JPH071311B2 - Method for peak separation of fluorescent X-ray spectrum - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for separating peaks (waveforms) from a fluorescent X-ray spectrum for separating peaks in the fluorescent X-ray spectrum.

[0002]

2. Description of the Related Art In fluorescent X-ray analysis, a sample is irradiated with radiation and the fluorescent X-ray generated from the sample is measured by an X-ray detector.
The elemental analysis of the sample is performed based on the measured value by the X-ray detector. An example of a measuring instrument used for such analysis is shown in FIG.

In FIG. 3, an X-ray tube 50 emits a primary X-ray B1 and irradiates a sample 51 with the primary X-ray B1. The irradiated primary X-ray B1 excites the atoms of the sample 51 to generate fluorescent X-ray B2 peculiar to the element. The fluorescent X-ray B2 from the sample 51 is
Only the fluorescent X-ray B2 having a predetermined wavelength that satisfies the Bragg equation below is incident on the dispersive crystal 52 and is diffracted at a diffraction angle θ that is the same as the incident angle θ. 2 dsin θ = nλ d: interplanar spacing of crystal λ: wavelength of fluorescent X-ray n: order of reflection

The above-mentioned diffracted fluorescent X-ray B3 is an X-ray detector.
It is incident on 53 and detected. On the other hand, by driving a goniometer (not shown), the dispersive crystal 52 and the X-ray detector 53 are set to 1: 2.
At a predetermined spectral angle 2θ (for example, every 0.02 °) by rotating the fluorescent X-ray B3 at an X ratio of X.
Measure the line strength.

FIG. 4 (a) shows a fluorescent X-ray spectrum obtained by measuring the components at a spectral angle 2θ. In this spectrum, the peak p1 is CrKβ ₁ from the spectral angle 2θ.
The line shows that peak p2 is the MnKα line.

However, since the two peaks p1 and p2 are overlapped as shown in FIG. 4 (b), the actual peak value (X-ray intensity) is immediately determined from the fluorescent X-ray spectrum of FIG. 4 (a). Can't figure out. Therefore, when performing X-ray fluorescence analysis, as shown in Fig. 4 (b), peaks (waveforms) p1 and p2
Need to be separated.

[0007]

The energy dispersive X-ray fluorescence analyzer is inferior in resolution to the wavelength dispersive X-ray fluorescence analyzer shown in FIG.
Considering the two peaks p1 and p2 as a symmetrical Gaussian function,
The peaks can be separated by roughly matching the sum of the two Gaussian functions with the waveform of the measured fluorescent X-ray spectrum. However, the peak p1 in Fig. 4 (a) is composed of multiple lines, and C is added to the bottom p11 of the peak p1.
The rKβ ₅ line is included and its shape is not symmetrical.
Moreover, the intensity ratio between the lines (for example, between CrKβ ₁ and CrKβ ₅ ) is not known exactly. Therefore, it was difficult to match the actually measured peaks p1 and p2 with the conventional bilaterally symmetric Gaussian function, and the peaks could not be separated accurately.

In particular, when measured with the wavelength dispersive measuring instrument of FIG. 3, the resolution is better than that of the energy dispersive type, and the distortion of the peak due to the multiple lines appears remarkably. Therefore, since each peak p1 and p2 does not match the Gaussian function, accurate peak separation was difficult.

An object of the present invention is to provide a method for separating peaks in a fluorescent X-ray spectrum, which enables accurate and simple peak separation in fluorescent X-ray analysis.

[0010]

In order to achieve the above object, the present invention provides a plurality of types of reference samples each made of a single element or compound, and a plurality of types of reference samples obtained by measuring fluorescent X-rays. The fluorescent X-ray spectrum is used as a reference waveform. The combined waveform of the corrected waveform and the background component obtained by correcting each of the reference waveforms using the shift amount of the spectral angle and the intensity ratio should match the measured waveform of the fluorescent X-ray spectrum obtained by measuring the sample. , And the correction value of the intensity ratio and the shift amount of the spectral angle with respect to each of the reference waveforms are calculated. The corrected waveforms are separated as peaks for each element.

[0011]

According to the present invention, since a plurality of types of fluorescent X-ray spectra obtained by measuring from the reference sample are used as the reference waveform, the reference waveform is corrected by taking into consideration the intensity ratio and the shift amount of the spectral angle. The waveform easily matches the measured waveform of the fluorescent X-ray spectrum having a complicated shape measured from the sample.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below. Figure 1
(a) shows a fluorescent X-ray spectrum obtained by measuring an actual sample with a measuring device (FIG. 3) and dividing it into spectral angle 2θ components. This spectrum includes a plurality of peaks p1 and p2 and a background component Ba, and the waveform of this spectrum is referred to as a measurement waveform F (x _i ). Note that x _i is the spectral angle 2θ of the _i- th data point. It is assumed that the sample contains the elements A1 and A2. Two reference waveforms G _{1 in} FIG. 1 (b)
(X _i ) and G ₂ (x _i ) are each a single element
The fluorescent X-ray spectra obtained by measuring fluorescent X-rays from two types of reference samples composed of pure substances A1 and A2 using a measuring instrument (FIG. 3) are shown.

Next, the peak separation method will be described. By irradiating the reference sample with X-rays in advance and measuring fluorescent X-rays from the reference sample, the reference waveform G ₁ (x
_i ), G ₂ (x _i ) is obtained. Then, for the actual sample, the fluorescent X-ray is measured by using the measuring instrument used for the measurement of the reference sample, and the measurement waveform F (x _i ) of FIG. 1 (a) is obtained. This measurement waveform F (x _i ) can be approximated by the following equation (1).

[0014]

[Equation 1]

The above-mentioned shift amount β _j is shown in FIG. 1 (a), (b), (c)
As shown in (1), it corresponds to the shift amount of the spectral angle of the peak p1 with respect to the reference waveform G _j (x _i ), and is caused by the chemical shift that occurs when the elements A1 and A2 contained in the actual sample are compounds and the device. Correct the angle deviation. That is,
The waveform G _j (x _i + β _j ) in FIG. 1C is the reference waveform G
A waveform in which _j (x _i ) is corrected by the shift amount β _j is shown. Figure 1
The intensity ratio α _j of (d) is the intensity ratio of the peak with respect to the reference waveform G _j (x _i ) as shown in FIGS. 1 (b) and 1 (d). In other words, the waveform _{α j G j (x i +} β j) in FIG. 1 (d) the waveform modified by the shift amount beta _j G _j the (x _i + β _j), was modified by addition intensity ratio alpha _j waveform Show. The waveform B (x _i ) of the background component Ba in FIG. 1 (e) is a function, for example, a quadratic polynomial shown in the following equation (2) is used. B (x _i ) = ax _i ² + bx _i + c (2)

Using the data of the waveform F (x _i ) and the reference waveform G _j (x _i ) of the actual sample, the well-known least square method is used to obtain α _j , β _j , of the above equations (1) and (2), Calculate a, b, and c. That is, the reference waveform G _j (x _i ) is corrected using the intensity ratio α ₁ and the shift amount β _1, and each corrected waveform α _j G _j (x _i + β _j ) in FIG. The synthesized waveform obtained by synthesizing the waveform B (x _i ) is the measured waveform F.
Intensity ratio α _j and shift amount β _j so as to match (x _i ).
And the background function B (x _i ) are calculated. From this calculation, each modified waveform α _j G _j (x _i +
β _j ) is obtained, and this is separated as a peak for each element An as shown in Fig. 1 (e).

Here, the peak (waveform) of the X-ray fluorescence spectrum has a similar shape without being significantly affected by the state of the sample and the state of chemical bond, provided that the optical conditions of the apparatus are constant. Therefore, the composite waveform of the corrected waveform α _j · G _j (x _i + β _j ) obtained by correcting the reference waveform G _j (x _i ) as described above matches the measured waveform F (x _i ) having a complicated shape. To do.
Therefore, the peaks can be separated easily and accurately.

Next, a concrete example of actual analysis will be shown. FIG. 2 (a) shows a fluorescent X-ray spectrum obtained by analyzing a low alloy steel containing chromium Cr, vanadium V, and titanium Ti. Figure 2 (b)
Shows a waveform obtained by peak separation by the above separation method. As the reference sample for measuring the reference waveform, pure chromium, V ₂ O ₅
Cr powder and pure titanium, respectively,
VKα ray, VKβ ray and TiKα ray were measured in advance.

In the measured waveform as shown in FIG. 2 (a), one peak p11 may include not only the CrKα line but also a slight VKβ ₁ line as shown in FIG. 2 (b). . In the case where the peaks overlap in the same spectral angle region, when the intensity of _one spectrum line (VKβ ₁ line) is significantly smaller than that of the other spectrum line (CrKα line),
With the conventional method using a function such as a Gaussian function, it is difficult to separate a spectrum line (VKβ ₁ line) having a small intensity. On the other hand, in this peak separation method, since the reference waveform for each element is used, the peaks p12 and P13 in FIG.
Therefore, by determining the VKα line in Fig. 2 (b), VKβ ₁
The line shape is also determined automatically. In other words, instead of grasping the VKβ ₁ line as a single peak, the VKβ ₁ line and VKα line
Since the line is grasped as an integral peak p12 (Fig. 2 (a)), VK
The peak intensity of VKβ ₁ ray is determined from the peak intensity of α ray. Therefore, even if the peaks overlap, the peaks can be easily separated.

For elements having a large influence (chemical shift) due to chemical bonding, it is preferable to use a reference sample made of a compound almost the same as the measurement sample. Further, in the above embodiment, the background component is quadratic function B (x
_{Although i} ) is calculated at the same time, it may be calculated after removing only the background component in advance.

[0021]

As described above, according to the present invention, since the reference waveform having a complicated shape obtained by measuring from the reference sample is used, the corrected waveform obtained by correcting the reference waveform is measured from the actual sample. Since it is easy to match the measurement waveform with a complicated shape, peak separation can be performed even in a wavelength dispersive X-ray fluorescence analyzer having good resolution. Further, since the peaks are separated for each element, the peaks can be easily separated even if two or more spectral lines overlap in the same wavelength region.

[Brief description of drawings]

FIG. 1 is a schematic diagram of a peak separation method showing an embodiment of the present invention.

FIG. 2 is a characteristic diagram showing a specific example of a waveform separated from a fluorescent X-ray spectrum.

FIG. 3 is a schematic configuration diagram showing an example of a measuring device used for fluorescent X-ray analysis.

FIG. 4 is a characteristic diagram showing an example of peak separation.

[Explanation of Codes] 51 ... Sample, B3 ... Fluorescent X-ray, Ba ... Background component,
F (x _i ) ... measurement waveform, G _j (x _i ) ... reference waveform, α _j
... intensity ratio, β _j ... wavelength shift amount, α _j G _j (x _i + β _j )
… Corrected waveform.

Claims (1)

[Claims] 1. A method for peak separation of a fluorescent X-ray spectrum for separating a fluorescent X-ray spectrum obtained by measuring a sample into peaks for each element, which comprises a plurality of types of standards consisting of a single element or compound. A plurality of types of fluorescent X-ray spectra obtained by measuring fluorescent X-rays from each sample are used as reference waveforms, and each of these reference waveforms is corrected using the shift amount and intensity ratio of the spectral angle and the background. The composite waveform with the components is calculated by calculating the deviation amount of the spectral angle and the correction value of the intensity ratio with respect to each of the reference waveforms so as to match the measured waveform of the fluorescent X-ray spectrum obtained by measuring the sample, A method for separating peaks of a fluorescent X-ray spectrum, in which each of the corrected waveforms is separated as a peak for each element.