JP2001027619A - Lattice strain measuring apparatus of local region - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an apparatus for automatically measuring the lattice strain in the local region of a crystal material as two-dimensional distribution. SOLUTION: A lattice strain measuring apparatus of a local region is constituted of a transmission electron microscope 1 capable of observing the converged electron diffraction figure from a crystal material, a CCD camera 3 for automatically taking in the converged electron diffraction figure, an electronic calculator 4 having a program, which recognizes the crossing points of HOLZ lines in the taken-in converged electron diffraction figure and calculated electron beam effective acceleration voltage and lattice strain from the distance between the crossing points of the HOLZ lines, built therein, a sample holder 2 having a sample automatic moving device for moving the sample two-dimensionally and automatically built therein and a display 5 for displaying a measured result.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus and a method for measuring lattice distortion of a crystal material, and more particularly to a method for measuring lattice distortion automatically and two-dimensionally with high spatial resolution using a convergent electron diffraction method. Apparatus and method.

[0002]

2. Description of the Related Art Stress in a crystal material and lattice strain accompanying the stress are one of important physical properties representing the properties of the material.

In general, the stress and lattice strain of a crystal material vary depending on the growth conditions and processing conditions even for the same material. Many of the other physical properties and properties of the crystalline material change with stress and lattice strain. Therefore, the state of stress / lattice strain of the crystalline material is evaluated, and processing and
Optimizing growth conditions is essential for the development of high-performance materials and products.

In particular, in the ULSI, since the structure of the element is fine, it is necessary to locally measure the stress and the lattice strain. Conventionally, local stress / lattice strain has been measured using X-ray diffraction, Raman spectroscopy, or convergent electron diffraction.

[0005]

In recent years, high integration and miniaturization of LSIs have been rapidly progressing, and development of devices in units of 100 nm has been promoted. Along with this, there is an increasing need to locally measure lattice strain, which is one of the physical properties that affect the element characteristics of a device.

However, the conventional methods for measuring stress and lattice strain, such as X-ray diffraction and Raman spectroscopy, have low spatial resolution and cannot perform sufficient measurements.

Therefore, instead of the X-ray diffraction method and the Raman spectroscopy, a lattice distortion measurement method using a convergent electron diffraction method capable of obtaining a spatial resolution of nm (nanometer) has been proposed. For example, JP-A-4-206694, JP-A-7-167719, JP-A-7-286915, JP-A-10-162768, JP-A-10-162
Japanese Patent Publication No. 176809 proposes an apparatus or a method for measuring lattice strain or stress near element isolation using a convergent electron diffraction method.

However, since only one-dimensional measurement is performed in the measurement of lattice strain or stress by the apparatus or method proposed in these publications, the two-dimensionally distributed impurity element and It was not enough to discuss the effect of lattice strain on electrical defects. To clarify the effect of such lattice distortion,
An apparatus capable of two-dimensionally measuring lattice strain is required.

In addition, many measurement points are required for two-dimensionally measuring lattice distortion, and lattice distortion measurement using a convergent electron diffraction method requires many measurement steps. Therefore, a large amount of analysis time is required at each of a large number of measurement points. Therefore, although it is possible in principle to measure the lattice strain distribution two-dimensionally, it is actually extremely difficult. there were.

[0010] The present invention has been made in view of the problems in such a conventional lattice strain measuring apparatus or method, and
It is an object of the present invention to provide a local region lattice distortion measuring device that can easily measure lattice distortion without requiring much analysis time.

[0011]

In order to achieve this object, a first aspect of the present invention is an apparatus for measuring lattice distortion using a HOLZ line in a HOLZ figure obtained when a crystal material is irradiated with convergent electrons. , A local region lattice distortion measuring device is provided, which has a device for recognizing an intersection between HOLZ lines.

[0012] As described in claim 2, it is preferable that the local area lattice distortion measuring apparatus includes an apparatus for measuring a distance between intersections.

Further, as described in claim 3,
It is preferable that the local region lattice strain measuring device has a sample automatic moving device.

[0014] As described in claim 4, the local region lattice strain measuring apparatus preferably has an electron beam effective acceleration voltage measuring means.

[0015] As described in claim 5, the local region lattice distortion measuring apparatus preferably has display means for displaying a two-dimensional distribution of lattice distortion.

A sixth aspect of the present invention provides a transmission electron microscope capable of observing a focused electron diffraction pattern obtained when a crystal material is irradiated with focused electrons, a CCD camera for imaging the focused electron diffraction pattern, and a CCD camera. Local area lattice distortion measurement comprising: means for recognizing an intersection between HOLZ lines in a captured convergent electron diffraction pattern; and means for calculating an electron beam effective acceleration voltage and lattice distortion based on the distance between the intersections. Provide equipment.

This local area lattice distortion measuring apparatus is characterized in that:
It is preferable to further include a sample moving device that moves the sample two-dimensionally as described in (1).

According to another aspect of the present invention, a CCD is provided.
The camera preferably captures the HOLZ figure as two-dimensional numerical data.

According to a ninth aspect of the present invention, a CCD
Preferably, the camera has more than one million pixels.

According to a tenth aspect, a distortion-free reference sample is used.
A first process of generating a HOLZ pattern and measuring the electron beam effective accelerating voltage, a second process of moving the sample to a measurement point for performing strain measurement, and generating a HOLZ pattern,
The third process of taking LZ figures into a CCD camera, and CC
Based on the HOLZ figure taken into the D camera, a fourth process of obtaining lattice distortion in the irradiation area, a fifth process of repeating the second to fourth processes a predetermined number of times, and measurement of lattice distortion at each measurement point And a sixth step of displaying the result as a two-dimensional distribution.

According to the present invention, for example, it is possible to automatically carry out from the taking of the HOLZ figure to the calculation of the lattice distortion by using an electronic computer having a predetermined program. For this reason, analysis time is shorter than the conventional method,
In addition, the lattice distortion can be calculated by a simpler method.

[0022]

FIG. 1 is an explanatory view of the convergent electron diffraction method. An electron beam 10 having a convergence angle of about 10 mrad
, Most of the electrons pass through the sample 11 to form a circular transmission disk 12, but electrons satisfying the diffraction condition are diffracted to the outside of the transmission disk 12 to form a diffracted wave 13. Such diffraction is particularly pronounced in higher Laue bands (H
Light Laue Zone (HOLZ) reflection. With the HOLZ reflection, a dark line called a HOLZ line 14 is generated in the transmission disk 12.

FIG. 1 shows only one diffraction condition, but in reality, a plurality of HOLZ reflections form a ring as shown in FIG. 2C. In FIG. 2, a transmission disk O exists inside a HOLZ reflection C that forms a ring. FIG. 3 is an enlarged view of the transmission disk O.

In response to the plurality of HOLZ reflections shown in FIG. 2C, a plurality of HOLZ lines 1
4 is formed. The HOLZ pattern changes according to changes in diffraction conditions, that is, changes in electron beam energy and lattice spacing. Therefore, by quantifying the change in the HOLZ figure, the lattice distortion of the irradiation area can be known.

One of the features of the HOLZ reflection is that the Bragg angle is large. It is about 10 times larger than the Bragg angle of a diffraction spot appearing in a normal selected area diffraction pattern. This means that HOLZ reflection is about 10 times more sensitive to lattice distortion than diffraction that appears in selected area diffraction.

Here, the change of the Bragg angle θ _β with respect to the lattice strain (the value of θ _β determines the position of the HOLZ line and the diffraction spot) is estimated from the Bragg equation 2dsin θ _β = λ (1). Get.

In Equations (1) and (2), d is the spacing between grating surfaces that cause diffraction, θ _{β is the Bragg angle, Δθ
β} and Δd represent the amounts of change in Bragg angle and lattice spacing, respectively.

[0028] Since [Delta] d / d is equivalent to the lattice strain, which when is constant, Bragg angle theta _beta is greater reflection as Bragg angle variation [Delta] [theta] _beta, that is, a large amount of change in the position of the HOLZ line or diffraction spots You can see that. Thus, it can be seen that HOLZ reflection is more sensitive to lattice distortion than diffraction spots appearing in selected area diffraction.

As described above, the diffraction condition also depends on the energy of the incident electrons, that is, the acceleration voltage. Being sensitive to lattice distortion means being sensitive to the acceleration voltage of incident electrons. Bragg's equation (1) and an equation representing the relationship between the acceleration voltage V _{0 of} incident electrons and the wavelength Estimating the change in Bragg angle with respect to the change in acceleration voltage Get.

In equations (3) and (4), e, m,
h represents an elementary charge, a static mass of an electron, and a Planck constant, respectively.

Equation (4) means that a correct lattice distortion value cannot be obtained unless the acceleration voltage is measured with the same accuracy as the lattice distortion prior to the lattice distortion measurement.

Also, due to the effect of multiwave diffraction, the incident electrons behave as if the energy (accordingly, the accelerating voltage) has changed in the crystal. Therefore, the acceleration voltage to be measured prior to the measurement of the lattice strain must include this apparent change in energy. The accelerating voltage that includes the apparent change in energy (hereinafter “
Depends on the incident direction of the electron beam, but if there is an unstrained reference sample that can take the same incident direction as the incident direction when measuring the lattice strain, HOLZ is determined based on the procedure described below. The effective acceleration voltage can be obtained from the figure.

In order to quantify the amount of change in the HOLZ figure and obtain the effective acceleration voltage and lattice strain, the distance between the intersections of the HOLZ lines is compared between theoretical calculation and actual measurement. FIG. 4 shows a schematic diagram of the procedure.

The equation of the HOLZ line derived from the Bragg equation is It is. Here, x and y are plane coordinates on the fluorescent screen,
g _x , g _y , g _z , and g represent the components and magnitude of the reciprocal lattice vector of the HOLZ line, and K represents the magnitude of the wave vector of the incident electrons.

Equation (5) is not strictly a straight line.
In the OLZ figure, x, y≪1 holds, so
The right side is (−g ² / 2K), which can be approximated as a straight line. Therefore, if the equation of the HOLZ line as a straight line can be obtained in the observed HOLZ figure, the distance between the intersections can be obtained. Actually, coordinates on the HOLZ line are taken out, and a straight line equation is obtained using the least squares method.

Assuming that the equation of the i-th HOLZ line thus obtained is y = a _ix + b _i (6), k formed by the i-th and j-th HOLZ lines
The coordinate (x _k , y _k ) is And also the m formed by the k, lth intersection
_Th inter-intersection distance D _{exp. m} is D _{exp. m} = is determined as ^{_{((x k -x l) 2}} + (y k -y l) 2) 1/2 (8). The distance between the intersections of the HOLZ lines based on the theoretical calculation is calculated using Expression (5).

By comparing the distance between intersections obtained by the observation and calculation obtained in this manner by the least squares method, the electron beam effective acceleration voltage and the lattice distortion can be obtained. That is, the distance D _cal. And the inter-intersection distance D _exp. And the residual sum of squares (The acceleration voltage when measuring the effective acceleration voltage and the respective strain components when measuring the lattice strain) so as to minimize. i is a number for each distance between intersections, N is the total number of distances between intersections used for measurement, σ _i is D _exp. represents the measurement error of _i .

In this analysis method, the accuracy of the required effective acceleration voltage and lattice strain can be improved by considering the distances between a large number of HOLZ line intersections. For example,
When the [230] direction of the single crystal Si is selected as the electron beam incident direction, at an acceleration voltage of about 200 kV, ± 1.8 × 1
It can be obtained 0 ^-4 lattice strain accuracy. Note that this value is a value obtained when measurement is performed using 55 sets of intersection distances by six HOLZ lines.

Effective acceleration voltage V _eff. Is the magnitude of the wavenumber vector There is a relationship.

Further, since the equation of the HOLZ line in the calculation is the equation (5), the change of the effective acceleration voltage is reflected on the equation of the HOLZ line through the change of the magnitude of the wave number vector _. Is reflected in

The lattice strain is synonymous with the change in the lattice constant, and has a relationship expressed by the following equation with the lattice constant.

In Equation 11, vectors a, b, and c are basic propulsion vectors of an unstrained crystal, and have lattice constants a,
b, c, α, β, and γ have the following relationship.

A ', b', and c 'are basic translation vectors of the distorted crystal. Matrix A is a matrix that defines a primitive translation vectors of the crystals of no strain, primitive translation vectors using the matrix A, the unit vector e _x of the coordinate system defining the primitive translation _vectors, e _y, e _z Is defined as Matrix R unit vectors _{e x, e y, e z}
Is a matrix representing the orthogonal coordinates between the coordinate system and the unit vectors f, g, h of the coordinate system that defines the lattice distortion, and R ^-1 represents the inverse matrix.

The coordinate system that defines the lattice strain is a coordinate system that takes into account the shape of the sample, the crystal orientation, and the like, and does not always match the coordinate system that defines the basic translation vector (hence, the lattice constant). An example of a coordinate system that defines the lattice distortion is shown below.

A device structure is manufactured on a Si (001) substrate, a sample having a (110) cross section is prepared from this substrate, the incident direction is set to [110], and lattice distortion is measured. In this case, [001] of the substrate normal, [110] of the surface normal of the cross-sectional sample, and [1 (-1)] perpendicular to both.
It is natural to define the lattice strain by using each crystal orientation of [0] as a coordinate axis. The coordinate system determined in this way is a coordinate system that defines the lattice distortion.

In addition, the matrix E is a matrix having lattice distortion as a sentence. The lattice distortion is reflected in the basic translation vector through Expression (13), and furthermore, the definition expression of the reciprocal lattice vector Through the reciprocal lattice vector. Note that h, k,
1 is the index of the reciprocal lattice vector.

Since the equation of the HOLZ line in the calculation depends on the reciprocal lattice vector in the form of equation (5), the lattice distortion is calculated by changing the reciprocal lattice vector to D _cal. Is reflected in

[0048]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below. FIG. 5 shows a block diagram of the local region lattice distortion measuring apparatus according to the present embodiment. The local area lattice strain measuring apparatus includes a transmission electron microscope 1 capable of forming an electron beam nanoprobe and a sample holder 2 provided with a sample moving apparatus capable of moving a sample in nanometer units.
A CCD camera 3 capable of capturing a HOLZ graphic as two-dimensional numerical data, and an electronic computer 4 having a program for obtaining an effective acceleration voltage and a lattice strain from the captured HOLZ graphic based on the above-described method. Become.

The computer 4 is provided with the sample holder 2 and the CC
It is also connected to the D camera 3 and performs these controls.
The computer 4 also serves as an input terminal for various numerical values to be designated by the user.

The transmission electron microscope 1 can form convergent electrons having a convergence angle of about 10 mrad necessary for forming a HOLZ figure, and can continuously change the convergence angle according to the material or the incident direction. Further, the beam diameter can be converged to 1 nm or less in order to enable measurement in a minute area. In addition, the transmission electron microscope 1 includes an electron gun capable of obtaining a sufficient luminance even when the beam diameter is set to 1 nm.

The sample holder 2 is provided with a goniometer that can be tilted to two-axis directions perpendicular to each other so that the incident direction can be adjusted. The sample moving device mounted in the sample holder 2 is configured to move the sample two-dimensionally in units of 10 nm.

The CCD camera 3 for taking in the HOLZ figure automatically sets an appropriate taking time in accordance with the signal intensity of the HOLZ figure, and takes in the HOLZ figure. Also, CC
As the D camera 3, 100 cameras are used for the following reasons.
Those with more than 10,000 pixels are used.

Here, a CCD of 1024 × 1024 pixels
It is assumed that a HOLZ figure from Si is captured using a camera. Since 25 mrad is a typical value of the scattering angle region of the HOLZ pattern obtained from Si, assuming that 1024 pixels correspond to a scattering angle of 25 mrad, one pixel corresponds to approximately 2.4 × 10 ⁻² mrad. If this value is used as the detection accuracy of the position of the HOLZ line intersection, the corresponding lattice distortion accuracy is 8 × 10 ⁻⁴ . Since the lattice strain exists in the order of 10 ⁻³ in the vicinity of the element separation, a strain accuracy of at least 10 ⁻⁴ is required. Therefore, the CCD camera 3 is required to have a pixel number of 1,000,000 pixels or more.

The electronic computer 4 has a built-in program for automatically calculating the effective electron beam acceleration voltage or the lattice distortion from the HOLZ figure captured by the CCD camera 3 based on the above-described method. This program is HO
It has a function that can automatically calculate the equation of the LZ line.

Further, the computer 4 displays the observed H
A display 5 for displaying the OLZ graphic is provided. The electronic computer 4 also serves as an input terminal for the measurement conditions (the incident direction of the electron beam, the approximate acceleration voltage of the electron beam, and the moving amount of the sample) designated by the user, and the sample holder 2
The CCD camera 3 is controlled based on the measurement conditions specified by the user.

Next, the specific procedure of the local lattice distortion measurement using the local region lattice distortion measurement apparatus will be described. FIG.
Shows the outline of the procedure.

First, the electron beam effective acceleration voltage is measured using a strain-free reference sample as follows.

The user inputs the electron beam incident direction, convergence angle and approximate acceleration voltage via the electronic computer 4 (step 1).

The electron microscope 1 generates a HOLZ figure based on the conditions input in step 1, and the CCD camera 3 takes in the HOLZ figure (step 2).

The fetched HOLZ figure is stored in the computer 4
Is displayed on the display 5 provided in the. The user observes this and measures the electron beam effective acceleration voltage and the index of the HOLZ line used in the subsequent lattice strain measurement (step 3).

The computer 4 uses the built-in electron beam effective acceleration voltage measurement program to calculate the electron beam incident azimuth, the approximate acceleration voltage, and the HOLZ line index input in step 1 and step 3. Then, the effective acceleration voltage of the electron beam is determined from the taken HOLZ figure (step 4).

After obtaining the electron beam effective acceleration voltage as described above, the lattice strain is measured as follows.

The user moves the sample to the first position where the strain measurement is performed, and inputs coordinate axes for specifying lattice strain, parameters for the sample movement (movement direction and distance), and the electron beam effective acceleration voltage measured in advance. (Step 5).

The electron microscope 1 uses the HOL based on the incident azimuth and the convergence angle input when measuring the electron beam effective acceleration voltage.
A Z figure is generated, and the CCD camera 3 captures the Z figure (step 6).

Using the built-in lattice strain measurement program, the computer 4 was loaded based on the parameters input in step 5 and the index of the HOLZ line specified at the time of measuring the electron beam effective acceleration voltage. The lattice distortion of the irradiation area is determined from the HOLZ figure (step 7).

Thereafter, the automatic sample moving apparatus performs step 5
The sample is moved based on the conditions input in (Step 8).

Thereafter, steps 6 and 7 are executed again to measure the lattice distortion.

Thereafter, by repeating Steps 6 to 8, the lattice distortion measurement at each measurement point is automatically performed, and after all the measurements have been performed, the result is displayed as a two-dimensional distribution.

As described above, according to the present embodiment, the input direction of the electron beam, the convergence angle, the approximate acceleration voltage, the coordinate axis for specifying the lattice distortion, and the distance between the measurement points can be obtained simply by inputting the distance.
The local two-dimensional distribution of lattice strain can be easily measured.

[0070]

According to the local region lattice distortion measuring apparatus and method according to the present invention, it is possible to easily measure the two-dimensional distribution of the lattice distortion in the local region. New knowledge can be obtained about the correlation with the characteristics, the correlation between the lattice strain in the wiring material and the electromigration, and the like.

[Brief description of the drawings]

FIG. 1 is a perspective view of an example of a lattice distortion measuring device that performs a convergent electron diffraction method.

FIG. 2 is an electron micrograph of a plurality of HOLZ reflections forming a ring.

FIG. 3 is an enlarged electron micrograph of a transmission disk existing inside the HOLZ reflection shown in FIG. 2;

FIG. 4 is a flowchart illustrating a procedure for obtaining an effective acceleration voltage and a lattice strain from a HOLZ figure.

FIG. 5 is a block diagram of a local region lattice distortion measuring device according to an embodiment of the present invention.

6 is a flowchart showing a procedure of local lattice distortion measurement using the local region lattice distortion measurement device shown in FIG.

[Explanation of symbols]

 1 electron microscope 2 sample holder 3 CCD camera 4 electronic computer 5 display

Claims (10)

[Claims] 1. An apparatus for measuring lattice distortion using a HOLZ line in a HOLZ figure obtained when a crystal material is irradiated with convergent electrons, comprising an apparatus for recognizing an intersection between the HOLZ lines. Local area lattice strain measurement device. 2. The local area lattice distortion measuring apparatus according to claim 1, further comprising an apparatus for measuring a distance between the intersections. 3. The local area lattice distortion measuring apparatus according to claim 1, further comprising an automatic sample moving apparatus. 4. The local region lattice distortion measuring apparatus according to claim 1, further comprising an electron beam effective acceleration voltage measuring unit. 5. The local area lattice distortion measuring apparatus according to claim 1, further comprising a display unit that displays a two-dimensional distribution of lattice distortion. 6. A transmission electron microscope capable of observing a focused electron diffraction pattern obtained when a focused electron is irradiated on a crystal material; a CCD camera for imaging the focused electron diffraction pattern; Means for recognizing an intersection between the HOLZ lines in the obtained convergent electron diffraction pattern, and means for calculating an electron beam effective acceleration voltage and a lattice distortion based on a distance between the intersections. measuring device. 7. The local area lattice distortion measuring apparatus according to claim 6, further comprising a sample moving device for moving the sample two-dimensionally. 8. The local area lattice distortion measuring apparatus according to claim 7, wherein said CCD camera captures a HOLZ figure as two-dimensional numerical data. 9. The local area lattice distortion measuring apparatus according to claim 6, wherein the CCD camera has one million or more pixels. 10. A first process of generating a HOLZ pattern using an unstrained reference sample and measuring an electron beam effective acceleration voltage, and a second process of moving the sample to a measurement point at which strain measurement is performed. A third step of generating a HOLZ figure and capturing the HOLZ figure into a CCD camera; a fourth step of obtaining a lattice distortion of an irradiation area based on the HOLZ figure captured by the CCD camera; A fifth process of repeating the fourth to fourth processes a predetermined number of times; and a sixth process of displaying a measurement result of the lattice strain at each measurement point as a two-dimensional distribution.