JPS63191004A - Fine dimension measurement - Google Patents

Abstract

PURPOSE:To measure small dimensions at a high accuracy with a high stability, by adjusting the degree of overlap of two light beams using a laser to analyze a reflected light intensity pattern. CONSTITUTION:Two light beams 15 having a Gauss's intensity distribution, separated at a close distance from each other, are condensed to a fine spot diameter, and with it the surface of an object to be measured is irradiated and scanned. Here, when there is a difference in the reflection state between a base material section 31 and a dimensions measuring section 30 of an object being measured, a reflected light intensity pattern is obtained in the integrated form of the Gauss's distribution corresponding to the difference while the intensity level of the reflected light intensity pattern changes relying on dimensions of the object being measured. Thus, dimensions can be determined by analyzing the reflected light intensity pattern.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for measuring minute dimensions in the submicron region.

[Background of the invention]

With recent advances in precision processing technology, workpieces have become increasingly micro-fabricated, and in fields such as integrated circuits and magnetic heads, the dimensions of micro-fabricated products have reached the order of 1 micron meter (μm) or less. There is an increasing need for high-precision measurement of the following minute dimensions.

[Conventional technology]

A commonly used method is to illuminate a minute portion to be measured, magnify it to a high magnification using a microscope, and receive the magnified image with a TV camera to measure the size. This process binarizes the video signal of the received image emitted by the TV camera at a preset slice level value, and the pixels included between the falling edge and vertical edge of the binarized signal. The dimensions are determined by counting the pitch.

[Problem that the invention seeks to solve]

In the above-mentioned measurement method using a TV camera, in order to perform stable binarization processing, it is necessary to obtain a video signal with a good S/N ratio of brightness and darkness and to have a stable slice level value.

If the optical contrast ratio between the edge of the convex or concave part of the part where the dimensions of the object to be measured are to be measured and the base material is poor, the slice level value for binarization becomes unstable and the measurement accuracy decreases. do.

In addition, as the dimensions of the object to be measured become smaller than the spot diameter of the illumination light, the amount of reflected light reflected from the small dimension part decreases, so a single slice level value is set for binarization processing. If so, the actual dimensions and the binarized dimensions no longer correspond (and therefore highly accurate dimension measurements cannot be performed).

SUMMARY OF THE INVENTION An object of the present invention is to provide a dimension measurement method that eliminates the problems of the conventional measurement methods described above and is capable of measuring dimensions smaller than 1 micrometer with high accuracy and stability.

[Means for Solving the Problems] In order to achieve the above object, the present invention comprises the following method.

Two laser beams emitted from a laser light source are placed close together.
The object whose dimensions are to be measured by separating the light beams into two light beams, setting the distance between the intensity peaks of each of the two beams to a predetermined distance, and converging the light into a minute spot diameter with an objective lens. irradiate the surface of the object to be measured, scan the surface of the object with the condensed two-beam light, detect the intensity of the reflected light at the DC level for each scanning state, and detect the reflected light obtained by scanning. The dimensions are measured from the intensity pattern and the intensity level value at a predetermined location of the reflected light intensity pattern.

[Effect]

When measuring dimensions using the above method, two beams of light separated at a close distance and having a Gaussian intensity distribution are focused on a minute spot diameter and irradiated onto the surface of the object to be measured. If there is a difference in reflection state between the base material part of the object to be measured and the dimension measurement part, a Gaussian distribution according to the difference can be integrated to obtain a reflected light intensity pattern, and Since the intensity level of the reflected light intensity pattern changes depending on the dimensions of the reflected light intensity pattern, the dimensions can be determined by analyzing the reflected light intensity pattern.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

Figure 1 is a diagram showing the light intensity distribution of the laser beam used for dimension measurement of the present invention. Figure 1 (a) is a light intensity distribution diagram of a single beam, and Figure 1 (b) is a diagram of the light intensity distribution of a two-beam light. It is an intensity distribution map.

Generally, the light intensity distribution of laser light emitted from a laser light source has a Gaussian distribution.

In FIG. 1(A), the laser light intensity at the maximum intensity point 11 of the Gaussian intensity distribution laser light 10 is I. When closed, the range of the intensity point 12 at 13.5% of the value of Io is defined as the diameter of the Gaussian distributed laser beam.

When the origin is the maximum intensity point 11, which is the center point of the Gaussian intensity distribution laser beam 10, and the radius of the beam diameter is W, the laser beam intensity at a distance r from the origin is I = I. exp (-2r/W) −
It is represented by ■.

FIG. 1(B) is a light intensity distribution diagram of two beams of light that are a combination of two single Gaussian distributed lights shown in FIG. When the beam diameters of the two laser beams 14 are each D, the dimensions of the object to be measured are the peak-to-peak distance CW between the first maximum intensity point 160 and the second maximum intensity point 1400 of each beam. Set the value accordingly. In Figure 1 (b), C
W=0.8D. Setting of this desirable inter-peak distance CW will be described later.

The two-beam light 15 set in this manner is irradiated as a probe light for dimension measurement onto the surface of the object whose dimension is to be measured.

FIG. 2 shows an embodiment of the method for generating the two-beam light 15 described in FIG. 1(b).

20 is an acousto-optic element (hereinafter abbreviated as AO), 21 is AO
The incident laser beam is incident on 20, 22 and 26 are A02
24 is an objective lens, 25 is an AC oscillator, and 26 is an acousto-optic element driver (AO driver).

27 is a DC voltage source for controlling the optical deflection operation of the AO 20. A020 modulates light through the interaction of light and ultrasonic waves, and can perform operations such as light deflection, light intensity modulation, and light frequency shift.

The AC oscillator 25 applies an AC signal having a frequency fm to the AO driver 26 . AO driver 26
Although various configuration examples can be considered, for example, if a high-frequency signal generation source with a frequency of f is built-in and the signals of two frequency components m and fa are AM-modulated, f a +
An AO drive signal having two frequency components f m can be created by the AO driver 26 .

When the signal of these two frequency components is converted into an ultrasonic wave by an ultrasonic transducer and propagated inside A020, the compression wave of the ultrasonic traveling wave is created by the refractive index of the medium due to the photoelastic effect of the medium constituting A020. It causes periodic fluctuations in the amount of light and acts as a diffraction grating for light.

When the incident laser beam 21 enters the AO 20 at a black angle θd, it is diffracted into +1st-order diffracted light 22 among the 1st-order diffracted lights.
A beam light 26 which is the same (-1st order diffracted light) as the beam light 22 is generated.

At this time, the two beams of light 22 and 26 undergo an optical frequency shift, and the frequency of the incident laser light 210 is f.

Then, the light wave of the beam light 22 is f o + f −+
f m.

The light wave of the beam light 26 has a frequency of fo-1-fa-frn.

Further, the angle θm between the two beams 22 and 26 is given by θm=λ−2fm/V, (2). Here, λ is the optical wavelength, V is the speed of the ultrasonic traveling wave propagating in the AO medium, and fm is the frequency of the above-mentioned alternating current.

Therefore, the angle θm between the two beams 22 and 26 is f
It changes linearly depending on the frequency m.

Regarding the operation of the AO20 mentioned above, see Optronics Magazine 1.
It is detailed in the August 1982 issue.

An objective lens 24 focuses the two beams 22 and 26 into a minute spot diameter.

At this time, the peak-to-peak distance CW between the two beams can be determined by the collection distance of the objective lens 24 and the above-mentioned angle θm, and once the optical system is decided, the AC oscillator 2
It can be easily varied with 50 frequency control.

Note that the above-mentioned two-beam light 22.
26 and detecting the alternating current component of the interference signal results in optical heterodyne interference.

FIG. 3 shows a state diagram when the two-beam light is irradiated onto the surface of the object whose dimensions are to be measured and the two-beam light is scanned.

Reference numeral 60 denotes a dimension measuring section whose dimension is to be measured, and its dimension is designated as G. 61 is a base material portion. For example, in the case of a boron head, the dimension measuring section 60 corresponds to the gap section and the base material section 61 corresponds to the track section.

The minute dimension measurement method of the present invention includes a dimension measurement section 600 and a base material section 6.
1 is within the fish collection depth range of the irradiated 2-beam light 15, and there is a difference in reflectance between the dimension measurement section 60 and the base material section 610, and the dimension measurement section 600 is smaller than the beam diameter of the 2-beam light 15. This is particularly effective when the dimension G is small.

Figure 3(b) shows the state where the two-beam light 15 is on the left side of the dimension measuring section 60 and before scanning starts, and Figure 3(b) shows the state where the two-beam light 15 is on the left side of the dimension measuring section 60.
is applied to the dimension measuring unit 60, FIG. 3(C) shows the case where the two-beam light 15 is on the right side of the dimension measuring unit 60 and the scanning is completed, and the two-beam light 15 changes from (A) to (B). →(c) is scanned.

Note that scanning with two beams of light can be realized by various methods. If the optical deflection of A020 explained in FIG. 2 is used, scanning of the two-beam light 15 can be easily performed by electrical control.

The frequency f of the high frequency signal is determined by the AO driver 26.

When emitting , if the frequency f is made variable, it is possible to deflect the two beam lights 22 and 26 according to the frequency band.

Alternatively, the scanning may be performed by driving an electromagnetic mirror (or the object to be measured may be placed on a moving stage and moved by the moving stage).

In this way, the reflected light when the two-beam light 15 scans the object to be measured is detected.

The optical scanning of the two-beam light 15 can be performed stepwise or continuously, but the reflected light is detected in each scanning state and the steps shown in Fig. 3 (A) → (B) → (C) are performed. Capture the data of the reflected light intensity pattern within the scanning range.

Note that the reflected light is detected by detecting the intensity at the DC level. 2
The amplitude of each laser beam 16.14 (shown in FIG. 1) constituting the beam light is set to A1. When A2, the intensities are A, , A, respectively, so the DC level intensities of the two beam lights are A, +A, and so on.

When scanning with two beams of light, the dimension measuring section 60 and the base material section 6
Since there is a difference in the intensity of reflected light between 1 and 1, the dimension G of the dimension measurement unit 600 is measured by analyzing the reflected light intensity pattern in the form of an integrated Gaussian distribution.

FIG. 4 shows a first calculation example of the reflected light intensity pattern when scanning the two-beam light 15 shown in FIG. 3. The amplitude of the two beams was set to A, =A.

The horizontal axis in FIG. 4 is the scanning amount, which is the distance scanned by the two beams, and the vertical axis is the reflected light intensity ratio (to).

The states shown in FIGS. 3(a) and 3(c) are assumed to have an intensity ratio of 100%. The reflectance of the dimension measuring section 600 with respect to the base material section 31 is 0.
．． 5. If the diameter of a single Gaussian beam is 1, then 2
The calculation was performed assuming that the distance CW between the peaks of the intensity of the beam light was 0.8D. For 41, the dimension G is Gaussian, 30% of the diameter of the -mm diameter, for 42 it is also 50%, for 46 it is also 70%, and for 4
4 is also 90%, and 45 is 100%.

As is clear from the graph in Fig. 4, the two-beam light 15
Once the beam shape is determined, the reflected light intensity pattern is uniquely determined depending on the size of the dimension G.

When the dimension G of the dimension measurement unit 600 is small, the change in the reflected light intensity is small and shows a W-shaped pattern, and as the dimension G becomes larger, the change in the reflected light intensity becomes thicker (becomes a U-shaped pattern or a V-shaped pattern). shows.

In this way, depending on the size of the dimension G, the intensity level of the reflected light intensity pattern and the minimum point 46 or maximum point 470 of the reflected light intensity changes.

FIG. 5 shows a calculation example of a change in the intensity level of the minimum point or maximum point of the reflected light intensity depending on the dimension G in the case of the calculation example shown in FIG. 4.

The horizontal axis of the graph is the relative value of the dimensions, and the vertical axis is the intensity ratio to the maximum intensity. Here, the intensity ratio of the vertical axis of the graph is 2
The intensity of reflected light when the beam light 15 is on the base material portion 61 is used as a reference.

In FIG. 5, the curve 51 is the intensity ratio at the maximum part, and the curve 520 is the intensity ratio at the minimum part. The curve indicated by the 560 dotted line is a case where the reflected light intensity pattern becomes VW and changes with the minimum portion and maximum portion coincident.

The dimension G can be easily determined by using the minimum/maximum intensity ratio data shown in FIG. 5 as reference data and comparing it with the minimum/maximum data of the measured values.

The results shown in Figures 4 and 5 are specific to certain conditions (CW = 0, 8
This corresponds to the shape of the two-beam light shown in D).

A second calculation example is shown in FIG. 6 for comparison.

FIG. 6 shows a calculation example in which the distance CW between the intensity peaks of the two beams is equal to the diameter of the Gaussian intensity distribution laser beam 10, and the reflectance of the dimension measuring section 300 is 50% of the reflectance of the base material section 61. be. Curve 61 indicates that the width of the dimension measuring section 600 is 30% of the Gaussian beam diameter, curve 62 indicates that the width is 50%,
Curve 66 is 70%, curve 64 is 90%, and curve 65 is 10%.
This is the case of 0%.

Although a W-shaped pattern as shown in FIG. 4 can be obtained in the same way, the intensity ratio of the minimum portion and maximum portion described above is generally different.

Therefore, 2-beam light 150 according to the dimensions to be measured.
Intensity shape, especially Gaussian beam diameter and peak-to-peak distance CW
It is important to set , but for example, if the rate of change of the intensity ratio of the minimum or maximum part of the reflected light intensity pattern shown in Fig. 5 with respect to the dimension G is large, and it changes to IJ = A, The shape of the two-beam light 15 may be determined depending on the state.

Further, the above-mentioned reflected light intensity pattern changes depending on the reflectance of the dimension measuring section 60 with respect to the base material section 61.

Therefore, first, prepare a reference sample that has a reflecting surface larger than the spot diameter of the two-beam light 15 and is made of the same material. If the reflectance of the dimension measurement unit 600 with respect to the base material part 61 is determined in advance by measuring it in advance, the conversion coefficient from the intensity information of the minimum and maximum parts of the reflected light intensity pattern shown in FIG. 5 to dimensions can be calculated. It can be easily determined.

At this time, if the shape of the two-beam light is known, the dimension G can be determined directly, but if the shape of the two-beam light is unknown, measure the dimensions of the reference sample in advance by other means. If the shape of the two-beam light 15 is determined by calculating backward from the dimensional values, the normal measurement as described above can be performed.

The analysis of the reflected light intensity pattern has been described in terms of the intensity level, but it is also possible to analyze the place where the pattern changes, the rate of change (derivative coefficient) of the W-shaped, U-shaped, or V-shaped curve, etc. .

Laser light has extremely high spatial coherency and excellent focusing ability, and can be focused to a Gaussian beam diameter of approximately 1 μm. It is possible to measure dimensions of minute areas down to about 0.1 μm.

〔Effect of the invention〕

As is clear from the above explanation, by using a laser beam with good convergence and a stable light intensity distribution, the degree of superposition of the two beams is adjusted, and the Analysis of the reflected light intensity pattern enables highly stable and highly accurate dimensional measurements.

[Brief explanation of the drawing]

FIGS. 1 to 6 show embodiments of the present invention, and the first
Figure (a) is a Gaussian intensity distribution diagram of laser light, Figure 1 (b)
is an intensity distribution diagram of two-beam light, Fig. 2 is a block diagram explaining the generation of two-beam light using an acousto-optic element, and Fig. 3 (
A), (B), and (C) are explanatory diagrams of the state in which two beams of light are scanned on the surface of the object to be measured, and Fig. 4 shows the first reflected light intensity and turn when two beams of light are scanned. Figure 5 is an intensity ratio-dimension graph that calculates the change in intensity level from the first calculation example shown in Figure 4, Figure 6 is a graph showing a calculation example of 2.
The second reflected light intensity pattern when scanning the beam light
It is a graph showing an example of calculation. 10... Gaussian beam light, 15... 2 beam light, 20... Acousto-optic element, 25... AC oscillator, 60... Dimension measurement section, 61...Base material section. Figure 1 (/i) (b) Distance (r) Figure 3 Figure 5 (

Claims (1)

[Claims] Two laser beams emitted from a laser light source are placed close together.
The object whose dimensions are to be measured by separating the light beams into two light beams, setting the distance between the intensity peaks of each of the two beams to a predetermined distance, and converging the light into a minute spot diameter with an objective lens. irradiate the object to be measured, scan the surface of the object with the condensed two beams, detect the reflected light intensity at the DC level for each scanning state, and scan the two beams. A method for measuring minute dimensions, characterized in that dimensions are measured from a reflected light intensity pattern and an intensity level at a predetermined location of the reflected light intensity pattern.