JPH0712525A - Thickness measuring instrument using confocal scanning system laser microscope - Google Patents

Abstract

PURPOSE:To provide a thickness measuring instrument which can measure the thickness data between two surfaces with a greatly different reflection factor without reducing the reproduction accuracy of data. CONSTITUTION:A 1/4 wavelength plate 53 is installed so that it can be loading into and out of the light path of reflection light from a beam splitter 50 for body in addition to a confocal scanning type laser microscope body 70 and then a beam splitter 54 for separation is installed at a later stage. Two laser beams 2c and 2d separated by the 1/4 wavelength plate 53 and the beam splitter 54 for separation are applied to a first photomultiplier 55 and a second photomultiplier 56, respectively.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a thickness measuring device for measuring the thickness of an object, for example, a thin film used in a semiconductor manufacturing process, using a confocal scanning laser microscope.

[0002]

2. Description of the Related Art Conventionally, a confocal scanning type laser microscope has been used mainly for measuring dimensions such as a width of a pattern formed on a wafer or a mask when manufacturing a semiconductor integrated circuit.

FIG. 6 shows the principle of a confocal scanning laser microscope. Laser light 2a emitted from the laser oscillator 1
Is reflected by the mirror 3 and squeezed by the lens 4 to pinhole 5
Passed through. The laser light 2a that has passed through the pinhole 5 is
The light passes through the beam splitter 6, is converged by the objective lens 7, and is irradiated onto the sample (wafer, mask, or reticle) 8. The reflected light 2b reflected on the surface of the sample 8 is reflected by the beam splitter 6, passes through the pinhole 9, and is received and amplified by the photomultiplier tube 10.

When performing dimension measurement with such a laser microscope, the position of the focal point F of the laser beam 2a is fixed to the position below (or above) the pattern 11 as shown in FIG. 6 (a). To do. Then, while the sample 8 is scanned in the X-axis direction at a high frequency by the scan controller 12, the sample 8 is slowly moved in the Y-axis direction by an XY stage (not shown). At this time, when the beam spot is passing under the pattern 11 as shown in FIG.
All pass through the pinhole 9, so the photomultiplier tube 1
The output of 0 becomes high. On the other hand, when the beam spot is passing over the pattern 11 as shown in FIG. 6B, the reflected light 2b is not focused at the position of the pinhole 9 and most of it is due to the pinhole 9. Since it is blocked, the output of the photomultiplier tube 10 becomes small. Therefore, by enlarging the brightness according to the output of the photomultiplier tube 10 on the television monitor 13 in synchronization with the scan by the scan controller 12, the enlarged image 14 of the pattern is obtained.
Is displayed on the television monitor 13, and the line width of the pattern 11 can be automatically measured based on the enlarged image 14 with high accuracy.

A specific example of a confocal scanning type laser microscope using the above principle will be described with reference to FIG. FIG. 7 is a device configuration diagram of a confocal scanning type laser microscope (trade name Siscan-IIA), which is a product of Siscan-Systems in the United States, and includes an optical module 16 interconnected with a signal cable 15. It is composed of a workstation 17.

In the optical module 16, an XY stage 20 is mounted on a base 19 supported by an air suspension 18, and a scanner 21 is installed on the XY stage 20. The sample is chucked and held on the scanner 21 by vacuum suction. A focus device 22 is arranged above the scanner 21, and the laser light emitted from the laser head 23 is converged by the objective lens 7 to irradiate the sample on the scanner 21. And
The laser head 23 accommodates the optical system shown in FIG. Further, a TV camera with a zoom lens 24 for taking an image of the vicinity of the laser irradiation position of the sample is provided adjacent to the laser head 23 so as to face downward. Further, below the base 19, the electric circuit chassis 2
5. Various indicators 26 such as a vacuum device for chucking the sample, a laser power source 27, a robot controller for handling the sample and setting it on the scanner 21 and removing it from the scanner 21 are provided. .

On the other hand, the workstation 17 includes an operation panel (keyboard) 28, a television monitor 13, and a printer 2.
9. A control device 30 including a CPU is provided so that all the optical modules 16 can be operated from this workstation 17, and a television monitor 13 is also provided.
On the screen, an observation image by laser light, an image taken by the television camera 24, a menu necessary for operation, and the like are displayed.

Next, FIG. 8 shows a system configuration of the confocal scanning type laser microscope of FIG. Sample 8 is a wafer
The handling robot 31 conveys it onto the scanner 21 and chucks it. The laser beam 2a from the laser oscillator 1 is emitted from the objective lens 7 through the optical module 16 and is applied to the sample 8. Then, the reflected light 2b is received by the photomultiplier tube 10.

Computer of workstation 17
The system 30 controls each unit via a bus 32. That is, the wafer handler control unit 33 drives the wafer handling robot 31 to carry out and carry in the sample 8. In addition, the XY stage motor control unit 34
The XY stage 20 is driven to position the desired inspected portion of the sample 8 as the scan range of the laser beam 2a. In addition, the Z-axis focus control unit 35 controls the objective lens 7
Is moved up and down to focus on the bottom or top of the pattern on the sample 8. Then, if there is a focus, the height of the objective lens 7 is fixed there.

The scan control / synchronization circuit 36 performs scanning waveform read and synchronization control as scan control. Line scan waveform memory 37
Stores a scan waveform such as a sine wave, and reads the stored scan waveform at high speed (for example, 2 kHz) according to a command from the scan control and synchronization circuit 36. The read scan waveform is converted into an analog waveform by the D / A converter 38, and the actuator (voice coil motor, piezoelectric element, etc.) of the scanner 21 is driven in the X-axis direction via the amplifier 39a, and laser light is emitted. 2a
Scan the irradiation position. The scanning speed is maintained at the specified speed by speed feedback.

The output of the photomultiplier tube 10 is the amplifier 3
It is converted into a digital signal by the A / D converter 40 via 9b. Pixel timing and synchronization circuit 41
Represents a correspondence relationship between the detection information of the photomultiplier tube 10 obtained continuously in each scan and the position on the X axis.
An address on one scanning line is given to the data sequentially output from the A / D converter 40, and the data is taken into the line scan pixel memory 42. In addition, line
The scan distortion memory 43 corrects the deviation between the scanning waveform and the actual movement of the scanner 21 so that an accurate pattern is captured in the line scan pixel memory 42.

The sample 8 is scanned by the XY stage 2.
It is performed while slowly moving the Y axis of 0 at a constant speed,
At this time, the contents of the line scan pixel memory 42 are stored in the video display memory 4 for each scan.
4, the pattern image in a desired range on the XY plane is displayed in the display area of the image monitor 45 on the television monitor 13. Further, in the display area of the graphic monitor 46 on the television monitor 13, the detected waveform of the photomultiplier tube 10 in the X-axis direction at the Y-axis position designated by the cursor on the image monitor 45 is displayed. When measuring the pattern width, the image monitor 45
For the range indicated by the cursor above,
The number of pixels of the pattern stored in the display memory 44 is counted, a value obtained by averaging the count values at each position in the Y-axis direction within the cursor is automatically calculated, and the result is displayed as a numerical value on the television monitor 13. It is like this.

[0013]

By the way, in the semiconductor manufacturing process, it is required to measure a plane dimension such as measuring a line width of a pattern formed on a wafer or a mask, and a wafer or a mask is required to be measured. Often it is required to measure the thickness of the formed film. However, although the conventional laser microscope can be used for measuring the planar dimension as described above, it cannot measure the dimension in the thickness direction.

Therefore, the inventors of the present invention have devised a method of measuring the thickness using the laser microscope. As described above, in the measurement of the planar dimension, the focus position of the Z axis is fixed and the laser light is scanned in the X and Y axis directions. The focus axis is operated so as to pass through the two surfaces of the upper and lower surfaces of the object to be measured, and a position where the reflection intensity of the laser light is maximum, that is, a so-called peak position is obtained for each surface, and from the difference, It is intended to obtain the thickness of the measurement object.

However, for example, when measuring the thickness of a chrome pattern formed on a glass substrate of a mask, the two surfaces have a chromium surface on the upper surface side and a glass surface on the lower surface side, and thus have a high reflectance for laser light. Since the reflectance of laser light differs greatly between the chrome that has it and the glass that has high transmittance, set the photomultiplier tube that detects the reflection intensity to an appropriate sensitivity for the reflected light from either side. Was difficult.

Therefore, as a measurement procedure, the sensitivity of the photomultiplier tube is adjusted to be appropriate for the reflected light from one surface, and then the focus axis is operated once, and then the other. It was necessary to perform the operation of the focus axis again after re-adjusting so as to be appropriate for the light reflected from the surface of.

However, when the peak position of the reflection intensity of each surface is to be obtained by the two movements of the focus axis in this manner, the position reproduction accuracy of the mechanism for operating the focus axis is affected by variations, There is a problem in that the accuracy of reproduction of measurement data decreases.

The present invention has been made to solve the above-mentioned problems, and even if the reflectances of two surfaces constituting the thickness of the object to be measured are greatly different, the measurement data can be reproduced. It is an object of the present invention to provide a thickness measuring device using a confocal scanning type laser microscope which can perform measurement without lowering accuracy.

[0019]

In order to achieve the above-mentioned object, a thickness measuring apparatus using a confocal scanning laser microscope according to the present invention is a thickness measuring and measuring method for an object to be measured for thickness. A confocal scanning type laser microscope main body for irradiating a laser beam by operating a focus axis at two points, and a separator for separating the reflected light of the laser beam reflected by the two points of the object; A first and a second photomultiplier tube for respectively measuring the reflection intensities of the two reflected lights obtained through the separator, wherein the first and second photomultiplier tubes are the object The sensitivity is set to be different depending on the reflectance of the two points.

[0020]

According to the thickness measuring apparatus using the confocal scanning laser microscope of the present invention, the reflected light of the laser light is separated by the separator and is incident on the first and second photomultiplier tubes, and Since each photomultiplier tube is set to have different sensitivities according to the reflectances of the two points of the target object, the reflection intensity of each of the two reflected lights can be obtained by only moving the focus axis once. The peak position of can be accurately detected.

[0021]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS. The thickness measuring device of the present embodiment is based on the conventional confocal scanning laser microscope described above,
A quarter wave plate and a beam splitter, which will be described later, are added to the main body, and two photomultiplier tubes are provided. FIG. 1 is a diagram showing an optical system in the thickness measuring apparatus of the present embodiment, in which reference numeral 1 denotes a laser oscillator, 4
Reference numerals 7, 48 and 49 are turning mirrors, 50 is a beam splitter for the main body, 51 is a beam expander, 52 is a collimator lens, and 7 is an objective lens.

The configuration of each device described above is similar to that of the conventional confocal scanning laser microscope, but in the present embodiment, in addition to the confocal scanning laser microscope main body 70 equipped with each device described above. , 1/4 wave plate (separator)
53, a beam splitter (separator) 54 for separation, and first and second photomultiplier tubes 55 and 56 are installed. The quarter-wave plate 53 is installed on the optical path of the reflected light 2b from the main-body beam splitter 50.
It is connected to the four-wave plate moving mechanism 57 so that it can be freely taken in and out of the optical path. Further, a separating beam splitter 54 is installed at the subsequent stage of the quarter wave plate 53. A first photomultiplier tube (hereinafter referred to as PMT1) is provided on the optical path of the reflected light 2c of the separating beam splitter 54.
55) is installed, and the beam splitter 54 for separation
A second photomultiplier tube (hereinafter referred to as PMT2) 56 is installed on the optical path of the transmitted light 2d via a turning mirror 58.

The quarter-wave plate 53 is formed by laminating a plurality of plates made of artificial quartz, and transmits the laser light inside thereof to change the polarization state of the laser light from linearly polarized light to circularly polarized light ( It is a polarizing element for converting into a polarization state having polarization components that are orthogonal to each other. The quarter-wave plate 53 may be a plate made of artificial quartz. On the other hand, the beam splitter 54 for separation reflects or transmits all of the linearly polarized laser light (depending on the polarization direction), but reflects half of the circularly polarized laser light. It is a polarizing prism that has the function of transmitting the other half. In the optical system of the present embodiment, the laser light 2a emitted from the laser oscillator 1 is reflected by the sample 8 and the quarter wavelength plate 53 is formed.
The polarized state is adjusted by passing through another polarizing element (not shown) before being incident on the optical system, and the reflected light 2b is linearly polarized at the time of being incident on the quarter wavelength plate 53.

Therefore, when this embodiment is used as a conventional laser microscope, if the quarter-wave plate 53 is out of the optical path, the reflected light 2b is incident on the beam splitter 54 for separation as linearly polarized light. Here, all of the laser light is reflected toward the PMTs 1 and 55 to create a pattern image, and the pattern width can be measured. When using this embodiment as a thickness measuring device,
If the quarter-wave plate 53 is placed in the optical path, the reflected light 2b is converted from linearly polarized light into circularly polarized light and is incident on the separating beam splitter 54, so that 1/2 of the laser light is separated by the separating beam splitter 54. Are reflected toward the PMTs 1 and 55, and the other half is transmitted through the beam splitter 54 for separation and is incident on the PMTs 2 and 56, and the reflection intensities of the reflected light are detected.

Here, the principle of thickness measurement will be described with reference to FIG. Here, it is assumed that the thickness d of the object (eg, pattern) 60 formed on the substrate (eg, mask or wafer) 59 is measured. As shown in FIG. 2A, the objective lens 7 is moved in the focus axis direction (Z) at the position A on the object 60 and the position B off the object 60.
Respectively) in the axial direction. The reflection intensity of the laser light obtained at this time is as shown in FIG. 2B, and the peak position on the Z axis differs between the A position and the B position. The peak 61 at the A position is due to the surface of the object 60, and the peak 62 at the B position is due to the surface of the substrate 59.
Therefore, the difference between the peak positions is obtained as the thickness d of the object 60.

However, when the thickness of the chrome pattern formed on the glass substrate is measured using the mask as a sample, laser light is emitted at the position A, that is, the chrome surface, and at the position B, that is, the glass surface. The reflectivity of the chrome surface is very different and the reflectivity of the chrome surface is much higher. Therefore, in actual measurement, the two peaks 61 and 62 are detected at the same height as shown in FIG. 2B. It is not possible, for example, as shown in FIG.
As shown in (d), only one peak can be detected. Therefore, two photomultiplier tubes are required so that detection can be performed with sensitivity according to the reflectance at each measurement position.

Next, an example of a specific method for measuring the thickness using the above-mentioned measurement principle will be described. The object to be measured in this example is the chrome pattern 64 formed on the glass substrate 63 of the mask as described above. FIG. 3 shows the flow of the procedure for measuring the thickness, which will be described below.

(1) An X, Y two-dimensional image in a predetermined range including the thickness measurement portion of the sample 8 is created by the same method as in the normal pattern width measurement. At this time, while scanning the laser beam in the X-axis direction, the XY stage 20 is moved at a constant speed in the Y-axis direction to obtain an X, Y two-dimensional image with a magnification of 1000 to 16000 times. At this point, the Z axis (focus axis) is kept fixed at a fixed position, and the quarter-wave plate 53 is out of the optical path.

(2) As shown in FIG. 4, the X, Y
The two cursors 65 and 66 are displayed on the two-dimensional image to measure the thickness, that is, the chrome pattern 6
The operator designates a predetermined position (position A) on the surface 4 and a predetermined position (position B) on the glass substrate 63.

(3) The XY stage 20 is moved to the designated place. In this case, the XY stage control circuit
The axes are controlled to position the XY stage 20 so that the laser beam is scanned in the X-axis direction at the Y-axis position designated by the cursors 65 and 66.

(4) The quarter-wave plate 53 is moved into the optical path so that the laser light is incident on each of PMT1, 55, PMT2, 56.

(5) In the PMTs 1 and 55, the contrast adjustment for the A position, that is, the chrome pattern 64
The sensitivity adjustment for the reflected light from is performed, and the signal from the position A is made large enough not to be saturated so that it can be detected with high sensitivity. This sensitivity adjustment is specifically performed by PMT1 and PMT5.
This is performed by adjusting the input voltage of No. 5 and, for example, the input voltage is set to 700V. In addition,
At this time, the sensitivity may be too low for the signal from the B position to make detection difficult.

(6) Similar to (5), PMTs 2, 56
At B, contrast adjustment for position B, that is, sensitivity adjustment for reflected light from the glass substrate 63,
The signal from the B position is made large enough not to be saturated. Here, the input voltage of the PMTs 2 and 56 is, for example, 1000V.
And set. At this time, the sensitivity of the signal from the position A may become too high and the signal may be saturated.

(7) The focus position of the Z axis is moved to the highest position Zmax above the chrome surface where the reflection intensity data is collected. However, the symbol Z representing the focus position used in this description does not represent the height of the focus position but represents the number of steps for moving the focus.

(8) Scanning in the X-axis direction is performed, data corresponding to the A position in the output signals of the PMTs 1 and 55 is sampled, data in a fixed area near the A position is averaged, and the value is calculated. The reflection intensity at the position A in Zmax is used. At the same time, the data corresponding to the B position in the output signals of the PMTs 2 and 56 is sampled, and the data in a certain area near the B position is averaged.
Let that value be the reflection intensity at the B position in Zmax.

(9) The Z axis is moved to reset the focus position to a position lower by 10 nm than the previous time. It should be noted that this moving distance is preferably set within the range of 4 to 16 nm, which provides good measurement reproducibility.

(10) The steps (8) and (9) are repeated until the focus position reaches the lowest position Zmin for data collection.

(11) When the focus position reaches Zmin, the scanning is finished. Therefore, at this point, two curves 67 and 68 having the peaks of the reflection intensity as shown in FIG. 5A can be obtained from each data.

(12) The focus positions of the peaks of the two curves 67, 68, that is, the maximum points of the reflection intensity, are obtained and set as Z ₁ and Z ₂ , respectively.

(13) Since Z ₁ and Z ₂ obtained in (12) are discrete data at intervals of 10 nm, interpolation processing is performed to improve accuracy. That is, FIG. 5B
As shown in, the peak position Z ₁ and the positions Z ₁ before and after the peak position Z ₁
Quadratic curve 69 from three points of data at −1, Z ₁ +1
And the peak position of the quadratic curve 69 is re-established as Z ₁ ′
Ask as. Similarly, Z ₂ ′ is calculated.

(14) The difference between the positions of Z ₁ ′ and Z ₂ ′ is calculated. However, as described above, Z ₁ ′ and Z ₂ ′ themselves indicate the number of steps, so when calculating the distance, Z
₁ ', Z _2' difference in step size ΔZ of, in this case 10
It is necessary to multiply by nm. That is, the distance W between Z ₁ ′ and Z ₂ ′ can be calculated as W = ΔZ × (Z ₁ ′ −Z ₂ ′).

(15) The final measurement result is obtained by using the correction coefficients c and d depending on the material of the measurement object and the structure. That is, the thickness measurement value W ₁ is W ₁ = c × W + d. The values of c and d are input by the operator according to each measurement object, but normally in the case of this example, c =
1.0 and d = 0.

(16) The measurement result W ₁ obtained as described above is displayed in the graphic monitor 46 area on the television monitor 13 as the thickness of the chrome pattern 64.

As described above, according to the thickness measuring apparatus of this embodiment, the peaks of the reflection intensity of PMT1, 55, PMT2 and 56 are appropriately increased depending on the reflectances of the chrome surface and the glass surface, respectively. Since the sensitivity is adjusted as described above, two peak positions can be reliably detected by combining the data from both PMTs 55 and 56. Therefore, it is necessary to move the focus axis only once when collecting data, which is a drawback of the conventional thickness measuring method. The position reproduction of the mechanism for operating the focus axis to perform the focus axis operation twice is performed. It is possible to solve the problem that the accuracy of reproduction of measured data is deteriorated due to the influence of variations in the characteristics.

Further, the thickness measuring device has a structure in which a conventional confocal scanning type laser microscope main body is equipped with a quarter wavelength plate (1/4 wavelength plate).
Since it can be realized only by adding the wave plate moving mechanism 57) 53, the beam splitter 54 for separation, and the PMT, the device itself can be easily manufactured. Further, since it is possible to change the use as a conventional pattern width measuring device and the use as a thickness measuring device just by moving the quarter-wave plate 53, a very compact system configuration is possible. .

Although one embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and it is needless to say that various design changes can be made in the concrete configuration of each part. is there.

[0047]

As described in detail above, according to the thickness measuring apparatus of the present invention, the reflected light of the laser light is separated by the separator and is incident on the first and second photomultiplier tubes. Since each photomultiplier tube is set to have different sensitivities according to the reflectances of the two points of the object, it is only necessary to move the focus axis once to reflect each of the two reflected lights. The peak position of the intensity can be accurately detected. Therefore, unlike the case of the conventional thickness measuring method, high reproducibility of measurement data can be obtained without being influenced by the position reproducibility in the two movements of the focus axis.

[Brief description of drawings]

FIG. 1 is a diagram showing an embodiment of an optical system of a thickness measuring device of the present invention.

FIG. 2 is a diagram showing the principle of thickness measurement in the present invention.

FIG. 3 is a flow chart showing a procedure of thickness measurement using the embodiment.

FIG. 4 is a front view showing a television monitor screen for setting a measurement position in thickness measurement.

FIG. 5 is a diagram showing an example of a reflection intensity curve of laser light.

FIG. 6 is a diagram showing the principle of a confocal scanning laser microscope.

FIG. 7 is a front view showing a specific example of a confocal scanning laser microscope.

FIG. 8 is a diagram showing a system configuration in the example.

[Explanation of symbols]

 8 sample 53 1/4 wavelength plate (separator) 54 separation beam splitter (separator) 55 first photomultiplier tube (PMT1) 56 second photomultiplier tube (PMT2) 63 glass substrate 64 chromium pattern 70 Confocal scanning laser microscope body

Claims (1)

[Claims] 1. A confocal scanning type laser microscope main body which irradiates a laser beam by operating a focus axis at two points which are thickness measurement measuring points for an object for which thickness measurement is to be performed; A separator for separating the reflected light of the laser light reflected at a point, and first and second photomultiplier tubes for respectively measuring the reflection intensities of the two reflected lights obtained through the separator are provided. The confocal scanning laser microscope is characterized in that the first and second photomultiplier tubes are set to have different sensitivities according to the reflectances of the two points of the object. The thickness measuring device used.