JPH05240809A - Method and device for fluorescent x-ray analysis - Google Patents

Abstract

PURPOSE:To analyze a sample with a thin film containing boron by allowing fluorescent X rays to go into a spectral element, reducing a measurement error due to Si-Lalpha line which is totally reflected by the spectral element, and then measuring X-ray intensity of B-Kalpha ray which is diffracted by the spectral element. CONSTITUTION:A sample 2 is BPSG film whose thickness is equal to or less than 3000Angstrom and a fourth measuring equipment 60 which measures intensity of Si-Lalpha line B6 regarding Si from the sample 2 and a compensation means 20 are provided. The measuring equipment 60 is provided with a spectral element 64, an X-ray detector 66, a wave-height analyzer 67, and a counter circuit 68. The analyzer 67 receives an output e from the detector 66 and then outputs a pulse signal p where noise is eliminated to the counter circuit 68. The circuit 68 counts the pulse signal p, measures X-ray intensity of Si-Lalpha line B6, and then outputs an intensity signal x4, while a first measuring equipment 30 outputs an intensity signal x1 of B-Kalpha line including a component of Si-Lalpha line to a compensation means 20. The compensation means 20 performs compensation operation by subtracting the component of Si-Lalpha line and then obtains X-ray intensity of B-KY ray.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention mainly relates to a fluorescent X-ray analysis method and apparatus for analyzing a sample having a thin film on a silicon substrate.

[0002]

2. Description of the Related Art In the process of manufacturing a semiconductor, a layer of a BPSG film containing B (boron), P (phosphorus) and Si (silicon) as main components may be formed on the surface of a wafer made of a silicon substrate. .. There is a demand to make this BPSG film as thin as possible in order to miniaturize the semiconductor.

On the other hand, it is necessary to measure the composition and film thickness of such a thin film from the viewpoint of semiconductor quality control. A fluorescent X-ray analyzer has been conventionally used for this measurement. 4,000Å ~ 9,000Å on the above silicon substrate
FIG. 10 shows an example of a sample analyzer having the BPSG film of FIG.

In FIG. 10, a sample 2 is a silicon substrate 2.
It has a BPSG film 2b on the surface of a. The primary X-ray B1 emitted from the X-ray tube 1 excites the atoms of the sample 2 to generate the sample 2
The fluorescent X-ray B2 peculiar to the element constituting the is generated. The generated fluorescent X-ray B2 is measured by the first, second and third measuring instruments 30, 40 and 50 as follows.

The fluorescent X-ray B2 is emitted from the sample 2,
Fluorescent light X having a predetermined wavelength that passes through the first solar slit or slit 3 and is incident on the B, P, and Si spectroscopic elements 4B, 4P, and 4S, each of which has a predetermined wavelength that satisfies the Bragg equation.
Lines (B-Kα, P-Kα and Si-Kα lines) B3, B
4, B5 is diffracted. This diffracted fluorescent X-ray B3
B5 to B5 pass through the second solar slit or the slit 5 and enter the X-ray detectors 6B, 6P and 6S, respectively. The X-ray detectors 6B, 6P, 6S detect the fluorescent X-rays B3, B4, B5, respectively, and output the detection output e to the pulse height analyzers 7B, 7P, 7S via the amplifier 9.
Each detection output e is subjected to wave height analysis by wave height analyzers 7B, 7P and 7S to remove noise and the like, and then input to counting circuits 8B, 8P and 8S as a pulse signal p. Each of the counting circuits 8B ... 8S counts the pulse signal p and outputs X-ray intensity signals x1 ... X3 to the calculator 10. Calculator 10
Is to analyze the composition and film thickness of the sample 2 from each X-ray intensity by the known fundamental parameter method.

[0006]

As described above, the analysis of the sample 2 having a film thickness of 3,000Å or less, particularly about 1,500Å to 3,000Å is requested from the demand for thinning the BPSG film 2b. The need arises. Therefore, the inventor conducted the analysis using the above apparatus. In this case, since the film thickness is reduced, the X-ray intensity of B (boron) should be reduced, but conversely, the measured intensity of B-Kα ray is increased. It turns out that no analysis can be done.

The present invention has been made in view of the above-mentioned conventional problems, and one of the objects thereof is to provide a sample having a thin film containing 3,000Å or less and containing boron on a silicon substrate. It is to provide a fluorescent X-ray analysis method capable of performing analysis.

By the way, the present inventor has conducted repeated studies on the phenomenon that the measured intensity of the B-Kα ray increases conversely as the film thickness decreases. As a result, it was discovered that a part of the Si-Lα ray was totally reflected by the spectroscopic element 4B, and this appeared as noise, and the present invention was completed.

The cause of the above phenomenon will be described below. The Si-Lα ray is generated by knocking out an electron in the L shell, and is greatly affected not only by the composition (the component ratio of elements) but also by the molecular structure. Therefore, the Si-Lα ray is different from the oxide of Si in the thin film 2b. Almost does not occur, but occurs from the substrate 2a. On the other hand, since the energy of Si-Lα rays is extremely small, when the thin film 2b has a thickness of 4,000 Å or more, most of the Si-Lα rays are absorbed by the thin film 2b and are not emitted from the sample 2. However, when the thickness of the thin film 2b is 3,000 Å or less, the absorption of Si-Lα ray in the thin film 2b becomes small, which is unexpectedly expected by ordinary fluorescent X-ray analysis, and therefore the intensity of Si-Lα ray becomes large. Become.

On the other hand, the spectroscopic element produces diffraction only when X-rays are incident at a constant angle determined by the lattice plane gap of the element and the wavelength, and utilizes the fact that the X-rays are reflected efficiently, and thus a predetermined range is obtained. X-rays having a wavelength are separated. However, BK
Spectroscopic element 4 used when detecting soft X-rays such as α rays
Since B is composed of an artificial cumulative film formed by laminating two or more kinds of substances by vacuum vapor deposition or the like, the surface thereof has a mirror surface state. Therefore, the Si-Lα ray having small energy is totally reflected on the surface of the spectroscopic element 4B and is incident on the X-ray detector 6B as a part of the fluorescent X-ray B3. That is, B-Kα
At the spectral angle for diffracting the rays, part of the Si-Lα rays is totally reflected, and the B-Kα rays and Si-Lα rays are incident on the X-ray detector 6B.

Here, as described above, the measurement target (B-Kα
Even if the fluorescent X-rays B3 other than the X-rays are incident on the X-ray detector 6B, they are usually selected by performing the pulse height analysis in the pulse height analyzer 7B as follows. That is, the wave height analyzer 7B
11A includes a proportional amplifier 70, an upper limit selector 71, a lower limit selector 72, and an inverse simultaneous circuit 73 as shown in FIG. 11A, and a signal including electrical noise n and the like as shown in FIG. 11B. As shown in FIG. 11C, only signals having a predetermined crest value between the upper limit value and the lower limit value are selected from among the above, and this is output to the counting circuit 8B in FIG.

However, since the peak values of the B-Kα line and the Si-Lα line are close to each other and there is a variation in these peak values, only the B-Kα line is selected in the peak height analyzer 7B. Therefore, the counting circuit 8B is
The signal of the i-Lα line is also counted. As a result, when the film thickness is extremely thin, the apparent measured intensity of the B-Kα ray increases.

By the way, as a method of reducing the degree to which the totally reflected fluorescent X-rays are incident on the X-ray detector, there is known a method of forming a layer of a substance having a density smaller than that of the spectroscopic element on the surface of the spectroscopic element. (For example, Japanese Patent Laid-Open No. 61-8
See Japanese Patent No. 9547). However, this prior art shifts the phase of the totally reflected fluorescent X-rays with respect to the spectral angle, and therefore the intensity of the totally reflected fluorescent X-rays cannot be sufficiently reduced. As a result, even if the above-described prior art spectroscopic element is used, the analysis accuracy is not improved.

Therefore, another object of the present invention is to provide a fluorescent X-ray analysis method and apparatus for weakening the intensity of the fluorescent X-rays of the elements other than the measurement target which are totally reflected on the surface of the spectroscopic element to improve the analysis accuracy. That is.

[0015]

In order to achieve the above object, the method of the present invention according to claim 1 has a thin film having a film thickness of 3,000 Å or less and containing boron on a silicon substrate. The sample is irradiated with primary X-rays, the fluorescent X-rays from this sample are made incident on the spectroscopic element, and the measurement error due to the Si-Lα rays totally reflected by this spectroscopic element is reduced, so that the spectroscopic element diffracts the light. Then, the X-ray intensity of the B-Kα ray is obtained. According to the method of the present invention, the measurement error due to the Si-Lα ray totally reflected by the spectroscopic element can be reduced, so that the X-ray intensity of B-Kα ray with less noise can be obtained.

According to the second aspect of the invention, the sample containing the first and second elements is irradiated with radiation, and the fluorescent X-rays from the sample are made incident on the spectroscopic element to emit fluorescent X-rays relating to the first element. Fluorescence X emitted from the spectroscopic element after being diffracted
The X-ray intensity of the fluorescent X-ray corresponding to the fluorescent X-ray of the second element totally reflected by the spectroscopic element is measured while the X-ray is detected, and the total reflected fluorescence is obtained based on the X-ray intensity. The correction calculation is performed by subtracting the X-ray component, and the first
The intensity of the fluorescent X-ray due to the element is calculated. According to the method of the present invention, the correction operation is performed by subtracting the component of the totally reflected fluorescent X-ray related to the second element to obtain the intensity of the fluorescent X-ray due to the first element. The component can be sufficiently removed, and the strong X-ray intensity related to the first element can be measured without being weakened.

According to a third aspect of the present invention, in the analysis of the sample containing the first and second elements, the optical path from the spectroscopic element that diffracts the fluorescent X-rays related to the first element of the sample to the X-ray detector, An absorbing member is provided that absorbs the fluorescent X-rays of the second element that are totally reflected by the spectroscopic element and has a high transparency of the fluorescent X-rays of the first element. According to the device of the present invention, since the absorbing member that absorbs the fluorescent X-rays related to the second element is provided, the total reflected fluorescent X-rays of the total reflected internal fluorescent X-rays are different from those of the conventional device for shifting the phase of the total reflected fluorescent X-rays. The strength can be reduced.

[0018]

Embodiments of the present invention will be described below with reference to the drawings. In each of the following embodiments, the same or corresponding portions as those of the conventional example of FIG. 10 are denoted by the same reference numerals, detailed description and illustration thereof will be omitted, and different points will be mainly described.

FIG. 1 shows a first embodiment of the present invention. Figure 1
Sample 2 is a BPSG film 2 with a thickness of 3,000Å or less
b on the silicon substrate 2a. In this embodiment, the difference from the conventional example of FIG. 10 is that a fourth measuring device 60 for measuring the intensity of the Si-Lα ray B6 relating to Si (second element) from the sample 2 of FIG. 20 is provided.

The fourth measuring device 60 is the Si-
The spectroscopic element 64 that diffracts the Lα ray B6, and the spectroscopic element 64
An X-ray detector 66 for detecting the fluorescent X-rays B6 from, a wave height analyzer 67, and a counting circuit 68 are provided. The wave height analyzer 67 receives the detection output e from the X-ray detector 66 as an input, and counts the pulse signal p from which noise and the like are removed by the counting circuit 6.
Output to 8. The counting circuit 68 counts the pulse signal p to measure the X-ray intensity of the Si-Lα line (fluorescent X-ray corresponding to the Si-Lα line totally reflected by the spectroscopic element 4B) B6, and the intensity signal x4. Is output to the correction means 20. On the other hand, the first measuring device 30 receives the Si-Lα from the counting circuit 8B.
The intensity signal x1 of the B-Kα line including the total reflection component of the line is output to the correction means 20.

The correcting means 20 calculates B- based on the intensity signals x1 and x4 from the first and fourth measuring instruments 30 and 60.
A correction calculation is performed by subtracting the component of the Si-Lα ray totally reflected in the spectroscopic element 4B that diffracts the Kα ray (B3) to obtain the X-ray intensity of the true (low noise) B-Kα ray. Is. The correction means 20 is composed of the microcomputer 11 together with the calculator 10.

Next, with the above configuration, the B-Kα ray (B
A method of correcting the intensity of 3) will be described. Fourth
The measuring device 60 measures the X-ray intensity I _S at the spectral angle θ _S of the spectrum of the Si-Lα line shown by the broken line in FIG. 2 (however, the figure shows the spectrum when the silicon substrate 2a is used alone). On the other hand, the first measuring device 30 has a Si-
Spectral angle θ _B of B-Kα line including total reflection component of Lα line
X-ray intensity I _{BS at} is measured. This X-ray intensity I
_{Since BS} contains a component of total reflection of Si-Lα rays,
It is larger than the spectrum for the thick film (thickness of 4,000 Å or more) shown by the chain line.

Therefore, the film thickness and composition are known and thin (
3,000 Å or less) A standard sample having a BPSG film 2b is prepared, and the X-ray intensity I _S of the Si-Lα line and the spectral angle θ _B are prepared.
In advance, the relationship between the X-ray intensity I _S1 and the X-ray intensity I _S1 due to the total reflection of the Si-Lα line is obtained, and the X-ray intensity I _B of the true B-Kα line in the sample 2 having a thin BPSG film and the X-ray intensity I.
_Find the relationship with _BS and I _S. This relational expression is, for example,
It becomes like the following formula (1). I _B = I _BS − (A ₁ · I _S + A ₂ ) ... (1) I _B : X-ray intensity of true B-Kα line I _BS : B-Kα line containing total reflection component of Si-Lα line X-ray intensity at the spectral angle θ _B of I _S : X-ray intensity of the Si-Lα line at the spectral angle θ _S where A _n : a constant

The correction means 20 of FIG. 1 uses the equation (1) to calculate the intensity signals x4 and x4 from the counter circuits 68 and 8B.
Based on x1, the true X-ray intensity I _B of B-Kα ray is obtained. That is, the correction unit 20 determines the true X-ray intensity I _B of the B-Kα line by reducing the measurement error due to the Si-Lα line totally reflected by the spectroscopic element 4B of the first measuring device 30. True X-ray intensity I obtained by the correction means 20
_B is input to the calculator 10 as an intensity signal x5. The calculator 10 performs regression calculation by the known fundamental parameter method based on the intensity signals x5, x2, and x3 of the B-Kα line, the P-Kα line, and the Si-Kα line to determine the film thickness of the sample 2. Calculate the composition.

Thus, this fluorescent X-ray analysis method
By reducing the measurement error due to the total reflected Si-L [alpha line by the spectral element 4B for -Kα line, since obtaining the X-ray intensity of the B-K [alpha line, the true X-ray intensity I _B of B-K [alpha line
Therefore, even if the film thickness is less than 3,000Å, it is possible to analyze Sample 2.

Further, this fluorescent X-ray analysis method measures the X-ray intensity of fluorescent X-ray B6 corresponding to the Si-Lα ray (a part of fluorescent X-ray B3) totally reflected by the spectroscopic element 4B, and further, Since the correction means 20 subtracts the total reflection component of the Si-Lα line from the intensity of the fluorescent X-ray B3 measured by the first measuring device 30 for the B-Kα line, the conventional (the above-mentioned prior art: Unlike Kai 61-89547), the component of total reflection can be removed accurately and the X-ray intensity of the measured element is not attenuated. Therefore, not only the sample 2 but also another sample is analyzed, the effect of improving the analysis accuracy is obtained as compared with the above-described prior art.

By the way, the Si-Kα ray is used for the spectroscopic element 4B.
When incident on, the atoms of Mo, which is a component element of the spectroscopic element 4B, are excited to generate Mo-Mx rays. Therefore, the X-ray intensity of the Mo-Mx ray generated in the spectroscopic element 4B is calculated by the X-ray intensity of the Si-Kα ray measured by the third measuring instrument 50, and the X-ray measured by the first measuring instrument 30. If the intensity I _BS is corrected by the intensity of the Mo-Mx ray, the analysis accuracy is further improved.

In the above embodiment, the first, second, third
Although the fourth measuring device 30 ... 60 is composed of different measuring devices, it is possible to form one measuring device by using the scanning type. That is, if the multiple wave height analyzer in which the spectral element and the X-ray detector are rotated in conjunction with the goniometer and the range of the peak value analyzed by the wave height analyzer is variable is used, the first, second, third and Two to four of the fourth measuring devices can be configured with one measuring device.
Further, in this case, each of the dispersive crystals 64, 4B, 4
When P and 4S cannot be shared, the dispersive crystals 64, 4B,
A known structure in which only 4P and 4S are automatically replaced may be used.

Next, a second embodiment shown in FIG. 3 will be described. The embodiment of FIG. 3 is different from the conventional example of FIG. 10 in that a pair of wave height analyzers 37S and 37B and counting circuits 38S and 38B are provided in the first measuring device 30, and
The correction means 20A is provided.

The spectroscopic element 4B is the same as the conventional one, receives the fluorescent X-ray B2 from the sample 2 and diffracts the B-Kα ray (fluorescent X-ray relating to the first element B), and also Si-Lα.
A part of the line (fluorescent X-ray related to the second element Si) is totally reflected. These B-Kα rays and Si-Lα rays are detected by the X-ray detector 6B as fluorescent X-rays B3, and the detection output e thereof is input to the wave height analyzers 37S and 37B.

The above two wave height analyzers 37S and 37B
Respectively, as shown in FIG. 4, performs pulse height analysis in areas A1 and A2 of adjacent peak values, that is, the baselines BL1 to BL2 and BL2 to BL3, and outputs the pulse signal p of FIG. 3 to the counting circuits 38S and 38B. Output to. Each of the counting circuits 38S and 38B counts the pulse signal p, and the correction means 20A counts the intensity signal x of the X-ray intensity.
It outputs 7, x6.

Here, the principle of correction in the correction means 20A will be described. What is plotted in FIG. 4 is the X-ray intensity for each crest value of the B-Kα line that includes the total reflection component of the Si-Lα line. These X-ray intensities are shown by the dashed line B- near the baseline BL2.
This is a combination of the intensities of the Kα line and the Si-Lα line (total reflection component) indicated by the broken line, and therefore the intensities of the B-Kα line and the Si-Lα line are unknown. However, the X-ray intensities I _A1 , I obtained in the region A1 or A2
_A2 contains a large amount of X-ray intensities of Si-Lα rays and B-Kα rays, respectively, and is determined by the composition and film thickness of Sample 2. Therefore, regarding both X-ray intensities I _A1 and I _A2 A relational expression between the composition and the film thickness is obtained. That is,
X of two unknowns, Si-Lα line and B-Kα line
A simultaneous equation is obtained for the line intensities I _S1, I _B. Therefore, these simultaneous equations and P-Kα line and Si
The true X-ray intensity I _B of the B-Kα line is obtained by performing regression calculation using the fundamental parameter method based on the simultaneous equations for the −Kα line.

According to the above principle, that is, the correcting means 20A in FIG. 3 described above, that is, based on the outputs from both the wave height analyzers 37S and 37B, the S reflected by the spectroscopic element 4B for B-Kα rays is totally reflected.
The true BK is obtained by subtracting the component of the i-Lα line (the component related to the second element Si) from the X-ray intensity of the apparent B-Kα line.
The intensity of α-ray (fluorescent X-ray by the first element B) is obtained.

In this fluorescent X-ray analysis method, the fluorescent X-ray intensity I _A1 (hatched portion) corresponding to the totally reflected Si-Lα line in FIG. 4 is measured, and the apparent B is calculated based on this X-ray intensity. −
From the intensity of Kα rays, the total reflection component I of Si-Lα rays I
_{Since the} correction calculation is performed by subtracting _2S (the shaded portion) and the intensity of the (true) B-Kα ray with less noise is obtained, a sufficiently accurate analysis can be performed.

Also in this embodiment, the wave height analyzers 37S and 38B shown in FIG. 3 may be multiple wave height analyzers.
In addition, the peak value areas A1 and A2 in FIG.
Although divided by L1, BL2, and BL3, four baselines may be provided and the regions A1 and A2 may be slightly separated. Other configurations are similar to those of the first embodiment, and the description thereof will be omitted.

By the way, in the first and second embodiments, the X-ray intensity of the Si-Lα line was measured, but in the present invention, it is not always necessary to measure the intensity of the Si-Lα line. An example of the method is shown in the third embodiment of FIG.

The mechanical construction of the third embodiment shown in FIG. 5 is as shown in FIG.
Similar to the conventional example of 0, the only difference is that the arithmetic unit 10A is provided with the correction means 20B. In this third embodiment, the first, second and third measuring devices 30, 40, 5 of FIG.
0, Si-Kα ray, P-Kα ray, B-
The intensities I _SK (m), I _P (m) and I _Bs (m) of the Kα ray are measured. The correction means 20B of the arithmetic unit 10A is the BK
The measurement error due to the Si-L α ray is subtracted from the apparent measured intensity I _Bs (m) of the α ray. Hereafter, this computing unit 1
A method of calculating 0A will be described.

Firstly, calculator 10A is, Si0 _2, P ₂ 0
For _5, B ₂ 0 ₃ component ratio of each appropriate concentration W _S, W _P, giving a W _B with (assuming), given the thickness T Te (assuming), these concentrations, the thickness , Based on theoretical X-ray intensity formula, fluorescent X-ray Si-Kα ray, PK
Theoretical intensities of α rays, B-Kα rays and Si-Lα rays I
_SK (th), I _P (th), I _B (th) and I _SL (th) are calculated. Then, for example, according to the following equation (2), S
The measured intensity I _{BS of the} B-Kα ray containing the i-Lα ray
(m) is corrected based on the theoretical intensity I _SL (th) of the Si-Lα ray, and the corrected intensity I _B 1 of the B-Kα ray is obtained. I _B 1 = I _BS (m)-{A ₃ · I _SL (th) + A ₄ } (2) where A _n : a constant (the film thickness and the structure are obtained in advance using a known standard material) , The intensity I _B 1 and the measured intensity I _SK (m),
I _P (m) is the theoretical strength I _B (th), I
_It is determined whether or not they match _SK (th) and I _P (th), and new concentrations W _S , W _P , W _B and film thickness T are given based on the difference between the measured intensity and the theoretical intensity. Thus, by a well-known fundamental parameter method repeating the above calculation, to calculate the density W _S, W _P, W _B and thickness T.

According to the third embodiment, the same three measuring devices 30, 40 and 50 as those of the conventional one are used, and the arithmetic unit 10 is used.
There is an advantage that only the program of A or the like needs to be changed from the conventional one.

FIGS. 6, 7 and 8 show the fourth, fifth and sixth embodiments, respectively. In these examples, as described below, the spectroscopic element 4 for B-Kα rays is used.
In the optical path from B (FIG. 10) to the X-ray detector 6B, the Si-Lα line (fluorescent X-ray related to the second element Si) totally reflected by the spectroscopic element 4B is absorbed, but the B-Kα line is transmitted. This is different from the conventional example in that the absorbing members 81 ... 83 having good properties are provided.

In the embodiment shown in FIG. 6, an absorbing member 81 made of polypropylene, for example, is provided in front of the second solar slit 5 to absorb a part of the totally reflected Si-Lα rays, thereby causing a measurement error. Is decreasing.

In the embodiment of FIG. 7, an absorbing member 82 having carbon or boron deposited is provided on the outer surface of the mylar 62 of the detector window member 61 in the X-ray detector 6B for B-Kα rays, and the whole is provided. A part of the reflected Si-Lα ray is absorbed to reduce the measurement error. The detector window material 61
Has a mylar 62 made of polypropylene and a conductive Al thin layer 63 deposited inside the mylar.

In the embodiment of FIG. 8, the absorbing member 83 made of an element having a density lower than that of the element forming the spectroscopic element 4B itself is provided on the reflecting surface 4a of the spectroscopic element 4B for B-Kα rays. The absorbing member 83 is formed by vapor-depositing carbon having a thickness of 500 Å, for example, when the spectroscopic element 4B is composed of a Mo dispersive crystal.

In the fourth to sixth embodiments, Si-L is used.
Since the absorption members 81 ... 83 which absorb α rays but have excellent B-Kα ray transmission are provided, the conventional technique for shifting the phase of fluorescent X-rays totally reflected (the above-mentioned prior art: JP-A-618954).
No. 7), the X-ray intensity of Si-Lα rays can be made smaller. Therefore, these examples can be applied not only to the sample 2 (FIG. 1) but also to other samples. In addition,
In the fourth to sixth embodiments, the absorption of Si-Lα rays is determined by the X-ray intensity of Si-Lα rays with respect to the X-ray intensity of B-Kα rays.
The line strength is set to 10% or less.

FIG. 9 shows a seventh embodiment. In this embodiment, the surface 4a of the spectroscopic element 4B made of a dispersive crystal is formed into a rough surface having a roughness of about 500 Å by sputtering with argon ions. In this embodiment, the diffused reflection of the fluorescent X-ray B2 reduces the total reflection component of the Si-Lα ray, thereby reducing the measurement error.
The X-ray intensity of B-Kα ray is obtained. The seventh embodiment is included in the invention of claim 1, but not included in other inventions.

[0046]

As described above, according to the method of the first aspect of the present invention, a sample having a thin film of 3,000Å or less on a silicon substrate and having a thin film containing boron is irradiated with primary X-rays. Since the X-ray intensity of the B-Kα ray is obtained by reducing the measurement error due to the Si-Lα ray totally reflected by the spectroscopic element, the X-ray intensity of the B-Kα ray with less noise can be obtained. Even if the thin film is thin, fluorescent X-ray analysis can be performed.

According to the method of the second aspect of the present invention, the X-ray intensity of the fluorescent X-rays corresponding to the fluorescent X-rays of the second element totally reflected by the spectroscopic element is measured, and the X related to the second element is measured. Since the correction calculation is performed by subtracting the component of the fluorescent X-rays totally reflected on the basis of the line intensity to obtain the intensity of the fluorescent X-rays by the first element, the total reflection component is sufficiently removed. It is possible to have strong X related to the first element
Since the line strength can be measured without being weakened, the analysis accuracy is improved.

According to the third aspect of the invention, an absorption member is provided in the optical path from the spectroscopic element to the X-ray detector to absorb the fluorescent X-rays of the second element totally reflected by the spectroscopic element. Therefore, the intensity of the totally reflected fluorescent X-rays can be made smaller than that of the conventional apparatus that shifts the phase of the totally reflected fluorescent X-rays. Therefore, analysis accuracy is improved.

[Brief description of drawings]

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

FIG. 2 is a characteristic diagram showing X-ray intensity with respect to a spectral angle.

FIG. 3 is a schematic configuration diagram of a fluorescent X-ray analyzer showing a second embodiment of the present invention.

FIG. 4 is a characteristic diagram showing a relationship between a crest value analyzed by a crest analyzer and an X-ray intensity.

FIG. 5 is a schematic configuration diagram of a fluorescent X-ray analyzer showing a third embodiment of the present invention.

FIG. 6 is a schematic configuration diagram of a main part showing a fourth embodiment of the present invention.

FIG. 7 is a sectional view showing a window member of an X-ray detector showing a main part of a fifth embodiment of the present invention.

FIG. 8 is a side view of a spectroscopic element and an absorbing member showing a main part of a sixth embodiment of the present invention.

FIG. 9 is a side view of a spectroscopic element showing a main part of a seventh embodiment of the present invention.

FIG. 10 is a schematic configuration diagram showing an example of a conventional X-ray fluorescence analyzer.

FIG. 11A is a schematic configuration diagram of a wave height analyzer, and FIG.
(C) is a characteristic diagram showing the function of the wave height analyzer.

[Explanation of symbols]

1 ... X-ray tube (radiation source), 2 ... Sample, 2a ... Silicon substrate, 2b ... Thin film, 4B ... Spectroscopic element, 6B ... X-ray detector,
81, 82, 83 ... Absorption member, B1 ... Primary X-ray (radiation), B2 ... Fluorescent X-ray, B3 ... Fluorescent X-ray emitted from the spectroscopic element.

Claims (3)

[Claims] 1. A primary X sample is formed on a silicon substrate having a film thickness of 3,000Å or less and a thin film containing boron.
In the fluorescent X-ray analysis method for analyzing the sample, the sample is analyzed by irradiating the sample with the X-ray, making the fluorescent X-ray from the sample incident on the spectroscopic element, and measuring the intensity of the B-Kα ray diffracted by the spectroscopic element. A fluorescent X-ray analysis method, wherein the X-ray intensity of the B-Kα ray is obtained by reducing a measurement error due to the Si-Lα ray totally reflected by the spectroscopic element. 2. A sample containing the first and second elements is irradiated with radiation, and fluorescent X-rays from this sample are made incident on a spectroscopic element to diffract fluorescent X-rays relating to the first element,
A fluorescent X-ray analysis method for detecting a fluorescent X-ray emitted from the spectroscopic element with an X-ray detector to analyze a sample, which corresponds to the fluorescent X-ray of the second element totally reflected by the spectroscopic element. A correction calculation is performed by measuring the X-ray intensity of the fluorescent X-ray and subtracting the component of the totally reflected fluorescent X-ray based on the X-ray intensity to obtain the fluorescent X-ray of the first element. A fluorescent X-ray analysis method, characterized by obtaining intensity. 3. A radiation source that irradiates a sample containing the first and second elements with radiation, a spectroscopic element that diffracts fluorescent X-rays related to the first element from the sample, and a fluorescent X from the spectroscopic element. An X-ray detector for detecting X-rays, which comprises an X-ray fluorescence analyzer for analyzing the sample, wherein a second element totally reflected by the spectroscopic element is provided in an optical path from the spectroscopic element to the X-ray detector. An X-ray fluorescence analyzer, which is provided with an absorbing member that absorbs the fluorescent X-rays regarding the first element and that has a high transparency to the fluorescent X-rays regarding the first element.