JP7134427B2 - Crystal lattice plane distribution measurement method - Google Patents

Description

The present invention relates to a crystal lattice plane distribution measuring method, and more particularly to a crystal lattice plane distribution measuring method for measuring a crystal lattice plane distribution with a high spatial resolution of 50 μm or less.

The crystallinity of semiconductor crystals, such as lattice strain and defects, affects the characteristics of semiconductor devices that are products. For example, it has been reported that lattice strain changes the bandgap of GaN and affects device characteristics (see Non-Patent Documents 1 and 2). In the field of semiconductor device manufacturing, as devices are miniaturized, it is industrially very important to observe dislocations and defects and feed back the results to crystal manufacturing to improve the quality of crystals.

As a method for observing crystal perfection, there is a method collectively called an X-ray diffraction microscopic method or an X-ray diffraction topography (XRDT) method (see Non-Patent Document 3). This method is further classified into the Berg-Barrett method, the Lang method, the Laue method, and others. These methods are commonly used by laboratory X-ray sources even before the use of synchrotron radiation (synchrotron X-rays) as an X-ray source. These methods can record the contrast of diffracted X-ray intensity due to dislocations and defects with a resolution of several μm, and determine the direction of the Burgers vector of dislocations.
An attempt has also been made to measure the in-plane distribution of crystal orientation using an X-ray source in a laboratory (see Patent Document 1).

With the development of the XRDT method using synchrotron X-rays, the distribution of lattice spacings and the curvature of lattice planes have been evaluated since 1980 (see Non-Patent Document 4). By combining the XRDT method with the laser scanning method, it was possible to understand curvature information including residual stress (see Non-Patent Document 5). For example, a recent report on lattice plane bending of 4H--SiC4 substrates observed by X-ray diffraction indicates the shape and curvature of the substrate (see Non-Patent Document 6).

JP-A-2-266249

J. Appl. Phys. , vol. 74, p. 6734 (1993) Phys. Rev. B: Condens. Matter, vol. 54, p. 17745 (1996) Journal of the Crystallographic Society of Japan, vol. 13, p. 358 (1971) J. Appl. Phys. , vol. 51, p. 6224 (1980) Thin Solid Films, vol. 415, p. p. 21-31 (2002) Cryst, Eng. Comm. , vol. 19, p. 3844 (2017)

The conventional crystal lattice plane orientation distribution measurement method described in the background is mainly aimed at obtaining macroscopic properties in a sense such as the curvature of the sample from the viewpoint of the crystal lattice plane, and the spatial resolution is high. It wasn't. Therefore, it is not suitable for measuring the variation distribution of the crystal lattice plane orientation of a sample with high spatial resolution.
SUMMARY OF THE INVENTION Under such circumstances, an object of the present invention is to provide a high-speed method for measuring the crystal lattice plane orientation distribution of a sample having a crystal structure with a high spatial resolution of 50 μm.

The configuration of the present invention is shown below.
(Configuration 1)
In a crystal lattice plane orientation distribution measuring method for measuring the crystal lattice plane orientation distribution of a sample having a crystal structure,
an X-ray irradiation step of irradiating each in-plane portion of the sample with a monochromatic parallel X-ray with a plurality of different incident angles for each of the two or more azimuths along the direction of the crystal axis of the sample. When,
Bragg reflection X for detecting the Bragg reflected light of the X-rays from the sample corresponding to each in-plane portion and measuring the Bragg reflected X-ray intensity in each of the two or more azimuths corresponding to each in-plane portion a line strength measurement step;
an incident angle calculating step of obtaining an incident angle that maximizes the reflection intensity of the Bragg-reflected X-ray for each of the two or more azimuths and each portion in the plane;
a crystal lattice plane orientation calculation step of calculating the crystal lattice plane orientation for each in-plane portion from the two or more orientations and the incident angle obtained for each in-plane portion.
(Configuration 2)
The crystal lattice plane orientation distribution measuring method according to configuration 1, wherein the number of orientations is two.
(Composition 3)
The crystal lattice plane direction of each in-plane portion is projected against the average direction of the crystal lattice plane direction distribution for each in-plane portion, and the difference in the crystal lattice plane direction of each in-plane portion with respect to the average direction of the crystal lattice plane direction distribution is calculated. 3. The crystal lattice plane orientation distribution measuring method according to configuration 1 or 2, wherein the in-plane distribution of the difference is obtained and vector-displayed.
(Composition 4)
4. The crystal lattice plane orientation distribution measuring method according to any one of configurations 1 to 3, wherein the X-ray is monochromatic parallel radiation with a wavelength of 0.01 nm or more and 0.25 nm or less.

According to the present invention, it is possible to provide a method for measuring the crystal lattice plane direction distribution of a sample having a crystal structure at a resolution of 50 μm or less at high speed.

An explanatory view showing a concept of the present invention in a schematic diagram. FIG. 4 is an explanatory diagram showing the relationship between incident X-rays and Bragg-reflected X-rays; FIG. 3 is a sample in-plane distribution diagram of Bragg reflected light intensity when the incident angle θ _k of X-rays is used as a parameter. In-plane distribution diagram of _x -component qx and _y -component qy of the crystal orientation plane normal. Fig. 2 is a sample in-plane distribution diagram of a projection image of a crystal lattice plane normal projected onto an ideal (00.1) plane. The inclination distribution diagram of the normal to the crystal lattice plane. The cross-sectional schematic diagram of a crystal plane shape. Explanatory drawing of the coordinate system of an Example.

A mode for carrying out the present invention will be described below with reference to the drawings.
<Outline of the measuring method of the present invention>
A sample having a crystalline structure is irradiated with X-rays from an orientation (first orientation) along the crystal axis while changing the incident angles to the sample. Then, the intensity of the Bragg reflected light from the sample is measured by a two-dimensional X-ray detector such as a flat panel detector, CMOS array sensor, CCD array sensor, etc. to obtain a so-called rocking curve. The measurement accuracy can be improved by setting the first orientation along the crystal axis.
Here, the X-rays used for the measurement are preferably monochromatic parallel X-rays in order to improve the measurement accuracy. In particular, synchrotron radiation (synchrotron X-rays), which can increase the intensity of X-rays, is more preferable. The wavelength of the X-ray is preferably 0.01 nm or more and 0.25 nm or less, and the higher the monochromaticity (the narrower the wavelength distribution), the higher the measurement accuracy.
The required monochromaticity and parallelism of the incident X-rays vary depending on the crystal perfection of the sample and the wavelength of the X-rays used. Here, the monochromatic X-rays are X-rays in which Δλ/λ _ave falls within the range of 10 ⁻⁴ when the spread Δλ of the wavelength λ of the X-rays is λ _ave as the center wavelength of the X-rays, Also, the parallel X-rays here refer to X-rays whose divergence angle is within the range of 30 arcsec.

A rocking curve for hkl diffraction from each site on the sample surface is then recorded to obtain its peak angular position. Then, the deviation angle Δθ ₁ from the average angular peak position obtained from the entire sample surface is obtained for all sites. That is, the above measurement is performed at each location in the plane of the sample to obtain data of Δθ(x, y).

Assuming that the sample is a single crystal and the sample surface is approximately the (00.1) plane, .DELTA..theta.(x,y) is the ideal (00.1) plane 11 as shown in FIG. It is the same as the angle formed by the normal 13 and the local normal direction 10 at each portion of the crystal lattice plane substantially parallel to the sample surface.
Here, x and y represent locations within the sample surface, and 11 in FIG. 1(a) is the ideal (00.1) plane. Reference numeral 12 denotes a plane formed by normals obtained by averaging the normals of crystal lattice planes from various locations over the entire sample surface and incident X-rays that cause diffraction, and is referred to as an average (00.1) diffraction plane A here. 14 is an image (map) obtained by projecting the normal to the crystal lattice plane at each location (x, y) onto the average (00.1) diffraction plane A. FIG. Note that the notation of (00·1) conforms to the notation method for hexagonal GaN used in the examples.

Rotate the sample φ around its surface normal (change the azimuth angle φ) and repeat the same measurement, and as shown in FIG. evaluate.
Here, 22 in FIG. 1(b) indicates that the normal of the crystal plane from each location is averaged over the entire sample surface and the incident X-ray that causes diffraction in the measurement at the second azimuth angle. The plane to be made, ie the average (00.1) diffraction plane B. 24 is an image (map) obtained by projecting the normal to the crystal lattice plane at each location (x, y) onto the average (00.1) diffraction plane B. FIG.

Next, using the obtained Δθ(x, y) and Δχ(x, y)), the normal q(q _x (x, y) , q _y (x, y), q _z (x, y)) are obtained from (Equation 1). It is preferable to set the second azimuth angle along the crystal axis of the sample in order to improve the measurement accuracy.

where R is a 3-by-3 matrix of coordinate transformations (equation 2).

As a result, the shape of the crystal lattice plane can be obtained over the entire surface of the sample.

In the above description, the case where there are two azimuth angles, the first and the second, is described. good too. Here, the number of orientations to be measured must be at least two.
Increasing the number of azimuths to be measured has the effect of improving the measurement accuracy. On the other hand, when the number of azimuths to be measured is two, the measurement time and the time required for calculation are shortened, and the measurement efficiency is improved.

Then, from the first, second or more orientations and the deviation angles Δθ (x, y) and Δχ (x, y) obtained for each portion in the plane of the sample, the crystal lattice plane direction for each portion in the plane of the sample is calculated. do.
Specifically, the crystal lattice plane direction of each portion of the sample plane is projected onto the average direction of the crystal lattice plane direction distribution of each portion of the sample plane, and the crystal lattice of each portion of the sample plane relative to the average direction of the crystal lattice plane direction distribution is calculated. A difference in the plane direction is obtained, and the distribution of the difference in the plane of the sample is displayed as a vector.

<Spatial resolution of the present invention>
The spatial resolution of each part is determined by the pixel size Δp of the two-dimensional X-ray detector used.
The present invention can evaluate within the range of X-ray extinction distance from the sample surface. The extinction distance depends on the wavelength of incident X-rays, the sample, and the diffraction index used, but is typically several micrometers.
It can be applied when the lattice constant difference δd between adjacent Δp regions on the sample surface satisfies |δd/d|<Δp/(2L·tan(θ _B )).
Here, L is the camera length defined by the distance between the sample and the X-ray detector. d is the interplanar spacing involved in hkl diffraction, and θ _B is the Bragg angle related by the wavelength λ of incident X-rays and d under the Bragg condition λ=2d·sin(θ _B ).
The deviation angle Δψ from the normal q of the crystal lattice planes averaged over the entire sample surface at each site on the sample can be observed for |Δψ|<Δp/L.

As will be described later, in the embodiment of the present invention, an X-ray detector with a pixel size of 50 μm×50 μm was used. Therefore, the spatial resolution in this case is 50 μm. However, this is just an example, and it is also possible to reduce the pixel size to, for example, 20 μm×20 μm.
The spatial resolution measurable by the technique of the present invention precisely depends on the pixel size per pixel and the pitch between adjacent pixels (inter-pixel pitch). Here, if the deviation angle from the normal line averaged over the entire sample surface is large and the X-ray is incident on the adjacent pixel beyond the pixel pitch, the location on the sample corresponding to the adjacent pixel is measured and distinguished. This limits the measurable deviation angle.

On the other hand, the smaller the pixel size, the smaller the measurable off-normal angle averaged over the sample surface. To solve this problem, the camera length L should be shortened.
Therefore, the minimum |Δψ| is substantially defined by the pixel size and the camera length L that can be arranged. The camera length L in Examples described later was 0.5 m. The technology of the present invention can control observable |Δψ| by optimizing the pixel size of the X-ray detector and the arrangement of the camera.

The present invention will now be described in more detail with reference to examples, but it should be noted that the present invention is not limited to the following examples, but is defined only by the claims. want to be

(Example 1)
In Example 1, as a sample 101 (see FIG. 2 described later), GaN was grown to a thickness of 1 μm on a gallium (00.1) nitrogen semiconductor wafer having a diameter of 2 inches, and magnesium-added GaN was grown to a thickness of 2 μm. used things. Its thickness is 330 μm.
Then, using the synchrotron radiation of the large synchrotron radiation facility SPring-8 as the radiation source, the entire surface of the sample is irradiated with almost monochromatic parallel X-rays from a silicon monochromator using two (111) plane crystals, and (hk l) The incident angle θ of X-rays was adjusted so that diffraction occurred from the crystal lattice plane. The measurement was also performed at SPring-8.

Figure 2 shows the relationship between incident X-rays and Bragg-reflected X-rays.
The incident X-ray wavelength was 0.1284 nm and excited 11·4 diffraction. The angle between the incident X-ray and the surface of the sample 101 was 0.6°. Since the ( _00.1 ) plane and the (11.4) plane form an angle of 39.1°, the incident angle is changed to A diffraction image was measured by the two-dimensional X-ray detector 102 every time the step was changed by 25 arcsec. FIG. 3 shows the in-plane distribution of the Bragg reflected light detected at that time, using the X-ray incident angle θ _k as a parameter.
The X-ray exposure time was 0.5 seconds. The X-ray detector 102 used was a flat panel detector (manufactured by Hamamatsu Photonics) with a pixel size of 50 μm×50 μm and 2368×2240 pixels. The x direction is parallel to [11.0] and the y direction is perpendicular to it. [11.0] is in the average diffraction plane.
The sample was then rotated φ=120° around the surface normal of the sample (rotating the azimuth angle by 120°) and similar measurements were repeated. The total measurement time was 32 minutes and 30 seconds.
|δd/d| in Example 1 is less than 6.0×10 ⁻⁵ and |Δψ| is less than 0.06°.

Next, using (Formula 3), normal lines q(q _x (x, y), q _y (x, y), q _z ( x, y)) were determined. The measured q _x (x, y) and q _y (x, y) are shown in FIGS. 4(a) and 4(b), respectively.

The projection of the normal q onto the ideal (00.1) plane is indicated by an arrow in FIG.
Moreover, the inclination distribution of the normal q was analyzed. The results are shown in FIG. Further, the shape of the crystal plane was determined, and the schematic shape thereof is shown in FIG. 7(a) for the x direction and in FIG. 7(b) for the y direction. The coordinate system used this time is shown in FIG.
As a result, it was confirmed that the in-plane distribution of the crystal orientation plane of the sample can be measured with a high spatial resolution of 50 μm in a short measurement time of the order of 30 minutes.

As described above, the present invention is a method for measuring the in-plane distribution of the crystal orientation plane of a sample with a high spatial resolution of 50 μm in a short measurement time of the order of 30 minutes. Conventional methods for measuring crystal orientation planes have focused on grasping macroscopic characteristics such as warpage and curvature of the entire sample. Distribution can be analyzed.
If this measurement method is applied to semiconductor substrates, semiconductor devices, etc., it will be possible to correlate the effects of crystal orientation plane variations on them, and high-quality semiconductor substrates, semiconductor devices, etc. with less in-plane crystal orientation variations can be obtained. be able to provide. For this reason, there is a possibility that it will be widely used in industrial fields including semiconductors.

10: Normal direction vector of the crystal orientation plane of the sample 11: Ideal (00.1) plane 12: Average (00.1) diffraction plane A
13: Normal to the ideal (00.1) diffraction plane 14: Projected image (mapping) onto the average (00.1) diffraction plane A
22: Average (00.1) diffraction plane B
24: Projected image (mapping) onto the average (00.1) diffraction plane B
101: Sample 102: X-ray detector (flat panel detector)

Claims (3)

In a crystal lattice plane orientation distribution measuring method for measuring the crystal lattice plane orientation distribution of a sample having a crystal structure,
Directional X-rays having a parallel speed of light are irradiated to each part in the plane of the sample with a plurality of different incident angles for each of the two or more azimuths along the direction of the crystal axis of the sample. an X-ray irradiation step;
The Bragg reflected light intensity of the X-rays from the sample is detected corresponding to each part in the plane, and the Bragg reflected light intensity is measured corresponding to each part in the plane for each of the two or more azimuths. a measuring step;
an incident angle calculating step of obtaining an incident angle that maximizes the reflection intensity of the Bragg reflected light for each of the two or more azimuths and each portion in the plane;
a crystal lattice plane direction calculation step of calculating the crystal lattice plane direction for each in-plane portion from the two or more orientations and the incident angle obtained for each in-plane portion ;
The crystal lattice plane direction of each in-plane portion is projected against the average direction of the crystal lattice plane direction distribution for each in-plane portion, and the difference in the crystal lattice plane direction of each in-plane portion with respect to the average direction of the crystal lattice plane direction distribution is calculated. A crystal lattice plane direction distribution measuring method, wherein the distribution of the difference in the plane is vector-displayed. 2. The crystal lattice orientation distribution measuring method according to claim 1, wherein the number of orientations is two. 3. The crystal lattice plane orientation distribution measuring method according to claim 1 , wherein said X-rays are monochromatic parallel radiation with a wavelength of 0.01 nm or more and 0.25 nm or less.

Images