JPH10170232A - Method for improving line width measuring accuracy in scanning microscope - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for improving line width measuring accuracy in a scanning type microscope which can accurately measure a short dimension such as a line width from the image data of a scanning microscope. SOLUTION: In improving method for line width measuring accuracy in a scanning microscope which mounts a sample on a stage, and moves the stage 31 at a fixed speed command value 52 to create the image of the sample by the scanning microscope, and measures the line width of the sample from an encoder signal 54 during the movement of a stage and the image data of the scanning type microscope, the pitch of the encoder signal 54 during the movement of the stage is counted by a high frequency standard clock signal 55 to obtain actual moving speed from that counted value, and the said command speed is compared with the actual moving speed to compensate the measured line width.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for measuring a short dimension such as a line width of an image of a mask or the like, and more particularly, to a method for generating a two-dimensional measurement image by a laser scanning confocal microscope. The present invention relates to a method for improving width measurement accuracy.

[0002]

2. Description of the Related Art In a SiSCAN (trade name, laser scanning confocal microscope), in order to generate a two-dimensional measurement image, a device called a scanner is used to rapidly change the relative position between a sample and laser light and scan in the X direction. While writing, the stage on which the sample is placed is scanned at a low speed in the Y direction to write data in the image memory. As a result, an X and Y two-dimensional image is completed.

[0003] Referring to FIG.
The sample on which the image has been formed is placed on a stage (not shown) of a scanning microscope, and the stage is slowly moved in the Y direction while being scanned in the X direction by a spring. , Are detected by the optical system 12. The optical system 12 passes through a semi-reflecting mirror 14 from a laser oscillator 13, is once condensed by a pair of collimator lenses 15, 15, is converted into parallel light, passes through a reflecting mirror 16, a concave lens 17, and is sampled by an objective lens 18. The light is focused upward, and the reflected light passes through the pinhole 19 provided at the focal position of the collimator lenses 15 and 15, is reflected by the semi-reflecting mirror 14, and is incident on the photoelectric tube 20 through the light path opposite to the above-described optical path. You. At this time, the focal point of the objective lens 18 is on the line 11 of the sample 10, and the reflected light of the line 11 is
9, the reflected light other than the line 11 is cut by the pinhole 19, so that the ON / OFF data of the photoelectric tube 20 becomes the light intensity data I for measuring the line width.

[0004] A two-dimensional image can be formed from the light intensity data I and the X and Y positions, and the line width can be measured.

FIG. 7 shows a speed control mechanism of the Y stage 22 on which the sample 10 is mounted.
The stage 23 is moved by the motor 23,
An S (laser measurement system) 24 detects the position and an encoder signal 25 at that position is input to a computer 26.

A computer 26 outputs a speed command value to a servo system 27. The servo system 27 controls the motor rotation speed based on the encoder signal 25 of the LMS 24 and the command value.

The computer 26 measures the line width of the sample 10 based on the light intensity data I obtained from the phototube 20 and the encoder signal 25 obtained from the LMS 24 described with reference to FIG.

As described above, in order to measure a target dimension in an image, data in an image memory is used.
In order to obtain accurate dimensions, scanning in the X and Y directions must be performed accurately and with good reproducibility.

[0009]

The sample 1
0 is scanned in the X direction by the scanner 28 composed of a spring, and the reproducibility is good. However, as described above, the scanning in the Y direction is performed by moving the stage at a low speed.

The operation of the Y stage is performed by two types of speed feedback control using a motor and a tachogen forming the servo system 27 and position feedback control using a laser light intensity change by a laser interferometer forming the LMS 24 as position information. I have. Laser interferometer
The position feedback system outputs a detection signal in units of 10 nm, always counts this encoder signal, and controls the motor drive current so that the deviation from the commanded speed is minimized (ideally, 1).

However, since the motor does not respond sufficiently to the control current, the speed fluctuation of the stage appears as disturbance of the encoder signal of the laser interferometer.

FIG. 8 shows the actual actual speed distribution and the LMS ideal pitch obtained from the command speed (ideal speed) distribution of the stage generated by the computer.
The MS actual pitch is shown. Even if an attempt is made to keep the speed constant in the measurement range, the actual speed fluctuates, which disturbs the LMS actual pitch and affects the measurement accuracy of the measured line width.

The movement of the stage is slow, so that the control accuracy of the motor is limited.
This is the limit of the measurement accuracy in the direction (especially measurement reproducibility).

However, it is almost impossible to demand a motor or a tachogenator with higher performance than the present, and the laser beam of the laser interferometer is also constantly changed due to the flow of air in the place where it passes and temperature changes. It is not constant.

An object of the present invention is to solve the above-mentioned problems and to provide a method for improving the line width measurement accuracy in a scanning microscope capable of accurately measuring a short dimension such as a line width from image data of the scanning microscope. It is in.

[0016]

In order to achieve the above object, according to the first aspect of the present invention, a sample is placed on a stage,
In the scanning microscope, the stage is moved at a predetermined speed command value to generate an image of the sample with the scanning microscope, and the line width of the sample is measured from the encoder signal during the movement of the stage and the image data of the scanning microscope. In the method of improving the line width measurement accuracy, the pitch of the encoder signal during the movement of the stage is counted using a high-frequency reference clock signal, the actual speed is obtained from the count value, and the actual speed is compared with the command speed and measured. This is a method for improving the line width measurement accuracy in a scanning microscope that corrects the line width.

According to a second aspect of the present invention, the encoder signal is reduced to 1 /
The frequency is divided by two, and the rising and falling edges of the 1/2 frequency-divided waveform are counted independently by the reference clock signal, and the odd and even pitches of the encoder signal are counted to obtain pitch data, based on the command speed. 2. The method according to claim 1, wherein the measurement line width is corrected based on a reference speed count value of an ideal speed and the pitch data.

According to a third aspect of the present invention, the sum of the reference speed count value of the ideal speed and the sum of the pitch data are obtained for the range corresponding to the line width obtained from the image data, and the ratio of these sums is calculated as follows. A method for improving line width measurement accuracy in a scanning microscope according to claim 2, wherein the measurement line width is corrected.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A preferred embodiment of the present invention will be described below in detail with reference to the accompanying drawings.

First, the XY stage and the basic circuit will be described with reference to FIG.

In FIG. 2, a Y stage 31 is mounted on an X stage 30, and a sample 10 such as a mask is mounted on the Y stage 31 via a scanner 28 in the X direction.

The X stage 30 and the Y stage 31 can be moved in the X and Y directions by DC motors 32 and 33, respectively.

The rotational speeds of these DC motors 32 and 33 are detected by tacho generators 34 and 35, and the speed is controlled by an XY servo system 36.

The stages 30 and 31 are moved by using a semi-reflecting mirror 38 and a reflecting mirror 39 as light sources from a laser oscillator 37.
The X and Y stages 30 and 31 are irradiated and reflected, respectively, and their interference light is detected by photoelectric tubes 40 and 41.
LM via amplifiers 42 and 43 and AD converter 44
The signal is input to the S interface circuit 45.

The XY servo system 36 and the LMS interface circuit 45 are mounted on a substrate 47 on which a CPU 46 is mounted, and an image signal from an image data generator 48 generated by a scanning confocal microscope (see FIG. 6) is
The data is input to the image data circuit 49.

In the LMS interface circuit 45,
A line width measuring circuit 5 that counts the pitch of the LMS signal, especially the LMS signal of the Y stage, detects the actual moving speed of the Y stage, corrects the line width based on this, and measures the line width.
0 is formed.

The details of the line width measuring circuit 50 will be described with reference to FIG.

In FIG. 1, a speed command value 52 from a computer 51 is input to a stage servo system 53 for driving and controlling the Y stage 31, and a driving means 53 composed of a motor is controlled based on the speed command value 52.

The LMS signal of the Y stage 31 is supplied to the amplifier 4
2. The encoder signal 54 is converted into an encoder signal 54 via the AD converter 44.
And speed control and position control are performed.

The line width measuring circuit 50 outputs an encoder signal 54
Counter 56 that counts the clock signal with a high frequency reference clock 55 (about 10 MHz), a pitch memory 57 that stores the count value (about 100,000 counts), a control circuit 58 that controls the pitch counter 56 and the pitch memory 57, , Information from control circuit 58 and image data 49
And a computer 51 for measuring the line width from the information from the computer.

The pitch counter 56 receives the encoder signal 5
4 is ON / OFF of the sine wave LMS signal by AD conversion
In order to correctly obtain one cycle of the signal, the encoder signal 54 is frequency-divided by 、, and the odd-numbered and even-numbered waveforms of the 分 frequency-divided waveform are independently obtained. It has a circuit configuration that can be counted.

The odd-numbered and even-numbered count values of the 1/2 frequency-divided waveform are stored in the pitch memory 57.

A block diagram of the encoder pitch measuring circuit will be described with reference to FIG.

In FIG. 3, reference numeral 60 denotes an even / odd address decoder for generating a 1/2 frequency division and an address;
A timing pulse generator, 62 is a pitch data compressor for counting the pitch of the 1/2 frequency-divided waveform, 63 is a RAM, 64 is a system bus interface, 65 is a data buffer, and 66 is an address decoder.

The measurement gate signal 67 is input to the even / odd decoder 60 and the timing pulse generator 61, and the even / odd address decoder 60 outputs the encoder signal 54
Is divided by に for each of even and odd pitch numbers, and is output to the pitch data compressor 62. At the same time, these addresses are sent to the address decoder 66 on the system bus interface 64 side by the timing pulse from the timing pulse generator 61.

In the pitch data compressor 62, an odd / even write signal (EVN / ODD WT) 68, an odd / even gate signal (EVN / ODD GATE) 69, and an odd / even clear signal (EVN / ODD CLR) 70,
The odd half-frequency-divided waveform is counted, and this count value is output to the data buffer 65.

The RAM 63 receives the write signal 71 from the pitch data compressor 62 and stores the count value transferred to the data buffer 65 and the address transferred to the address decoder 66.

The timing chart of FIG. 3 will be described with reference to FIG.

First, when the measurement gate signal indicating the measurement range in (a) is turned on, the encoder signal in (b) is divided into even and odd numbers by 1/2, as shown in (c) and (d), respectively. Then, as shown in (e) and (f), an address of each 1/2 frequency-divided waveform is generated and counted as shown in (g) and (h). This count is generated at the falling edge of the 1/2 frequency-divided waveform of (c) and (d) (i),
Count up by the write signal of (g), and (k),
The gate signal of (l) is stored in the data buffer of (m), the counter is reset by the clear signal of (n) and (o), and the RAM write signal of (p) is generated by the clear signal. The count value is stored in the RAM together with the address of each pitch of the encoder signal.

FIG. 5A shows the pitch count value stored in the RAM as data during one scan of the image 70 in the Y direction, where 71 is a box indicating the measurement range, and 72 is a line. This is a pattern for measuring the width.

Now, in the Y direction of the image, the starting point of the image is A0, the position of the measurement range of the box is Ab1, Ab2,
The positions of the line widths of the pattern are Ap1 and Ap2, the ideal speed obtained by the command speed in the Y direction is Vb, and the actual speed distribution V obtained by the pitch count value of the encoder signal.

The ideal speed V in the measurement range Ab1 to Ab2 is constant, and its value is generated by a computer. Assuming that the actual speed V (pitch count value) fluctuates with respect to the ideal speed Vb (a reference speed count value of the apparatus) as shown in FIG. Since the ratio of the integrated values is the total moving distance, the sum of the measured count values in the measurement area in the RAM (integrated value) is compared with the integrated value of the ideal speed (known) to obtain a coefficient. Correct it by acting on the data.

That is, the speed integral value SbI of the measurement range at the ideal speed is SbI = (Ab2−Ab1) × Vb, and the speed integral value SpI of the line width at the ideal speed is SpI = (Ap2−Ap1). × Vb.

The speed integral value S in the measurement range at the actual speed
bR is

[0045]

(Equation 1)

And the speed integral value Sp of the line width at the actual speed.
R is

[0047]

(Equation 2)

Is as follows.

Therefore, the ratio Rb of the speed integral value SbI of the ideal speed in the measurement range to the speed integral value SbR of the actual speed is Rb = SbI / SbR, and the speed integral value SpI of the ideal speed of the line width and the speed of the actual speed are obtained. The ratio Rp of the integrated value SpR is as follows: Rp = SpI / SpR, and these Rb and Rp are the speed correction values.

Based on the speed correction value, the data is corrected by acting on the actual measured value to obtain final data.

[0051]

In summary, according to the present invention, the error correction with respect to the ideal speed is always performed by counting the pitch of the encoder signal, so that the measurement reproducibility can be improved, and as a result, the measurement accuracy of the device can be improved. .

[Brief description of the drawings]

FIG. 1 is a schematic view of a main part showing an embodiment of the present invention.

FIG. 2 is a block diagram showing the overall configuration of the present invention.

FIG. 3 is a diagram showing a block diagram of a pitch measurement circuit of an encoder signal in the present invention.

FIG. 4 is a diagram showing a timing chart of the measurement operation of FIG. 3;

FIG. 5 is a diagram for explaining measurement data correction after pitch counting in the present invention.

FIG. 6 is a diagram illustrating image generation by a scanning confocal microscope.

FIG. 7 is a diagram illustrating conventional stage control.

FIG. 8 is a diagram showing a change between a command speed of a stage and an actual speed.

[Description of Signs] 10 samples 31 Y stage 52 Speed command value 54 Encoder signal 55 Reference clock signal

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Inventor Masaru Morita Inside Ishikawajima System Technology Co., Ltd. 1-1-17 Kamiosaki, Shinagawa-ku, Tokyo

Claims (3)

[Claims] A sample is placed on a stage, the stage is moved at a predetermined speed command value, an image of the sample is generated by a scanning microscope, an encoder signal during the movement of the stage and image data of the scanning microscope. In the method of improving the line width measurement accuracy in a scanning microscope that measures the line width of a sample from, the pitch of the encoder signal during the movement of the stage is counted with a high-frequency reference clock signal, and the actual speed is obtained from the count value. A method for improving line width measurement accuracy in a scanning microscope, wherein the measurement line width is corrected by comparing the command speed with the actual speed. 2. The encoder signal is frequency-divided by 1/2 for each of even and odd numbers, and the rising and falling of the even and odd 1/2 frequency-divided waveforms are independent of the reference clock signal. 2. The scanning microscope according to claim 1, wherein the counting is performed to count even and odd pitches of the encoder signal as pitch data, and to correct a measurement line width from the reference speed count value of the ideal speed based on the command speed and the pitch data. How to improve line width measurement accuracy. 3. The sum of the reference speed count value of the ideal speed and the sum of the pitch data are obtained for the range corresponding to the line width obtained from the image data, and the measured line width is corrected by the ratio of these sums. A method for improving line width measurement accuracy in the scanning microscope according to claim 2.