JPH0750042B2 - State analysis method in spectrum analyzer - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Field of Industrial Application The present invention relates to a state analysis method in an apparatus capable of spectrum analysis such as EPMA and ESCA.

(B) Conventional technology and its problems Conventionally, for example, in the state analysis by EPMA, a spectrum of an unknown sample is measured to create a wavelength profile, and characteristic data of the profile, for example, a half-value width of a main peak, a peak The automation is falsified up to the stage of calculating and outputting the intensity ratio, each peak wavelength or the asymmetry index, and the analyst compares the output characteristic data with the characteristic data of the standard sample to determine the state of the unknown sample. Are doing. As described above, there is a problem in that it takes time for processing because the analyst determines the state.

The present invention has been falsified in view of the above points, and an object thereof is to make it possible to automatically make a determination in a state analysis.

(C) Means for Solving the Problems In the state analysis method in the spectrum analyzer of the present invention, in order to achieve the above object, in advance, for each element,
Create a database of characteristic data over multiple items by performing spectral analysis on multiple standard samples and importance data indicating the importance at the time of determination in the state analysis of each item for each element, and The characteristic data of the plurality of items obtained by performing the spectral analysis and the characteristic data for each standard sample of the corresponding element of the database are sequentially compared to obtain the difference,
The state of the unknown sample is determined based on the evaluation data obtained by adding the importance data to each difference.

(D) Action According to the above configuration, since a database of characteristic data over a plurality of items of a standard sample and importance data indicating the importance at the time of determining a plurality of items is created in advance, these data and unknown Based on the characteristic data of the sample,
It is possible to automatically determine the state analysis of an unknown sample.

(E) Example Hereinafter, an example of the present invention will be described in detail with reference to the drawings. This embodiment will be described by applying it to EPMA. First
FIG. 1 is a block diagram of EPMA used for carrying out the method of the present invention. In the figure, 1 is a CPU that performs data processing and predetermined control, and 2 is an RO in which a control program is stored.
M, 3 is a RAM as a work area, 4 is an interface, 5 is an analyzer for spectral analysis of a sample, and 6 is a database described later.

The state analysis method of this embodiment is performed according to the following procedure.

First, a plurality of standard samples (known samples) are prepared for each element in advance.
Is performed to prepare a database 6 of characteristic data over a plurality of items and importance data indicating the importance at the time of determination in the state analysis of each item for each element.

This spectral analysis is performed on a plurality of standard samples for each element. For example, for Fe, Fe, FeO, Fe ₂ O ₃ , Fe ₃ O ₄ , FeS, FeC, Fe ₂ N ... Are used as standard samples. In addition, at the time of spectrum analysis, the peak for measuring the wavelength profile, that is, the type of X-ray (K
α, Kβ, Lα, Lβ ...) is defined as one type for each element,
The choice of this peak is empirically determined to be the most appropriate. For example, in the case of Fe, the FeKα line is selected.

Wavelength profile data obtained by spectral analysis is the background correction, normalization, and preprocessing.
In addition, statistical fluctuations are corrected.

From the data of the wavelength profile which has been subjected to such pre-processing, characteristic data for the following plural items are calculated.

Half peak width of main peak When high-order peaks exist, peak intensity ratios Peak wavelengths (main peak, high-order peak, etc.) Asymmetry index A database 6 of characteristic data above the analysis line coefficient of the midpoint method is created. .

FIG. 2 is a flow chart of the above operation,
At n1, the element and the type of X-ray are determined and the process proceeds to step n2. In step n2, a standard sample is selected, and in step n3
Then, the wavelength profile of the standard sample is measured and the process proceeds to step n4. In step n4, preprocessing such as background correction and standardization is performed, and then the process proceeds to step n5. In step n6, the characteristic data items 1 to 3 above are calculated, and the process proceeds to step n5 to create the database 6. This series of operations is performed for each standard sample of each element.

The above-mentioned properties (1) to (5) have different degrees of importance in the determination of the state analysis depending on the element. This importance is obtained semi-empirically. In this embodiment, the database 6 of the importance coefficient showing the importance of each item for each element is created. Therefore, the database 6 stores the characteristic data and the importance coefficient. For example, regarding Fe, each peak wavelength is the most important at the time of judgment, then the peak intensity ratio, and then the asymmetry index, and the importance coefficient for each item is determined accordingly. It

The spectral analysis of an unknown sample is performed after the above preparations. The wavelength profile data obtained for the unknown sample is subjected to background correction, normalization, and statistical fluctuation correction as pretreatment, as in the case of the standard sample.

Next, the characteristic data of each of the above items 1 to 3 is calculated from the data of the wavelength profile of the unknown sample.

Characteristic data and database 6 for this unknown sample
Characteristic data of each standard sample of the corresponding element of is compared sequentially, the difference is calculated respectively, the difference is multiplied by the importance coefficient of the database 6, and the reciprocal thereof is used as the score, which is the most similar to the unknown sample. The standard sample, that is, the standard sample having the largest total score, is determined and output.

FIG. 3 is a flowchart of the analysis operation for the above unknown sample. In step n1, the element to be analyzed of the unknown sample and the type of X-ray are determined, and the process proceeds to step n2. In step n2, the wavelength profile is measured and then step
Move to n3. In step n3, background correction,
Perform preprocessing such as standardization and move to step n4.
Characteristic data is calculated and the process proceeds to step n5. Step
At n5, the difference between the characteristic data of each standard sample of the corresponding element in the database 6 and the characteristic data obtained at step n4 is calculated, and the process proceeds to step n6. In step n6,
The importance coefficient for the corresponding element in the database 6 is multiplied by the difference obtained in step n5 and the reciprocal thereof is calculated to calculate the score, and the process proceeds to step n7. In step n7, the standard sample having the largest total score is searched for, and the process proceeds to step n8, where the standard sample is output.

In this way, the determination of the state analysis will be automatically performed, the processing can be performed faster than the conventional example performed by the analyst, and furthermore, the state is determined based on the objectively calculated score, The objectivity of the judgment of the state analysis is improved.

Note that the importance coefficient is obtained semi-empirically, so the output result may not be valid depending on the unknown sample.In that case, change the setting of the importance coefficient value. Alternatively, the accuracy of analysis may be improved by adding characteristic data of standard samples.

(F) Effect As described above, according to the present invention, spectral analysis is performed in advance on a standard sample to create a database of characteristic data over a plurality of items and importance data indicating the importance of each item, and unknown samples are prepared. Based on the evaluation data obtained by comparing the characteristic data obtained by conducting the spectrum analysis with the characteristic data of each standard sample in the database to find the difference, and adding the importance data to each difference. Since the state of the unknown sample is determined, the state analysis can be automatically determined.

[Brief description of drawings]

FIG. 1 is a block diagram of a spectrum analyzer used for implementing the present invention, and FIGS. 2 and 3 are flow charts for explaining the operation. 1 ... CPU, 5 ... Analysis device, 6 ... Database.

Claims (1)

[Claims] 1. Preliminarily performing spectral analysis on a plurality of standard data for each element to show characteristic data over a plurality of items and importance indicating the importance at the time of judgment in the state analysis of each item for each element. Data database, the characteristic data of the multiple items obtained by conducting the spectrum analysis of the unknown sample and the characteristic data of each standard sample of the corresponding element of the database are sequentially compared and the difference is found. A state analysis method in a spectrum analyzer, wherein each state is determined, and the state of an unknown sample is determined based on evaluation data obtained by adding the importance data to each difference.