as it is, despite the fact that the peaks adjacent to each other are partially overlapped, The intensity values of the spectral lines obtained as a result of the identification process will include an error.

In order to eliminate such an error, when overlapping peaks occur adjacent to each other, it is necessary to separate the two peaks and remove the influence of the mutual peaks.

Therefore, in the prior art, function fitting processing is performed as such peak separation processing.

In this function fitting processing, a curve given by a function such as a Gaussian function or a Lorentz function is enlarged or reduced to fit one detected peak, and a peak position and a peak intensity of the curve given by the fitted function are obtained. After subtracting the value of the fitted function from the detected peak to remove the peak,
Further, peak separation and peak intensity are sequentially performed by similarly performing a function fitting process on the remaining detected peaks to obtain a peak position and a peak intensity.

[0012]

By the way, when a plurality of detected peaks exist close to each other, particularly,
There is no need to perform peak separation processing.

For example, as shown in FIG. 4A, even if there are two peaks P ₁ and P ₂ close to each other, when each of the peaks P ₁ and P ₂ is steep and the half width is small, Since there is little influence of the overlapping of the peaks P ₁ and P _{2 at} the foot portions thereof, there is no need to perform peak separation processing.

Further, as shown in FIG. 4B, when the background removal processing is performed in advance, the effect of the background existing over each of the peaks P ₁ ′ and P ₂ ′ is not sufficiently removed. It appears that there is an overlap at the base of both peaks P ₁ ′ and P ₂ ′. In this case,
What is necessary is just to repeat a background removal process, and there is no need to dare to perform a peak separation process.

However, conventionally, there is no clear standard for determining whether or not the peak separation processing should be performed. Therefore, as shown in FIGS. If there are peaks that are close to each other, including unnecessary cases, all peak separation processing is performed unconditionally.

Since the function fitting calculation for such a peak separation process takes a certain amount of time, extra time is required for the peak separation.
Inconveniences such as a long time until the final identification result is obtained have been caused.

The present invention solves the above-mentioned problems, improves the efficiency of automatically separating adjacent spectral lines, and determines the spectral line intensity more accurately and within a short time. It is an object to improve the analysis accuracy and shorten the analysis time.

[0018]

According to the present invention, there is provided an X-ray fluorescence spectrometer for detecting a peak from a measured spectral waveform to identify a spectral line of an element in order to achieve the above object. It is composed.

That is, according to the present invention, when a plurality of peaks are detected, the distance D between the right and left peaks adjacent to each other is determined.
Is smaller than a predetermined threshold value η ₁ , and the valley intensity I between the two peaks is smaller than a predetermined threshold value η ₂
Means for judging whether or not the two peaks are larger than, and only when it is judged that both conditions (D ≦ η ₁ and I ≧ η ₂ ) are satisfied, the two peaks are separated by the function fitting calculation. And a peak separating means.

According to this configuration, when a plurality of peaks are detected, it is determined whether or not the peaks adjacent to each other satisfy a certain condition. Since the lines are separated, unnecessary peak separation processing can be omitted, and the intensity value of the spectrum line close to the true value can be obtained more efficiently.

[0021]

FIG. 1 is a schematic configuration diagram of an X-ray fluorescence analyzer according to an embodiment of the present invention.

In the figure, 1 is an X-ray tube, 2 is a sample, 3
Is a primary solar slit, 4 is a spectroscope, 5 is a secondary solar slit, 6 is an X-ray detector, 7 is a device that analyzes and processes data and has a function as a separating unit that separates adjacent spectral lines as described later. A computer 8 is a display unit such as a CRT for displaying analysis results.

In such a fluorescent X-ray analyzer, primary X-rays from the X-ray tube 1 are irradiated on the sample 2, and correspondingly, fluorescent X-rays generated from the sample 2 are transmitted through the primary solar slit 3 to the spectroscope 4. And split the X-rays into a secondary solar slit 5
Is detected by the X-ray detector 6 via the.

In the computer 7 to which the output of the X-ray detector 6 is given, the data is analyzed according to the procedure shown in the flowchart of FIG.

Next, the data analysis processing will be described specifically.

The measured data of the measured spectrum waveform is subjected to a smoothing process to remove a statistical error (step n1), the background intensity included in the measured data is removed (step n2), and the measurement is further performed. Detect peaks in the data and perform peak detection processing to find the detected peak positions and peak intensities (step n
3).

Next, in order to specify each peak detected as described above, for example, serial numbers N are assigned in order from the peak with the smaller wavelength, and the peak number N is first set to "1" (step n4). ). At first, M is set to "1" as a count value M for counting the number of peaks requiring peak separation (step n5).

Subsequently, as shown in FIG. 3, for two adjacent peaks P _N and P _{N + 1} , the wavelength interval D = |
λ _{N + 1} −λ _N | is obtained (step n6), and furthermore, the intensity I of the valley between the two peaks P _N and P _{N + 1} is obtained (step n6).
7).

Next, the wavelength interval D obtained in step 6 is smaller than a predetermined threshold value η ₁ , and the valley intensity I obtained in step 7 is set to a predetermined threshold value η.
_It is determined whether it is larger than ₂ (step 8). That is, it is determined whether or not the conditions of D ≦ η ₁ and I ≧ η ₂ are satisfied.

Here, as shown in FIG. 4A, when there are _two peaks P ₁ and P ₂ close to each other, the condition that the wavelength interval D ₁ is smaller than the threshold value η ₁ is satisfied. However, when the peaks P ₁ and P ₂ are steep, the half width is small, and the influence of the overlap at the foot of each peak P ₁ and P ₂ is small, the valley intensity I ₁ is threshold. Since the value becomes smaller than the value η ₂ and does not satisfy the condition of I ≧ η ₂ , it can be determined that there is no need to perform the peak separation process at this time.

As shown in FIG. 4B, even if background removal processing is performed in advance, each peak P ₁ ′, P ₂ ′
Is not sufficiently removed, the condition that the valley intensity I ₁ is larger than the threshold η ₂ is satisfied, but the wavelength interval D ₁ is equal to the threshold η ₁ When it is larger than the above, the condition of D ≦ η ₁ is not satisfied. In this case, the background removal may be performed again, and it can be determined that there is no need to perform the peak separation processing.

Therefore, in step 8, only when the conditions of D ≦ η ₁ and I ≧ η ₂ are satisfied, it is determined that the peak separation processing by the function fitting calculation is necessary.

Therefore, the count value M for counting the number of peaks requiring peak separation and the peak number N are both increased by one (step n9 and step n9).
10) Return to step n5.

This step n6 → step n7 → step
By repeating the routine of n8 → step n9 → step n10, peaks which need to be subjected to the peak separation processing because the adjacent peaks overlap and influence each other are specifically specified. Therefore, Steps 6 to 10 correspond to the determining means in the claims.

When the peaks requiring the peak separation processing are specified in this way, next, in step n11, it is determined whether or not the count value M for counting the number of peaks requiring the peak separation is "2" or more. Is determined (Step 1
1).

Here, if M ≧ 2, it means that there are a plurality of adjacent peaks that need peak separation processing, so that the fitting range is determined (step n1).
2) A function fitting calculation is performed on the data in the range (step n13).

In this function fitting calculation, fitting is performed using a VOIGT function represented by the following equation, which is a composite function of a Gaussian function and a Lorentz function, as a theoretical function.

[0038]

(Equation 1)

Here, the fitting parameters are:
c is the center of the peak, H is the height of the peak, G is the Gaussian factor (0 <G <1), W is the half width, and θ is the angle.

By performing the above function fitting calculation, for example, in the case of FIG. 3, each peak P _N , P _{N + 1}
Positions λ _N , λ _{N + 1} and peak intensities I _N , I _{N + 1}
Is obtained from the fitting function, and is re-registered with the result of the peak detection processing obtained in step 3 to update the data (step n14).

Accordingly, the above steps 12 to 1
4 corresponds to a peak separating means in the claims.

When the peak separation processing is completed, the step
After n15, it is determined whether or not all necessary processing has been completed for all detected peaks.

On the other hand, if M = 1 in step 11, it means that there is no peak necessary for peak separation, and the process is similarly completed for all detected peaks after step n15. It is determined whether or not.

If the processing for all the detected peaks has not been completed, N is further increased by one (step n16), and the process returns to step n5.

When the processing for all the detected peaks is completed, the names of the elements and the names of the spectral lines are identified for the detected peaks using the wavelength table and the line intensity ratio table stored in advance. (Step n1
7) Output the result of the element identification processing to the display unit 8 and end the analysis processing.

[0046]

According to the present invention, if there are peaks that are close to each other, the peak separation processing is not performed unconditionally by the function fitting calculation but the peak separation processing is performed only when a specific condition is satisfied. It is possible to increase the efficiency of automatically separating the adjacent spectral lines.

Therefore, the spectral line intensity can be determined more accurately and within a short time, so that the analysis accuracy can be improved and the analysis time can be shortened.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of an X-ray fluorescence analyzer showing an embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of the embodiment of FIG. 1;

FIG. 3 is a spectrum waveform diagram showing a state in which peaks of a plurality of spectral lines overlap.

FIG. 4 is a waveform diagram for explaining a case where there is no need for peak separation processing even when a plurality of spectral line peaks exist.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 X-ray tube 2 Sample 4 Spectrometer 6 X-ray detector 7 Computer

Claims (1)

[Claims] 1. An X-ray fluorescence spectrometer for detecting peaks from a measured spectral waveform to identify spectral lines of an element, wherein when a plurality of peaks are detected, an interval D between right and left peaks adjacent to each other is determined. preset smaller than the threshold value eta _1, and means for determining whether the intensity I of the valley between the two peaks is greater than the threshold value eta ₂ which is set in advance, both by the determination means An X-ray fluorescence analyzer comprising: a peak separating unit that separates both peaks by function fitting calculation only when it is determined that the condition (D ≦ η ₁ and I ≧ η ₂ ) is satisfied. .