KR20230139294A - Polarized microscopy and intra image field correction method - Google Patents

Abstract

A polarizing microscope device capable of measuring magnetization characteristics within a field of view with high precision and a within-field correction analysis method are provided.
The polarizing microscope device according to the embodiment includes a light source that generates illumination light, a polarizer through which the illumination light is incident and transmits the illumination light including linearly polarized light in the first polarization direction, and simultaneously illuminating a sample with the illumination light including linearly polarized light. An objective lens that transmits the reflected light reflected from the sample, an analyzer that transmits the linearly polarized component of the reflected light in the second polarization direction, an image acquisition unit that acquires an image of the reflected light, and an external magnetic field applied to the sample. It is provided with a generating magnet and an image processing unit that processes the acquired image, wherein the image processing unit calculates a device constant including a polarization rotation angle distribution and an ellipticity square distribution for each ROI in the field of view, and uses the device constant and a hysteresis loop. From the analysis, the rotation angle of the rotation is calculated for each ROI.

Description

Polarized microscope device and intra-field correction analysis method {Polarized microscopy and intra image field correction method}

The present disclosure relates to a polarizing microscope device and an in-field correction analysis method.

Kerr effect microscopy and magnetic domain observation microscopy are based on classical polarization microscopy and are widely used.

Patent Document 1: US Patent No. 4410277 Specification Patent Document 2: Japanese Patent Application Publication No. 2009-042040

Non-patent Document 1: Meguro, “Development of a Kerr effect microscope capable of detecting local magnetization and application to spatial magnetic field detection”, Microscope, Vol.52, No.3 (2017). Non-patent Document 2: P. Wolniansky, et.al, "Magneto-optical measurements of hysteresis loop and anisotropy energy constants on amorphous TbxFe1-x alloys", Applied Physics 60, 346 (1986).

The present disclosure has been made in view of the above problems, and provides a polarizing microscope device capable of measuring magnetic properties with high precision and an in-field correction analysis method.

In order to solve the above-described problem, the technical idea of the present disclosure is to include a light source that generates illumination light; a polarizer through which the illumination light generated by the light source is incident and transmits the illumination light including linearly polarized light in a first polarization direction; an analyzer that transmits a component of linearly polarized light in a second polarization direction of the reflected light reflected from the sample; an image acquisition unit that acquires an image of the reflected light; and an image processing unit that processes the acquired image, wherein the image processing unit calculates a device constant, acquires a plurality of hysteresis loops for each region of interest, and uses the device constant and the plurality of hysteresis loops. A polarizing microscope device is provided, characterized in that it calculates the rotation angle of Kerr rotation for each region of interest.

In order to solve the above-described problem, the technical idea of the present disclosure is to include a light source that generates illumination light; a polarizer through which the illumination light generated by the light source is incident and transmits the illumination light including linearly polarized light in a first polarization direction; an analyzer that transmits a component of linearly polarized light in a second polarization direction of the reflected light reflected from the sample; an image acquisition unit that acquires an image of the reflected light; A magnet that generates an external magnetic field applied to the sample; an image processing unit that processes the acquired image, wherein the image processing unit rotates the angle formed by the first polarization direction and the second polarization direction at predetermined intervals within a predetermined range, while rotating the sample without applying a magnetic field. From the plurality of images acquired by illuminating the polarized illumination light on a sample that can be regarded as a non-magnetic mirror surface by using the state, the polarization rotation angle distribution for each region of interest in the field of view including the plurality of regions of interest, and other When calculating a device constant including the square distribution of the ellipticity by the original image and setting the angle formed by the first polarization direction and the second polarization direction as the first angle, a hysteresis loop of the luminance value for each region of interest is acquired, and when the angle formed by the first polarization direction and the second polarization direction is set to a second angle, the hysteresis loop of the luminance value for each region of interest is acquired, the device constant, the first A polarizing microscope device is provided, wherein a rotation angle of rotation is calculated for each region of interest from analysis using the hysteresis loop at the first angle and the hysteresis loop at the second angle.

In order to solve the above-described problem, the technical idea of the present disclosure is to include a light source that generates illumination light, a polarizer through which the illumination light generated by the light source is incident and transmits the illumination light including linearly polarized light in a first polarization direction, and the linearly polarized light. illuminating a sample with the illumination light including an objective lens that transmits reflected light reflected from the sample, an analyzer that transmits a component of linearly polarized light in the second polarization direction in the reflected light, and an image of the reflected light. An in-field correction analysis method using a polarization microscope device including an image acquisition unit to acquire an image, a magnet to generate an external magnetic field applied to the sample, and an image processing unit to process the acquired image, wherein the first polarization direction and the first polarization direction are 2 By rotating the angle formed by the polarization direction at predetermined intervals within a predetermined range and using the non-magnetic sample or the sample containing a magnetic material in a state without applying a magnetic field, polarized light is applied to a sample that can be regarded as a non-magnetic mirror surface. A first step of calculating device constants including a polarization rotation angle distribution for each region of interest in a field of view including a plurality of regions of interest, and a square distribution of ellipticity due to ellipticalization, from the plurality of images acquired by illuminating the illumination light. ; When the angle formed by the first polarization direction and the second polarization direction is set to the first angle, the polarized illumination light is illuminated on the magnetic material portion of the sample including the magnetic material, and the external When a hysteresis loop of luminance values for each region of interest is acquired from a plurality of images acquired while sweeping a magnetic field, and the angle formed by the first polarization direction and the second polarization direction is set as a second angle, the magnetic body A second step of illuminating the magnetic portion of the sample with the polarized illumination light and acquiring a hysteresis loop of the luminance value for each region of interest from a plurality of images acquired while sweeping the external magnetic field; And a third step of calculating the rotation angle of the cursor rotation for each region of interest from analysis using the device constant, the hysteresis loop at the first angle, and the hysteresis loop at the second angle. Provides an in-field correction analysis method.

According to the present disclosure, a polarizing microscope device capable of measuring magnetization characteristics within a field of view with high precision and a correction analysis method within a field of view are provided.

1 is a configuration diagram illustrating a polarizing microscope device such as a magnetic domain microscope according to a comparative example.
Figure 2 is a schematic diagram illustrating the Kerr effect in a polarizing microscope device such as a magnetic domain microscope according to a comparative example.
Figure 3 is a graph illustrating hysteresis loops in five regions of interest in the field of view measured by a polarizing microscope such as a magnetic domain microscope according to a comparative example, where the horizontal axis represents the external magnetic field and the vertical axis represents the luminance value.
Figure 4 shows the hysteresis loop in five ROIs in the field of view measured with a polarizing microscope device such as a magnetic domain microscope for a comparative example, normalized to the average of the luminance values of each ROI, and further normalized to the average in the ROI at the center of the field of view. will be.
FIG. 5 is a graph illustrating hysteresis loops in five ROIs in the field of view measured with a spot meter according to a comparative example, where the horizontal axis represents the external magnetic field and the vertical axis represents the magnitude of the physical quantity normalized to the average value in the ROI at the center of the field of view. Indicates the rotation angle of rotation.
Figure 6 is a configuration diagram illustrating a polarizing microscope device according to Embodiment 1.
Figure 7 is a flowchart illustrating the in-field correction analysis method according to Embodiment 1.
Fig. 8 is a flowchart illustrating the in-field correction analysis method according to Embodiment 1.
Fig. 9 is a flowchart illustrating the in-field correction analysis method according to Embodiment 1.
FIG. 10 is a diagram illustrating images acquired by the polarizing microscope device according to Embodiment 1.
FIG. 11 is a diagram illustrating conversion of a plurality of images acquired by the polarizing microscope device according to Embodiment 1 into analyzer transmission angle data for each ROI.
Figure 12 is a graph illustrating analyzer transmission angle data acquired by the polarizing microscope device according to Embodiment 1, where the horizontal axis represents the angle of the polarization direction of the analyzer, and the vertical axis represents the luminance value.
Figure 13 shows (a) to (c) the distribution of luminance unevenness a(x, y), distribution of polarization rotation angle Θ ₀ (x, y) in the polarizing microscope device 1 according to Embodiment 1, and a diagram illustrating the distribution of the square of ellipticity η η ² (x, y).
FIG. 14 is (a) a diagram illustrating an image including a plurality of ROIs acquired by the polarizing microscope device according to Embodiment 1, and (b) is a diagram illustrating images acquired by the polarizing microscope device according to Embodiment 1 for each ROI. This is a graph illustrating the extracted hysteresis loop, where the horizontal axis represents the external magnetic field and the vertical axis represents the luminance value.
Fig. 15 is a graph illustrating the hysteresis loop and fitting function acquired by the polarizing microscope device according to Embodiment 1, where the horizontal axis represents the external magnetic field and the vertical axis represents the luminance value.
FIG. 16 is a diagram illustrating equations used in the in-field correction analysis method according to Embodiment 1.
Fig. 17 is a graph illustrating the definitions of contrast C(x, y), coercive force Hc(x, y), and slope α(x, y) acquired by the polarizing microscope device according to Embodiment 1.
Fig. 18 is a graph illustrating the definitions of contrast C(x, y), coercive force Hc(x, y), and slope α(x, y) acquired by the polarizing microscope device according to Embodiment 1.
FIG. 19 is a diagram illustrating the rotation and ellipticalization of the polarization plane in the polarizing microscope device according to Embodiment 1 when no external magnetic field is applied.
FIG. 20 is a diagram illustrating the optical path until the reflected light reflected from the sample reaches the imaging surface of the image acquisition unit in the absence of a beam splitter.
FIG. 21 is a diagram illustrating the optical path until the reflected light reflected from the sample reaches the imaging surface of the image acquisition unit in the presence of the beam splitter.
FIG. 22 is a diagram illustrating the rotation and ellipticalization of the polarization plane when an external magnetic field is applied in the polarizing microscope device according to Embodiment 1.
Figure 23 shows (a) normalizing the high and low gray level difference range of the hysteresis loop to the gray level at the center of the field of view when the angle of the polarization direction of the analyzer is set to +4°, and (b) shows the hysteresis loop The contrast is simply divided by the average luminance value and normalized based on the value at the center of the field of view, (c) is a diagram illustrating the distribution of magnetization contrast of the sample acquired by the polarizing microscope device according to Embodiment 1, (d) ) is a diagram illustrating the distribution of magnetization contrast acquired by the spot meter.
24, (a) to (c) are diagrams illustrating the results of quantitative analysis of the polarizing microscope device 1 according to Embodiment 1, and (d) to (f) are diagrams illustrating the results of spot measurement. It is a drawing.
FIG. 25 is a diagram illustrating the characteristics of the polarizing microscope device and the in-field correction analysis method according to Embodiment 1.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. The same reference numerals are used for the same components in the drawings, and duplicate descriptions thereof are omitted.

The Kerr effect microscope is capable of obtaining a magnetic hysteresis loop at one point on a sample by measuring the amount of change in the polarization component of the reflected light by incident laser light while sweeping an external magnetic field on a sample containing a magnetic material such as a magnetic thin film. there is. On the other hand, in a magnetic domain observation microscope, an external magnetic field is applied to a sample containing a magnetic material such as a magnetic thin film, the light from an incoherent light source is polarized by a polarizer, and the change in the polarization component of the reflected light when incident on the sample is observed. By recording as a change in luminance of the image sensor, the magnetic domain pattern within the field of view can be recorded.

Even with a magnetic domain observation microscope, a hysteresis loop for each pixel can be obtained by recording the luminance for each pixel while sweeping an external magnetic field. However, in a magnetic domain observation microscope, the gap in polarization characteristics within the field of view is larger than the polarization change occurring in the sample. Therefore, it is difficult to quantitatively evaluate the hysteresis loop within the field of view, making it impossible to measure magnetization characteristics with high precision.

(Comparative example)

Before explaining the polarizing microscope device according to Embodiment 1, the polarizing microscope according to Comparative Example and its problems will be explained. This makes the polarizing microscope of this embodiment more clear. In addition, the configuration and problems of the comparative example are also included in the scope of the technical idea of the embodiment.

1 is a configuration diagram illustrating a polarizing microscope device such as a magnetic domain microscope according to a comparative example. As shown in FIG. 1, the polarizing microscope device 101 according to the comparative example includes a light source 10, a lens 11, a polarizer 12, and a beam splitter 13. , Objective Lens (14), Sample Stage (15), Magnet (16), Analyzer (17), Imaging Lens (18), and Image Acquisition It is equipped with a unit (19). A sample 20 is placed on the sample stand 15.

Here, for convenience of explanation of the polarizing microscope device 101, the XYZ orthogonal coordinate axis system is introduced. For example, the direction perpendicular to the upper surface of the sample stand 15 is referred to as the Z-axis direction, and the plane parallel to the upper surface of the sample stand 15 is referred to as the XY plane. Below, each configuration will be described.

The light source 10 generates illumination light 21. The light source 10 is, for example, a halogen lamp, a white LED (Light Emitting Diode), or the like. The illumination light 21 may be white light. The light source 10 emits the generated illumination light 21 to the polarizer 12. A lens 11 may be disposed between the light source 10 and the polarizer 12. Illumination light 21 converted into parallel light by the lens 11 is incident on the polarizer 12.

The polarizer 12 is disposed at or near the illumination pupil 23 of the illumination light 21. Illumination light 21 generated by the light source 10 is incident on the polarizer 12 . The polarizer 12 converts the incident illumination light 21 into illumination light 21 containing linearly polarized light. For example, the polarizer 12 transmits illumination light including linearly polarized light in the first polarization direction. For example, the first polarization direction is the Z-axis direction. In this case, the polarizer 12 converts the illumination light 21 to include linearly polarized light in the Z-axis direction. That is, the transmission axis of the polarizer 12 is in the Z-axis direction. The illumination light 21 converted to include linearly polarized light by the polarizer 12 is incident on the beam splitter 13.

The beam splitter 13 reflects a part of the incident illumination light 21 toward the objective lens 14. The illumination light 21 reflected by the beam splitter 13 includes linearly polarized light in the X-axis direction, for example. The objective lens 14 is disposed between the beam splitter 13 and the sample 20. The objective lens 14 is disposed at or near the pupil plane 24. The objective lens 14 focuses the illumination light 21 reflected by the beam splitter 13 onto the sample 20. The objective lens 14 illuminates the sample 20 with illumination light 21 containing linearly polarized light. In addition, the magnetic thin film sample in the present invention is a non-patterned thin film sample or an island pattern of a sufficiently large size (for example, Φ is several hundred μm).

The sample 20 contains a magnetic material such as a magnetic thin film. The sample 20 may be, for example, a magnetic thin film formed on a wafer. Additionally, the sample 20 may be a magnetic thin film used in MRAM (Magnetoresistive Random Access Memory). An external magnetic field in the Z-axis direction may be applied to the magnetic material of the sample 20 placed on the sample stand 15, for example. When the direction of the external magnetic field changes upward or downward, the magnetization direction of the magnetic material may change upward or downward. In addition, the sample 20 in the calibration described later may be a non-magnetic sample 20.

The magnet 16 is disposed between the sample 20 and the objective lens 14. The magnet 16 is, for example, an electromagnet that can modulate the strength and direction of the magnetic field on the sample. Additionally, the magnet 16 may be a permanent magnet. The magnet 16 generates an external magnetic field applied to the sample 20. Accordingly, the magnet 16 applies an external magnetic field to the sample 20. For example, when examining the free layer composed of a perpendicular magnetic anisotropy material of MRAM, the Polar Kerr Effect (also called the Polar Kerr Effect) is detected. Accordingly, the magnet 16 applies a magnetic field in a direction perpendicular to the surface of the sample 20. When the magnet 16 is an electromagnet, an external magnetic field within a predetermined range from the +Z direction to the -Z direction is applied to the sample 20 by controlling the current value and direction flowing through the electromagnet.

In addition, in order to detect the polarization effect, the incident light incident on the sample 20 must be linearly polarized so that it is incident perpendicular to the surface of the sample 20 and the oscillation direction of the electric field is within the XY plane. The polarization state of the sample 20 slightly changes due to the Kerr effect depending on the magnetic field strength (H) of the external magnetic field. The Kerr effect includes rotation of the polarization axis (called Kerr rotation) and Kerr ellipticity.

Figure 2 is a schematic diagram illustrating the Kerr effect in a polarizing microscope device such as a magnetic domain microscope according to a comparative example. As shown in FIG. 2, the Kerr effect includes ellipticalization including the rotation angle ΔΘ of the rotation of the polarization axis (also called the rotation angle ΔΘ of the rotation of the polarization direction) and the change in ellipticity Δη. The change Δη in ellipticity is, for example, tan ^-1 (OB/OA). Here, OA is the minor axis and OB is the major axis. The reflected light 22 reflected from the sample 20 is modulated into a polarized state by the Kerr effect caused by the sample 20.

Returning to Figure 1, the reflected light 22 reflected from the sample 20 is incident on the objective lens 14. The objective lens 14 transmits the reflected light 22 in which the illumination light 21 is reflected from the sample 20. The reflected light 22 passes through the objective lens 14 and the beam splitter 13 and enters the analyzer 17.

The analyzer 17 is disposed between the beam splitter 13 and the image acquisition unit 19. The analyzer 17 transmits the linearly polarized component of the reflected light 22 in the second polarization direction. The second direction is, for example, the Y-axis direction. In this case, the analyzer 17 converts the reflected light 22 to include linearly polarized light in the Y-axis direction. That is, the transmission axis of the analyzer 17 is in the Y-axis direction. The direction (transmission axis) of linearly polarized light transmitted by the analyzer 17 may be arranged orthogonal to the direction (transmission axis) of linearly polarized light transmitted by the polarizer 12. For example, the direction of linearly polarized light transmitted by the polarizer 12 may be the X-axis direction, and the direction of the linearly polarized light transmitted by the analyzer 17 may be the Y-axis direction. In this way, an arrangement in which the transmission axis of the analyzer 17 is orthogonal to the transmission axis of the polarizer 12 is called a cross Nicole arrangement. The orthogonal Nicol arrangement can detect changes in polarized light, including linearly polarized light, as changes in luminance with high sensitivity. The reflected light 22, which includes slightly rotated and elliptical polarized light from the sample 20, is transmitted at an intensity equal to the square of the amplitude projected on the transmission axis of the analyzer 17. The reflected light 22 that has passed through the analyzer 17 is incident on the image acquisition unit 19 through the imaging lens 18, for example.

The image acquisition unit 19 detects a change in the polarization component of the reflected light 22 as a change in intensity. Thereby, the image acquisition unit 19 acquires the image of the reflected light 22. For example, the image acquisition unit 19 acquires an image within the field of view. The image acquisition unit 19 is, for example, a camera. The image acquisition unit 19 may be, for example, an image sensor including a PD (Photodiode) array. Additionally, the image acquisition unit 19 is not limited to cameras and image sensors as long as it can acquire images. The imaging surface 26 of the image acquisition unit 19 is in an image conjugate relationship with the measurement surface 25 of the sample 20.

η의 정량치에는 주목하지 않았다. 또 자구 현미경은, 시야내의 장소간의 시인성(Pattern Contrast)의 정량 해석 오차에 주목하지 않았다.Here, the rotation angle ΔΘ of polarization direction rotation occurring in the above-described MRAM sample 20 is typically 0.1° or less. Additionally, the change in ellipticity Δη is also a slight change at the same level as the rotation angle ΔΘ. The ellipticity η is an ellipticity angle (angular quantity) defined as arctan, which is the flatness of an ellipse. A magnetic domain microscope is a microscope that focuses on magnetic domains (patterns) within the field of view. Therefore, it is sufficient for a magnetic domain microscope to be able to visually see the magnetic domain within the field of view, and the change in rotation angle ΔΘ and ellipticity

No attention was paid to the quantitative value of η. Additionally, magnetic domain microscopy did not pay attention to quantitative analysis errors in visibility (pattern contrast) between places in the field of view.

As shown in the upper right corner of FIG. 1, in the case of the MRAM sample 20, the rotation angle ΔΘ and the change in ellipticity Δη are more than one order of magnitude smaller than the error due to the device (tool), making it difficult to detect the Kerr effect generated in the sample. It can be difficult. Meanwhile, in the case of a typical magnetic sample, the Kerr effect occurring in the sample is larger than the error caused by the device (tool), so the Kerr effect can be detected.

First, in the following description, spot measurement using laser light, which is a standard device, will be described. This is also a well-known device and most of the products are also available in the market. Hereinafter, a device that performs spot measurement is referred to as a spot measuring device.

A spot meter measures using the differential polarization method. The spot meter sweeps the magnetic field applied to the sample by incident laser light containing linearly polarized light on the sample. And while the spot measuring instrument sweeps the magnetic field, the reflected light, including the polarized light modulated by the sample, is split into s-polarized light and p-polarized light at the polarization separation prism. In this way, the rotation angle in the Kerr rotation generated in the sample is measured by lock-in differential processing by the two photo detectors of the spot meter. Therefore, a hysteresis loop can be obtained by plotting the magnetic field on the horizontal axis and the rotation angle on the vertical axis. In general, the MOKE device (Magneto-optical Kerr Effect), a Kerr effect measurement device, refers to these. Since the obtained hysteresis loop measures a physical quantity, the quantitative characteristics of the magnetic thin film can be evaluated. In addition, as a method of changing the evaluation position while moving the sample, only the center of the optical system is used and all reflected light flux is detected by a photodetector, so the evaluation position does not include device error.

However, in the magnetic domain microscope described above, when an image is acquired while sweeping an external magnetic field, a hysteresis loop can be obtained by plotting the external magnetic field on the horizontal axis and the luminance value on the vertical axis. The hysteresis loop obtained in this way is roughly correlated with the hysteresis loop in spot measurement using laser light. Magnetic domain microscopy can also obtain hysteresis loops as a function. However, the ability to obtain a hysteresis loop in magnetic domain microscopy is only a standard. For example, in magnetic domain microscopy, the inventor discovered that hysteresis loops obtained from a plurality of different measurement coordinates in the field of view have large differences (errors) between the measurement coordinates. This will be explained using Figure 3 below.

FIG. 3 is a graph illustrating hysteresis loops in five regions of interest (hereinafter referred to as ROI) in the field of view measured by a polarizing microscope such as a magnetic domain microscope according to a comparative example, and the abscissa is the external axis. It represents the magnetic field, and the vertical axis represents the luminance value. The five ROIs in FIG. 3 include the center of the visual field (CE), upper right (UR), upper left (UL), lower right (LR), and lower left (LL) of the visual field.

Figure 4 shows the hysteresis loop in five ROIs in the field of view measured with a polarizing microscope such as a magnetic domain microscope for a comparative example, normalized to the average of the luminance values of each ROI, and further, the average in the ROI at the center of the field of view (CE). It is standardized. In Figure 4, the values after normalization are shown in a table. That is, it represents the difference between the maximum and minimum luminance values in each ROI and the ratio of the average of the luminance values in each ROI to the center of view (CE).

Figure 5 is a graph illustrating hysteresis loops in five ROIs in the field of view measured with a spot meter related to a comparative example, where the horizontal axis represents the external magnetic field and the vertical axis is normalized to the average value in the ROIs at the center of the field of view (CE). It is a physical quantity and represents the rotation angle of rotation. In Figure 5, the values after normalization are shown in a table. That is, it represents the rotation angle of the curr rotation in each ROI and the ratio to the center of view (CE).

As shown in Fig. 4, in magnetic domain microscopy, the magnetization contrast of each ROI in the field of view has a difference of approximately ±20%. On the other hand, as shown in FIG. 5, in a spot measuring instrument, the rotation angle of each ROI within the field of view has a difference of approximately ±5%. In this way, in magnetic domain microscopy, the difference in magnetization contrast of each ROI in the field of view makes it difficult to quantitatively detect and evaluate the hysteresis loop.

Therefore, if the difference (error) in magnetization contrast can be corrected for each position (ROI coordinates or image height) in the field of view of the magnetic domain microscope, local hysteresis loops in the field of view can be treated quantitatively. Therefore, measurement of the local hysteresis loop in a plane (two-dimensional) using a magnetic domain microscope is established. This means that treatment time is greatly improved for spot measuring instruments using laser light. In this embodiment, local hysteresis loops are measured in a plane (two-dimensional) using a polarizing microscope such as a magnetic domain microscope. Hereinafter, this measurement will be referred to as Imaging MOKE (Imaging MOKE).

(Embodiment 1)

Next, the polarizing microscope device and in-field correction analysis method according to Embodiment 1 will be described. Figure 6 is a configuration diagram illustrating a polarizing microscope device according to Embodiment 1. As shown in FIG. 6, the polarizing microscope device 1 of the present embodiment further includes an image processing unit 30 in addition to the configuration of the polarizing microscope device 101 of the comparative example. The image processing unit 30 processes the image acquired by the image acquisition unit 19.

For example, the image processing unit 30 performs image processing in the in-field correction analysis method using the polarizing microscope device 1 shown below.

7 to 9 are flowcharts illustrating the in-field correction analysis method according to Embodiment 1. The in-field correction analysis method of this embodiment includes (i) pre-calibration without magnetic field sweeping, shown in FIG. 7, and (ii) image acquisition by magnetic field sweeping on a sample containing a magnetic material and analysis for each ROI, shown in FIG. 8. , (iii) shown in FIG. 9, and a correction process for the Kerr effect using the measurement data (i) and (ii) and the physical model equation. In addition, in the following description, the direction of rotation is defined as + for clockwise (CW) and - for counterclockwise (CCW) when viewed from the direction of light travel.

<(i) Pre-calibration without magnetic field sweeping>

In pre-calibration, first, a mirror surface sample is placed on the sample table 15 (step S11 in FIG. 7). The mirror surface sample may be an unpatterned magnetized thin film, a silicon wafer, or the like. In this way, the non-magnetic sample 20 may be disposed as a mirror surface sample, or the sample 20 may be disposed as a non-magnetic mirror surface sample using an unpatterned magnetic thin film in a state where the illumination light 21 is not applied as a magnetic field.

Next, the positions of the polarizer 12 and the analyzer 17 are adjusted to achieve orthogonal Nicol arrangement (step S12 in FIG. 7). Typically, on the upper surface of the sample stage 15, the polarization direction of the polarized light included in the illumination light 21 is aligned with either the X-axis direction or the Y-axis direction, for example, the X-axis direction.

Next, the polarization direction of the polarizer 12 is fixed. For example, the polarization direction of the polarizer 12 is fixed so that it is in the X-axis direction on the upper surface of the sample stand 15. In this case, the luminance of the image acquired by the image acquisition unit 19 becomes low at the position of the analyzer 17 in the orthogonal Nicol arrangement.

At this time, the position of the analyzer 17 is set to the zero point. Hereinafter, the angle Θ _a of the polarization direction of the analyzer 17 at the zero point is defined as 0°. In addition, a of angle Θ _a is a subscript of Θ.

Next, the illumination light 21 is illuminated on the mirror surface sample while rotating the angle between the polarization direction of the polarizer 12 and the polarization direction of the analyzer 17 at predetermined intervals within a predetermined range. Thereby, a plurality of images are acquired (step S13 in FIG. 7). The predetermined range of the angle formed by the polarization direction of the polarizer 12 and the polarization direction of the analyzer 17 is, for example, within a range of ±10°, and the predetermined interval is, for example, 1°. In that case, 21 images are acquired while rotating at 1° intervals within a range of approximately ±10°. Additionally, the predetermined range is not limited to ±10°. Additionally, the predetermined interval is not limited to 1°. Additionally, the number of images acquired is not limited to 21.

FIG. 10 is a diagram illustrating images acquired by the polarizing microscope device 1 according to Embodiment 1. Figure 10 simply shows 11 images. As shown in Fig. 10, in the case of an orthogonal Nicol arrangement where the angle Θ _a = 0 of the polarization direction of the analyzer 17 at the zero point, the luminance value is, in principle, the minimum.

Next, from the plurality of acquired images, a data group of analyzer transmission angle data (Through angle of analyzer data) that passed through the analyzer 17 is calculated for each ROI in the field of view including the plurality of ROIs, and calculation processing is performed ( Step S14 in FIG. 7).

FIG. 11 is a diagram illustrating conversion of a plurality of images acquired by the polarizing microscope device 1 according to Embodiment 1 into analyzer transmission angle data for each ROI. As shown in Fig. 11, the plurality of images are converted into analyzer transmission angle data for each ROI. Here, the ROI is not 1 pixel, and in order to reduce noise, it is desirable to do binning, for example, into a rectangle (10 pixels Ø 10 pixels) or a circle (ϕ20 pixels). In addition, some symbols are omitted so as not to complicate the drawing.

FIG. 12 is a graph illustrating analyzer transmission angle data acquired by the polarizing microscope device 1 according to Embodiment 1, where the horizontal axis represents the angle Θa in the polarization direction of the analyzer 17, and the vertical axis represents the average luminance within the ROI. Indicates the numerical value (Gray Level). As shown in FIG. 12, the smaller the angle Θ _a of the polarization direction of the analyzer 17, the smaller the luminance value tends to be. However, the position where the luminance value becomes smallest for each ROI is shifted from the angle Θ _a = 0 for each ROI. The smallest angle Θ _a is called the polarization rotation angle Θ ₀ .

Next, the analyzer transmission angle data for each ROI is fitted with a predetermined function. As will be described later, the function used for fitting is the well-known Malus law or its equivalent in the field of polarization optics (step S15 in FIG. 7).

13(a) to 13(c) show, in the polarizing microscope device 1 according to Embodiment 1, distribution of luminance unevenness a(x, y), distribution of polarization rotation angle Θ ₀ (x, y), and , This is a diagram illustrating the distribution of the square of ellipticity η η ² (x, y). As shown in Figure 13, by fitting the analyzer transmission angle data for each ROI with a predetermined function, the distribution of luminance unevenness a(x, y) for each ROI, distribution of polarization rotation angle Θ ₀ (x, y), And the distribution of the square of the ellipticity η η ² (x, y) can be calculated (step S16 in FIG. 7). In the following description, the distribution of the luminance non-uniformity a(x, y) for each ROI, the distribution of the polarization rotation angle Θ ₀ (x, y), and the distribution of the square of the ellipticity η η ² (x, y) are used as device functions or These are called device constants.

In this way, the image processing unit 30 rotates the angle formed by the polarization direction of the polarizer 12 and the polarization direction of the analyzer 17 at predetermined intervals within a predetermined range, while rotating the non-magnetic sample 20 or the sample containing a magnetic material. A plurality of images are acquired by illuminating the sample used as a non-magnetic mirror surface sample with illumination light 21 without applying a magnetic field to (20). Then, the image processing unit 30 calculates device constants including the luminance non-uniformity distribution for each ROI in the field of view including the plurality of ROIs, the polarization rotation angle distribution, and the square distribution of the ellipticity due to ellipticalization from the acquired plurality of images.

At this time, the analyzer 17 may be rotated within a predetermined range based on the orthogonal Nicol arrangement with respect to the polarizer 12. Additionally, the image processing unit 30 may calculate a device constant (device function) for each ROI using Malus' law. Additionally, it goes without saying that it is equivalent to acquiring data by fixing the orientation of the analyzer 17 and rotating the polarizer 12 within a predetermined range.

<(ii) Image acquisition by magnetic field sweeping and analysis for each ROI>

Next, image acquisition by magnetic field sweeping and analysis for each ROI will be explained. First, the sample 20 containing the magnetic material to be evaluated is placed on the sample stand 15 (step S21 in FIG. 8). The sample 20 is, for example, a wafer on which a magnetic thin film is formed.

Next, the polarizer 12 and the analyzer 17 are adjusted to the orthogonal Nicol arrangement and set to the zero point (step S22 in FIG. 8). After that, the angle Θ _a of the polarization direction of the analyzer 17 is set to some angle β ₁ within the angle range of the above-described calibration (step S23 in FIG. 8). The angle β ₁ is usually about 5°, for example.

Next, an image is acquired while sweeping the magnetic field at approximately ±10 times the coercive force H _c of the sample 20 to be evaluated (a level that reaches saturation magnetization with a margin) (step S24 in FIG. 8). At this time, images are acquired with sufficient time resolution capability. When the magnet 16 is an electromagnet, the control value of the magnetic field and the imaging time of the image acquisition unit 19 are linked by a timestamp. For the interval of image acquisition, the exposure time, frame rate, and magnetic field sweeping time per image are set so that a number of measurement points fall within the rising/falling slope of the hysteresis loop.

FIG. 14(a) is a diagram illustrating an image including a plurality of ROIs acquired by the polarizing microscope device 1 according to Embodiment 1. FIG. 14(b) is a graph illustrating a hysteresis loop extracted for each ROI from an image acquired by the polarizing microscope device 1 according to Embodiment 1, where the horizontal axis represents an external magnetic field and the vertical axis represents a luminance value. As shown in FIGS. 11(a) and 11(b), the magnetic field sweeping data (Through Magnetic field) group of magnetic field swept data is processed for each ROI of the acquired image (step S25 in FIG. 8). It is desirable to align ROI coordinates and pixel binning processing within the ROI through calibration. By plotting the magnetic field on the horizontal axis and the luminance value on the vertical axis of the processed data group, an uncorrected hysteresis loop can be obtained for each ROI, as shown in Fig. 14(b).

Next, the hysteresis loop for each ROI is fitted with an approximation function. FIG. 15 is a graph illustrating the hysteresis loop measurement value and its fitting function curve acquired by the polarizing microscope device 1 according to Embodiment 1, where the horizontal axis represents the external magnetic field and the vertical axis represents the luminance value. As shown in Fig. 15, the hysteresis loop for each acquired ROI is fitted with an approximation function. The approximate function may be an empirical function. The approximation function does not have to be a physical model equation.

FIG. 16 is a diagram illustrating equations used in the in-field correction analysis method according to Embodiment 1. As shown in Fig. 16, in this embodiment, the hysteresis loop for each ROI shown in Figs. 14 and 15 is fitted with the following equation (0).

I=D+Rtanh(α(H±H _c )) (0)

By this fitting, from the shape of the hysteresis loop, the magnetization indices of Contrast: C(x, y), Coercivity: H _c (x, y), and Slope: α(x, y) are obtained for each ROI. A map within the field of view can be obtained (step S26 in FIG. 8).

17 and 18 show definitions of contrast C (x, y), coercive force H _c (x, y), and slope α (x, y) acquired by the polarizing microscope device 1 according to Embodiment 1. This is an example graph. As shown in Figure 17, contrast C(x, y) can be calculated by contrast C=0.5×(R/D) from range R and luminance D when fitting equation (0) to the hysteresis loop. there is. Here, the range R is the difference between the luminance value in the saturation magnetization state on the plus side and the luminance value in the saturation magnetization state on the minus side. The luminance D is the average luminance value of the saturation magnetization states of the plus side and the minus side.

The coercive force H _c (x, y) is obtained from the intersection of the midpoint of the two saturation magnetizations of the hysteresis loop and the hysteresis loop, as shown in FIGS. 17 and 18. The slope α(x, y) is obtained from the slope of the hysteresis loop rise, as shown in FIGS. 17 and 18. Additionally, slope is also called stiffness.

In this way, in (ii) of the present embodiment, when the image processing unit 30 sets the angle formed by the polarization direction of the polarizer 12 and the polarization direction of the analyzer 17 to angle β ₁ , the image processing unit 30 includes the magnetic material. A plurality of images are acquired while illuminating the magnetic portion of the sample 20 with the illumination light 21 and simultaneously sweeping an external magnetic field. Then, the image processing unit 30 acquires a hysteresis loop of the luminance value for each ROI from the plurality of acquired images. The image processing unit 30 further adjusts the contrast C(x, y), coercive force H _c (x, y), and and slope α(x, y) is calculated. The contrast C (x, y), the coercive force H _c (x, y), and the slope α (x, y) in the case of the angle β ₁ of the polarization direction of the analyzer 17 are respectively defined by the contrast C ₁ (x, y), the coercive force H _c1 (x, y), and the slope α ₁ (x, y).

Next, the angle Θ _a of the polarization direction of the analyzer 17 is defined as the angle An angle different from ₁ Set to ₂ (step S27). β ₂ is an angle within the angle range of the above-described calibration. For example, β ₂ = -β ₁ , etc., but there is no particular limitation as long as it is within the angle range of the calibration. Next, as in the case of angle β ₁ , an image is acquired while sweeping the magnetic field at approximately ±10 times the coercive force H _c of the sample 20 to be evaluated (step S28 in FIG. 8). At this time, the image acquisition operation is the same as in the case of angle β ₁ , such as acquiring the image with sufficient time resolution ability. Then, the magnetic field sweeping data group is processed for each ROI on the acquired image (step S29 in FIG. 8). Then, by fitting the hysteresis loop for each ROI with an approximation function, the contrast C (x, y), coercive force H _c (x, y), and slope α (x, y) can be obtained for each ROI (step S30 in FIG. 8). .

In this way, in (ii) of the present embodiment, the image processing unit 30 determines the angle formed by the polarization direction of the polarizer 12 and the polarization direction of the analyzer 17. Even when set to ₂ , a plurality of images are acquired while illuminating the magnetic material portion of the sample 20 containing a magnetic material with the illumination light 21 and simultaneously sweeping an external magnetic field. Then, the image processing unit 30 acquires a hysteresis loop of the luminance value for each ROI from the plurality of acquired images. The image processing unit 30 further adjusts the contrast C(x, y), coercive force H _c (x, y), and and slope α(x, y) is calculated. Angle of polarization direction of analyzer 17 The contrast C (x, y), coercive force H _c (x, y), and slope α (x, y) in the case of ₂ are respectively contrast C ₂ (x, y) and coercive force H _c2 (x, y). , and the slope α ₂ (x, y).

RAM 디바이스로서의 최종 성능(동작 속도, 소비 전력, 신뢰성)에 관련된 지표이다. 아울러, 이 3개의 지표 이외에 다른 지표를 정의해도 상관없다. 발명자에 의한 해석에 의하면, 자화 반전이 발생하는 외부 자장의 지표인 보자력H_c(x, y)는, 보정 없이 스폿 계측과 거의 유사한 결과를 얻을 수 있다. 그러나, 포화 자화량 M_s(x, y)는 자화 콘트라스트에 상관하는 특성량이지만, 전술한 비교예에서의 도 3 및 도 4에서 설명한 것처럼, 스폿 계측과의 괴리가 커서 보정이 필요하다. 또 슬로프α(x, y)는, 보자력H_c(x, y)와 M_s(x, y)가 뒤섞인 지표로서, 역시 보정이 필요하다.Contrast C (x, y), coercive force H _c (x, y), and slope α (x, y) show the important and characteristic shape of the hysteresis loop. All, for example:

This is an indicator related to the final performance (operation speed, power consumption, reliability) of a RAM device. Additionally, it is okay to define other indicators in addition to these three indicators. According to the inventor's analysis, the coercive force H _c (x, y), which is an indicator of the external magnetic field where magnetization reversal occurs, can obtain results almost similar to spot measurements without correction. However, although the saturation magnetization amount M _s (x, y) is a characteristic quantity that is correlated with the magnetization contrast, as explained in FIGS. 3 and 4 in the above-mentioned comparative example, the difference from spot measurement is large and correction is necessary. In addition, the slope α(x, y) is an index that is a mixture of the coercive forces H _c (x, y) and M _s (x, y), so correction is also required.

<(iii) Correction processing using measurement data and physical model equation>

Θ(x, y)를 산출한다(도 9의 단계 S31). 구체적으로는, 화상 처리부(30)는, 장치 정수, 각도β₁에서의 히스테리시스 루프 및 각도β₂에서의 히스테리시스 루프를 이용한 해석으로부터 ROI마다 커 회전의 회전각ΔΘ(x, y)을 산출한다. 물리 모델식을 도 16의 (1)식 및 (2)식에 나타낸다. 또, 하기에도 (1)식 및 (2)식을 나타내는데, 지면 관계상, (1)식을 분할하여 (1-1)~(1-4)식으로 나타낸다.Next, using the physical model equation, for each ROI (x, y), the rotation angle of the polarization direction rotation

Θ(x, y) is calculated (step S31 in FIG. 9). Specifically, the image processing unit 30 calculates the rotation angle ΔΘ(x, y) of the cursor rotation for each ROI from analysis using the device constant, the hysteresis loop at the angle β ₁ , and the hysteresis loop at the angle β ₂ . The physical model equations are shown in equations (1) and (2) in FIG. 16. In addition, equations (1) and (2) are shown below, but for reasons of space, equation (1) is divided into equations (1-1) to (1-4).

2ΔΘ(x, y)=(P1-P2)/P3 (1-1)

P1=C ₂ (x, y){(1-2η ² (x, y)) sin ² (β ₂ -Θ(x, y))+η ² (x, y)}

(1-2)

P2=AC ₁ (x, y){(1-2η ² (x, y)) sin ² (β ₁ -Θ(x, y))+η ² (x, y)}

(1-3)

P3=(1-2η ² (x, y)){sin(2(β ₂ -Θ(x, y))-Asin(2(β ₁ -Θ(x, y))}

(1-4)

Here, A in equations (1), (1-3), and (1-4) is the following equation (2).

A={1+cos(2(β ₁ -Θ(x, y)}/{1+cos(2(β ₂ -Θ(x, y)}

(2)

The explanation and derivation of equations (1) and (2) are shown in the examples. In (ii), the polarization direction of at least the analyzer 17 is set to two angles β ₁ and β ₂ and an image is captured while sweeping the magnetic field. Then, by analyzing the hysteresis loop for each ROI in the image, two contrast values C ₁ (x, y) and C ₂ (x, y) are calculated for each ROI.

Accordingly, the image processing unit 30 determines the calibration result of the device constant in (i) described above, the setting values of angles β ₁ and β ₂ of the polarization direction of the analyzer 17 in (ii), and By substituting the contrast C ₁ (x, y) and C ₂ (x, y) into the physical model equations (1) and (2), the measured value of the rotation angle ΔΘ (x, y) of the corrected Kerr rotation can be obtained. there is.

In equations (1) and (2), the values measured in <(i) pre-calibration>, <(ii) image acquisition by magnetic field sweeping and analysis for each ROI>, and the set angles β ₁ and β ₂ are . Therefore, the rotation angle ΔΘ(x, y) of the Kerr rotation occurring in the sample 20 in ROI(x, y) becomes a corrected quantitative value.

<Propagation of polarization error occurring in the polarization microscope itself to Kerr rotation>

Next, we will explain how the polarization error (device constant) generated in the polarization microscope itself, which was the subject discussed above, is propagated by the Kerr rotation generated by the sample 20.

First, consider the case where no external magnetic field is applied to the sample 20. Also, consider ROI(x, y) as an arbitrary coordinate. An ideal orthogonal Nicol arrangement is set in which the polarization direction of the polarizer 12 is in the X-axis direction and the polarization direction of the analyzer 17 is in the Y-axis direction. At this time, as linearly polarized light with an amplitude in the

FIG. 19 is a diagram illustrating the rotation and ellipticalization of the polarization plane in the polarizing microscope device 1 according to Embodiment 1 when no external magnetic field is applied. The direction of the incident linearly polarized light is the X-axis, and the direction of the analyzer is the Y-axis. As shown in FIG. 19, the polarization direction OA of the reflected light 22 is rotated about the

This Θ rotation and ellipticalization are caused by the curvature of the objective lens 14, the anti-reflective film, the amplitude difference (ellipticization) between s-polarized light and p-polarized light caused by the coating of the beam splitter 13, and the s-polarized light and This is due to the phase difference (rotation of the polarization plane) of p-polarized light. The image acquisition unit detects the sum of the projection components of OA and OB on the Y axis. Additionally, since the optical path until it reaches the coordinates (image position) of the imaging device of the image acquisition unit 19 is different for each higher value, the amount of change in polarization changes depending on the higher value.

FIG. 20 is a diagram illustrating an optical path until the reflected light 22 reflected from the sample 20 reaches the imaging surface 26 of the image acquisition unit 19 in the absence of the beam splitter 13. . FIG. 21 is a diagram illustrating the optical path until the reflected light 22 reflected from the sample 20 reaches the imaging surface 26 of the image acquisition unit 19 in the presence of the beam splitter 13. . As shown in FIGS. 20 and 21, when the beam splitter 13 is present, the optical path until it reaches the imaging surface 26 of the image acquisition unit 19 is different for each higher value. Additionally, the angles at which the three light rays are incident on the beam splitter 13 are different depending on the object position (sample position). In a typical beam splitter 13, reflectance and transmittance are controlled using a dielectric multilayer film, but polarization dependence is not taken into consideration.

Therefore, the degree of rotation and ellipticalization of the polarization plane changes depending on this angle difference. As an example, the dielectric multilayer film of the beam splitter 13 is used, but this is the same even if it is an anti-reflection coating on each side of the objective lens constituting the optical path, and the variation of the polarization component as a whole of the optical path reaching the upper value varies depending on the higher value. . Therefore, the amount of change in the polarization direction changes depending on the difference value. This is, to varying degrees, inevitable.

sinΘ, ηcosΘ가 된다. 이들은, 서로 직교되는 진동면을 갖기 때문에 비간섭이다. 따라서, 검광자(17)를 투과하는 강도는 (1-η²) sin²Θ+η²cos²Θ가 되며, 또한 변형되어 이하 및 도 16의 (3)식을 얻는다. When the energy is set to OA ² + OB ² = 1, the projection components of OA and OB to the analyzer 17 are respectively

becomes sinΘ, ηcosΘ. They are non-interfering because they have vibration planes that are orthogonal to each other. Therefore, the intensity passing through the analyzer 17 becomes (1-η ² ) sin ² Θ+η ² cos ² Θ, and is further modified to obtain equation (3) below and in FIG. 16.

I(Θ,η)=(1-2η ² ) sin ² Θ+η ² (3)

This equation is known as Malus' law, which indicates the intensity of light passing through the analyzer 17. Therefore, in the above-mentioned pre-calibration (i), when the luminance value is recorded in the image acquisition unit 19 while shaking the angle Θ _a of the analyzer 17, equation (4) below and in FIG. 16 is obtained.

I(Θ,η)=a[(1-2η ² ) sin ² (Θ _a -Θ)+η ² ] (4)

Here, I, a, η ² , and Θ have different values for each position (x, y) within the field of view. That is, I(x, y). Hereinafter, the notation of (x, y) is omitted. Also, it is expressed as I(Θ,η). a is a coefficient for the total energy of 1, and corresponds to so-called luminance unevenness. η ² /(1-2η ² ), which is the ratio between the amplitude 1-2η ² of sin ² and the DC component η ² , corresponds to the so-called extinction ratio. (i) In preliminary calibration, according to the inventor's verification, this experimental curve is different for each ROI in the field of view. In this way, a(x, y), η ² (x, y), and Θ(x, y) are obtained as device constants from fitting according to equation (4) for each ROI. In tests conducted by the inventor, the gap in the field of view a(x, y) is approximately ±10%, η ² (x, y) is approximately ±20%, and Θ(x, y) is approximately ±0.5° (see Fig. 13) ). Additionally, the absolute values of Θ(x, y) and η ² (x, y) are quantities that are more than one order of magnitude larger than the Kerr rotation and Kerr ellipticalization that occur in the magnetic thin film of the measurement object.

Next, we will consider adding minute Kerr rotation and Kerr ellipticalization in the sample 20 containing a magnetic material such as a magnetic thin film when an external magnetic field is swept. FIG. 22 is a diagram illustrating the rotation and ellipticalization of the polarization plane when an external magnetic field is applied in the polarizing microscope device 1 according to Embodiment 1. As shown in Figure 22, the Kerr rotation is modulated by ±ΔΘ with Θ as the center, and the ellipticity is modulated by ±Δη with η as the center. In all, it is sufficient to think of the ± saturation magnetic field in the sweeping of the external magnetic field.

The device constants are considered the same as in the case of (i) pre-calibration and are expressed in equation (5) below and in Fig. 16.

I±=a[(1-2(η±Δη) ² ) sin ² (Θ _a -(Θ±ΔΘ))+(η±Δη) ² ]

(5)

Since ΔΘ<<Θ, Δη<<η, approximation and equation transformation are performed, and for simplicity, it is written as Θ _a -Θ=Θ and a is omitted. Then, equation (6) below and in Figure 16 becomes.

I±≒[(1-2(η ² ±2ηΔη) ² ) sin ² (Θ±ΔΘ)]+η ² ±2ηΔη]

(6)

Since the indicator for quantitative evaluation is ΔI=I ₊ -I _- , it is expressed as ΔI from equation (6) and transformed.

ΔI=(1-2η ² +4ηΔη) sin(2Θ) sin(2ΔΘ)+4ηΔη (1+cos(2(Θ+ΔΘ))) (7)

From this equation, ΔΘ and Δη have device constants as coefficients and cannot be separated. In addition, since there are two undetermined coefficients, ΔΘ and Δη, it can be seen that it is impossible to derive each from one measurement.

Therefore, as the easiest means of changing the state, the angle β of the analyzer 17 is measured twice at two levels (Θ ₁ = β ₁ -Θ, Θ ₂ = β ₂ -Θ). However, up to equation (7), although it is a dimensionless light intensity, what is actually recorded is the pixel value (luminance value: gray level) of the imaging device of the image acquisition unit 19. Therefore, a dimensionless contrast value is introduced by dividing equation (7) by equation (4). At this time, a in equation (4) also occurs in equation (7), so it is canceled out. That is, contrast C is defined by equations (8-1) to (8-4) below and equation (8) in FIG. 16.

C=(Q1+Q2)/Q3 (8-1)

Q1=(1-2η ² +4ηΔη) sin(2Θ) sin(2ΔΘ)) (8-2)

Q2=4ηΔη(1+cos(2(Θ+ΔΘ))) (8-3)

Q3=[(1-2η ² ) sin ² (Θ)+η ² ] (8-4)

In actual measurement data, this corresponds to the height difference (luminance difference)/average value of the hysteresis loop when sweeping an external magnetic field. Contrast C is the contrast in FIG. 17. By using the contrast derived from the hysteresis loop at the second level of the analyzer 17 as contrast C ₁ and C ₂ and eliminating Δη from the two simultaneous equations of equation (8) to derive ΔΘ, the following (9- Equations 1) to (9-4) and (9) in FIG. 16 are used.

U1=U2 (9-1)

U1=(1-2η ² )(C ₂ sin ² Θ ₂ -C ₁ sin ² Θ ₁ )+η ² (C ₂ -C ₁ ) (9-2)

U2=(1-2η ² )[sin(2Θ ₂ ) sin(2ΔΘ)-sin(2Θ ₁ ) sin(2ΔΘ)·U3]

(9-3)

U3=(1+cos(2Θ ₂ ) cos(2ΔΘ))/(1+cos(2Θ ₁ ) cos(2ΔΘ))

(9-4)

Here, since ΔΘ is a sufficiently small quantity, the equation is modified by approximating sin(2ΔΘ)≒ΔΘ, cos(2ΔΘ)≒1. By doing so, equations (1) and (2) in Figure 16 are obtained.

In addition, Δη(x, y) is also derived as equations (10-1) to (10-4) below and equation (10) in FIG. 16.

2Δη(x, y)=(V1-V2)/V3 (10-1)

V1=C ₁ {(1-2η ² (x, y)) sin ² (β ₁ -Θ(x, y))+η ² (x, y)} (10-2)

V2=(1-2η ² (x, y)) sin(2(β ₁ -Θ(x, y))×2ΔΘ(x, y) (10-3)

V3=2η(x, y){1+cos(2(β ₁ -Θ(x, y)))} (10-4)

However, the sign of η(x, y) in the denominator needs to be identified. This requires separate measurement using a λ/4 wave plate. In addition, the above-mentioned equations and the expressions in FIG. 16 can be appropriately changed without departing from the spirit. For example, if you are expressing an essential physical phenomenon, you may use an approximate notation such as sinΘ≒Θ. Specifically, the above description and the series of equations shown in FIG. 16 are representations that do not make approximations as much as possible, but the essential physical notation does not change even if approximate representations according to the order of actual measured values and set values entered into each are used.

According to this embodiment, it is possible to correct the disparity within the field of view and measure the distribution of polarization characteristics within the field of view with an accuracy equivalent to spot measurement. For example, Patent Document 2 describes a polarizing microscope that analyzes polarization characteristics such as Kerr rotation angle, but does not describe correcting differences in the field of view. The polarizing microscope device 1 of this embodiment can realize measurement with an accuracy equivalent to spot measurement (Φ is several tens of μm) by surface measurement (Φ is several thousand μm) through measurement and correction, thereby reducing the processing time. It can be significantly shortened.

<Example 1>

Next, the results of actually acquiring device constants through pre-calibration in the method of the above embodiment, image acquisition by magnetic field sweeping, and analysis for each ROI will be explained. That is, the angles of the two polarization directions of the analyzer 17 ₁ and The magnetization contrast C was evaluated by imaging MOKE measured in ₂ .

In addition, the rotation angle of the cursor rotation within the field of view was measured in advance by spot measurement, and this was compared as the correct answer data. The sample 20 has a free layer of CoFeB, which is commonly used in MRAM. Specifically, the sample 20 includes a free layer of CoFeB, a layer for generating perpendicular magnetic anisotropy disposed above and below the free layer, and a metal layer laminated as a cap layer on the outermost layer. there is.

The external magnetic field was set to a level that sufficiently reaches saturation magnetization by taking more than 10 times the coercive force H _c . In addition, the ROI coordinates within the field of view of imaging MOKE were the same as the measurement spot coordinates of spot measurement. In addition, the magnetic field sweeping speed (unit: Oe/sec) of spot measurement and imaging MOKE is also common.

Figure 23(a) shows the gray level difference range of the hysteresis loop of each ROI when the angle of the polarization direction of the analyzer 17 is set to +4°, normalized to the gray level difference range centered on the field of view. Figure 23(a) is uncorrected raw data. In this case, there was a spatial gap of 9.7% in σ/AVE.

In Figure 23(b), the contrast of the hysteresis loop is simply divided by the average luminance value of each ROI and normalized based on the value at the center of the field of view. In this case, the spatial gap is 11.4%, but there is a large gap between the raw data in (a) and the distribution of correct answers in (d) by spot measurement.

FIG. 23(c) is a diagram illustrating the contrast distribution of the sample 20 acquired by the polarizing microscope device 1 according to Embodiment 1. As shown in FIG. 23(c), the distribution of contrast in this embodiment is the result of two measurements in which the angle of the polarization direction of the analyzer 17 was set to β ₁ = +4° and β ₂ = -4°. It is derived from the results. In this case, the spatial gap is 3.2%.

It can be seen that the characteristic shape (Fingerprint) that is thought to be due to device constants as shown in Figures 23(a) and (b), but not in the correct answer distribution in (d), has been successfully removed.

Figure 23(d) is a diagram illustrating the distribution of contrast acquired by the spot meter.

As shown in Fig. 23(d), in the results of spot measurement, the gap is reduced and there is no characteristic pattern that is believed to be due to device constants.

24(a) to 24(c) are diagrams illustrating the results of quantitative analysis of the polarizing microscope device 1 according to Embodiment 1, and (d) to 24(f) are diagrams illustrating the results of spot measurement. am. Figures 24(a) to 24(c) show the distribution of saturation magnetization M _s (∝Δ2Θ), coercive force H _c , and slope α within the field of view. Figures 24(d) to 24(f) show the results of spot measurement at the same locations as (a) to (c).

As shown in FIGS. 24(a) to 24(c), this embodiment can quantitatively analyze all the indicators of the hysteresis loop shown in FIG. 18. In addition, although the distribution values of each indicator in this embodiment are multiplied by scaling, there is a correlation with the distribution by spot measurement. In the field of semiconductor manufacturing, coefficient management is possible, so there is no problem in practical terms.

As shown in FIG. 24(b), the coercive force H _c does not include scaling of the magnetization direction, which is a device constant. Therefore, correction is unnecessary. As shown in Fig. 24(c), the slope α can be corrected by multiplying the scaling coefficient of the magnetization direction obtained with the saturation magnetization M _s by R in the empirical model equation (0) in Fig. 16. Specifically, ΔΘ(x, y) is normalized to the center of the field of view to calculate the relative distribution ΔΘ'(x, y) (step S32 in FIG. 9). Next, the uncorrected range R(x, y) is normalized to the center of the field of view to calculate the relative distribution (step S33 in FIG. 9). Then, the Y-axis correction coefficient of the slope is calculated for each ROI. The slope distribution is corrected by ΔΘ'(x, y)/R(x, y) Slope(x, y) (step S34 in FIG. 9).

Next, the effects of this embodiment will be explained. The polarizing microscope device 1 of this embodiment can measure the polarization characteristic distribution within the field of view with an accuracy equivalent to spot measurement, and can realize quantitative surface measurement.

FIG. 25 is a diagram illustrating the effect of the in-field correction analysis method according to Embodiment 1. As shown in FIG. 25, the in-field correction analysis method of the present embodiment sets the angle β of the polarization direction of the analyzer 17 to, for example, -4° and +4° with respect to the sample 20. Acquire a gray level difference range image (raw data) when sweeping an external magnetic field. The raw data in this case has gaps of about σ = 8.6% and σ = 10%, respectively. Contrasts C ₁ (x, y) and C ₂ (x, y) are obtained by normalizing each data, but each contrast contains errors due to the device. These C ₁ and C ₂ are correction results based on the so-called conventional method of standardization based on luminance.

In this embodiment, such device-induced errors are corrected using device constants acquired in advance. If so, the gap can be reduced to σ = 3, about 2% (3 in Figure 25). The characteristic slope distribution due to device constants has also been removed. This allows access to the results of spot measurements. In addition, the polarizing microscope device 1 of the present embodiment enables mass measurement in a short time and is effective when managing the magnetization characteristics of the magnetic film of the MRAM.

Like the magnetic domain polarization microscope of the comparative example, in a magnetic domain observation microscope that uses a polarization microscope as its base, the function remains in the observation of magnetic domains (magnetic domains). On the other hand, according to this embodiment, the hysteresis loop within the field of view can be quantitatively evaluated.

In addition, while the Kerr effect measuring device and spot measuring device of the comparative example are point measurements, the present embodiment is surface measurements. Additionally, this embodiment can quantitatively analyze the hysteresis loop for each ROI in the field of view. In addition, from the shape of the hysteresis loop, performance indicators directly related to the final performance of the MRAM device, such as saturation magnetization M _s , coercive force H _c , and slope α, can be quantitatively evaluated in terms of surface. Therefore, since the evaluation score can be dramatically increased, yield management can be performed using volume data (Volume date).

The present invention is not limited to the above-mentioned Embodiment 1 and Examples, and may be appropriately modified without departing from the spirit. For example, the components of Embodiment 1 and Examples can be combined with each other.

Additionally, the image processing unit 30 described above may be an information processing device such as a personal computer, for example. In addition, the image processing unit 30 is not limited to a personal computer, and may be a server, tablet, mobile terminal, etc., or may be on the cloud as long as it processes information. The image processing unit 30 may include a processor, memory and storage device, and communication device as components not shown. Additionally, the storage device may store the processing performed by the image processing unit 30 as a program. Additionally, the processor may read a program from a storage device into memory and execute the program. Thereby, the processor realizes the functions of the image processing unit 30.

The image processing unit 30 may be realized with dedicated hardware. In addition, part or all of the image processing unit 30 may be realized by a general-purpose or dedicated circuit, a processor, or the like, or a combination thereof. These may be comprised of a single chip, or may be comprised of a plurality of chips connected across a bus. Part or all of the image processing unit 30 may be realized by a combination of the above-described circuits, etc., and programs. Additionally, as a processor, a CPU (Central Processing Unit), GPU (Graphics Processing Unit), FPGA (field-programmable gate array), quantum processor (quantum computer control chip), etc. can be used.

Additionally, when part or all of the image processing unit 30 is realized by a plurality of information processing devices or circuits, the plurality of information processing devices or circuits may be centrally arranged or dispersedly arranged. For example, information processing devices, circuits, etc. may be implemented as a client server system, cloud computing system, etc., each connected through a communication network. Additionally, the functions of the image processing unit 30 may be provided in SaaS (Software as a Service) format.

In addition, the following within-field-of-view correction analysis program that reads and executes the above-described within-field-of-view correction analysis method into a computer is also within the scope of the technical idea of the embodiment. The in-field correction analysis program may be stored in a non-transitory computer-readable medium or a tangible storage medium.

By way of example, and not limitation, computer readable media or tangible storage media may include random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technology, CD-ROM, Includes ROM, digital versatile disc (DVD), Blu-ray (registered trademark) disk or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices. The in-field correction analysis program may be transmitted on a temporary computer-readable medium or communication medium. By way of example, and not limitation, transitory computer-readable or communication media includes signals in an electrical, optical, acoustic, or other form.

(Appendix 1)

A light source that generates illumination light,

a polarizer through which the illumination light generated by the light source is incident and transmits the illumination light including linearly polarized light in a first polarization direction;

An objective lens that illuminates a sample with the illumination light including the linearly polarized light and transmits reflected light reflected from the sample by the illumination light;

an analyzer that transmits a component of linearly polarized light in a second polarization direction in the reflected light;

an image acquisition unit that acquires an image of the reflected light;

A magnet that generates an external magnetic field applied to the sample,

an image processing unit that processes the acquired image;

An in-field correction analysis program using a polarizing microscope device equipped with,

The non-magnetic sample or the sample containing a magnetic material is considered to be a non-magnetic mirror surface by using the non-magnetic sample or the sample containing a magnetic material without applying a magnetic field while rotating the angle formed by the first polarization direction and the second polarization direction at predetermined intervals within a predetermined range. For a sample that can be used, from the plurality of images acquired by illuminating the polarized illumination light, the distribution of the polarization rotation angle for each region of interest in the field of view including the plurality of regions of interest, and the square distribution of the ellipticity by ellipticalization are obtained. A first step of calculating a device constant including;

When the angle formed by the first polarization direction and the second polarization direction is set to the first angle, the polarized illumination light is illuminated on the magnetic material portion of the sample including the magnetic material, and the external Obtaining a hysteresis loop of luminance values for each region of interest from a plurality of images acquired while sweeping a magnetic field,

When the angle formed by the first polarization direction and the second polarization direction is set to a second angle, the polarized illumination light is illuminated on the magnetic material portion of the sample including the magnetic material and the external magnetic field is simultaneously illuminated. a second step of acquiring a hysteresis loop of the luminance value for each region of interest from a plurality of images acquired while sweeping;

A third step of calculating a rotation angle of Kerr rotation for each region of interest from analysis using the device constant, the hysteresis loop at the first angle, and the hysteresis loop at the second angle,

An in-field correction analysis program that runs on a computer.

(Appendix 2)

In the first step,

The analyzer rotates the predetermined range based on an orthogonal Nicol arrangement with respect to the polarizer.

In-field correction analysis program described in Appendix 1.

(Appendix 3)

In the first step,

Calculating the device constant including the angle distribution and the square distribution for each region of interest using Malus' law,

In-field correction analysis program described in Appendix 1 or 2.

(Appendix 4)

In the first step,

The device constants include a luminance distribution,

calculating the device constant including the luminance distribution for each region of interest,

The in-field correction analysis program described in any one of Appendices 1 to 3.

(Appendix 5)

In the second step,

Calculating the contrast, coercivity, and slope for each region of interest by fitting the hysteresis loop at the first angle and the hysteresis loop at the second angle to an empirical approximation function,

The in-field correction analysis program described in any one of Appendices 1 to 4.

(Appendix 6)

In the third step,

Calculating the rotation angle for each region of interest based on the first angle, the second angle, the luminance contrast at the first angle, and the luminance contrast at the second angle,

In-field correction analysis program described in Appendix 5.

Above, the present disclosure has been described in detail using exemplary embodiments, but the present disclosure is not limited to the above embodiments, and various modifications may be made by those skilled in the art within the technical spirit and scope of the present disclosure. and change is possible.

1, 101 polarizing microscope device, 10 light source, 11 lens, 12 polarizer, 13 beam splitter, 14 objective lens, 15 sample stage, 16 magnet, 17 analyzer, 18 imaging lens, 19 image acquisition unit, 20 sample, 21 illumination light, 22 reflected light, 23 illumination pupil, 24 pupil plane, 25 measurement plane, 26 imaging plane, 30 image processing unit

Claims (10)

A light source that generates illumination light;
a polarizer through which the illumination light generated by the light source is incident and transmits the illumination light including linearly polarized light in a first polarization direction;
an analyzer that transmits a component of linearly polarized light in a second polarization direction of the reflected light reflected from the sample;
an image acquisition unit that acquires an image of the reflected light; and
An image processing unit that processes the acquired image,
The image processing unit,
Calculate the device constant,
Obtaining a plurality of hysteresis loops for each region of interest, and
A polarizing microscope device, wherein the rotation angle of Kerr rotation for each region of interest is calculated using the device constant and the plurality of hysteresis loops. According to claim 1,
A polarizing microscope device, wherein the device constant includes a polarization rotation angle distribution for each region of interest in a field of view including a plurality of regions of interest from the plurality of images, and a square distribution of ellipticity due to ellipticalization. According to claim 1,
The device constant further includes a luminance distribution,
A polarizing microscope device, wherein the image processing unit calculates the device constant including the luminance distribution for each region of interest. According to claim 1,
A polarizing microscope device, characterized in that the analyzer rotates a predetermined range based on an orthogonal Nicol arrangement with respect to the polarizer. A light source that generates illumination light;
a polarizer through which the illumination light generated by the light source is incident and transmits the illumination light including linearly polarized light in a first polarization direction;
an analyzer that transmits a component of linearly polarized light in a second polarization direction of the reflected light reflected from the sample;
an image acquisition unit that acquires an image of the reflected light;
A magnet that generates an external magnetic field applied to the sample;
An image processing unit that processes the acquired image,
The image processing unit,
Polarized light is applied to a sample that can be regarded as a non-magnetic mirror surface by rotating the angle formed by the first polarization direction and the second polarization direction at predetermined intervals within a predetermined range and using the sample in a state without applying a magnetic field. From the plurality of images acquired by illuminating the illumination light, calculate device constants including a polarization rotation angle distribution for each region of interest in a field of view including a plurality of regions of interest, and a square distribution of ellipticity due to ellipticalization,
When the angle formed by the first polarization direction and the second polarization direction is set as a first angle, obtain a hysteresis loop of the luminance value for each region of interest,
When the angle formed by the first polarization direction and the second polarization direction is set to a second angle, obtain a hysteresis loop of the luminance value for each region of interest,
A polarizing microscope device, wherein a rotation angle of Kerr rotation is calculated for each region of interest from analysis using the device constant, the hysteresis loop at the first angle, and the hysteresis loop at the second angle. According to clause 5,
The image processing unit, without applying an external magnetic field,
A polarizing microscope device, characterized in that it calculates through angle of analyzer data transmitted through the analyzer for each region of interest in a field of view including a plurality of regions of interest. According to clause 5,
The image processing unit calculates contrast, coercivity, and slope for each region of interest using the hysteresis loop at the first angle and the hysteresis loop at the second angle. A light source that generates illumination light,
A polarizer through which the illumination light generated from the light source is incident and transmits the illumination light including linearly polarized light in a first polarization direction;
An objective lens that illuminates the sample with the illumination light including the linearly polarized light and transmits reflected light reflected from the sample by the illumination light;
An analyzer that transmits a component of linearly polarized light in a second polarization direction in the reflected light,
an image acquisition unit that acquires an image of the reflected light;
A magnet that generates an external magnetic field applied to the sample, and
An in-field correction analysis method using a polarizing microscope device including an image processing unit that processes the acquired image, comprising:
The non-magnetic sample or the sample containing a magnetic material is considered to be a non-magnetic mirror surface by using the non-magnetic sample or the sample containing a magnetic material without applying a magnetic field while rotating the angle formed by the first polarization direction and the second polarization direction at predetermined intervals within a predetermined range. From the plurality of images acquired by illuminating the polarized illumination light for a sample that can be used, a polarization rotation angle distribution for each region of interest in a field of view including a plurality of regions of interest, and a square distribution of ellipticity due to ellipticization are included. A first step of calculating device constants;
When the angle formed by the first polarization direction and the second polarization direction is set to the first angle, the polarized illumination light is illuminated on the magnetic material portion of the sample including the magnetic material, and the external Obtaining a hysteresis loop of luminance values for each region of interest from a plurality of images acquired while sweeping a magnetic field,
When the angle formed by the first polarization direction and the second polarization direction is set to a second angle, the polarized illumination light is illuminated on the magnetic material portion of the sample including the magnetic material, and the external a second step of acquiring a hysteresis loop of the luminance value for each region of interest from a plurality of images acquired while sweeping a magnetic field; and
A third step of calculating a rotation angle of rotation for each region of interest from analysis using the device constant, the hysteresis loop at the first angle, and the hysteresis loop at the second angle. How to interpret my calibration. According to clause 8,
In the second step,
An in-field correction analysis method wherein the image processing unit calculates a group of magnetic field sweeping data for each region of interest in a field of view including a plurality of regions of interest. According to clause 8,
In the second step,
The image processing unit calculates the contrast, coercivity, and slope for each region of interest by fitting the hysteresis loop at the first angle and the hysteresis loop at the second angle to an empirical approximation function. Characterized in-field correction analysis method.

Images