JP2000009443A - Method and device for measuring form - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method and device for measuring a form which can measure a fine pattern accurately. SOLUTION: A laser beam outputted from a laser outputting part 11 is condensed at a focal point 14a of an objective lens 14 set on a substrate 100. The condensed laser beam is reflected near the focal point 14a. The light distribution which is a brightness distribution in directions within a space caused by the reflection depends on a pattern form. A light distribution detecting part 15 detects the light distribution. A CPU 21 collects detected light-distribution data sequentially detected when the focal point 14a so moves as to cross the pattern formed at the substrate 100. Among a plurality of logical light-distribution data obtained logically from a known form, such logical light-distribution data most approximate to the collected detection light-distribution data is obtained. The known form based on the logical light-distribution data is decided as a pattern formed on the substrate 100.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for measuring a shape of a pattern formed on a substrate such as a semiconductor wafer or a glass substrate for a liquid crystal display.
In particular, the present invention relates to a technique capable of measuring the shape of a fine pattern.

[0002]

2. Description of the Related Art Conventionally, as a shape measuring device for measuring the shape of a pattern formed on a substrate such as a semiconductor wafer or a glass substrate for a liquid crystal display, there are known, for example, a confocal microscope, a laser microscope, a surface roughness meter, an optical step meter. and so on.
In this conventional example, a laser microscope shown in FIG. 10 will be described. The laser microscope includes a microscope unit 60 in which an optical system is housed, an arithmetic unit 61 that performs arithmetic processing according to an output from the microscope unit 60, and a microscope unit 60 that moves up, down, left, and right in accordance with the arithmetic result of the arithmetic unit 61. And a driving unit 62 for driving.

The laser light output from the laser output section 63 is reflected by a half mirror 64 and enters a first objective lens 65. The first objective lens 65 focuses the incident laser light on a focal point 65 a of the first objective lens 65. The tube lens 66 condenses the laser light reflected on the substrate 69 via the first objective lens 65. At the position of the focal point 66a of the tube lens 66, the slit plate 6
7, a slit 67a is provided.
The laser light that has passed through 7a is received by light receiving element 68. The light receiving element 68 outputs an output according to the amount of received laser light.

When measuring the shape of the pattern formed on the substrate 69, the focal point 6 of the first objective lens 65 is measured.
5a is scanned on the substrate 69. At this time, if the pattern surface on the substrate 69 deviates from the focal point 65a, the slit plate 6
The laser light is no longer focused at the position of the slit 67a, and the amount of laser light received by the light receiving element 68 decreases. Therefore, the calculation unit 61 controls the drive unit 62 to move the microscope unit 60 in the vertical direction (z direction) so that the amount of laser light received by the light receiving element 68 is maximized. By obtaining the amount of vertical movement of the microscope unit 60 at this time, the height of the pattern formed on the substrate 69 can be obtained. Further, by moving the microscope unit 60 in the horizontal direction (X direction or Y direction), the width of the pattern can also be obtained. Therefore, by moving the microscope unit 60 on the substrate 69, the shape of the pattern formed on the substrate 69 can be measured.

[0005]

However, the prior art having such a structure has the following problems. If the wavelength of the measuring light in the measuring device is sufficiently short with respect to the pattern formed on the substrate as in the conventional example, the diffracted wave diffracted by the pattern can be ignored in geometric optics. Therefore, as in the above-described laser microscope, the distance (Z direction) between the substrate and the first objective lens is adjusted so that the amount of laser light reflected on the substrate is maximized, and the distance between the substrate and the first objective lens is adjusted within the horizontal plane. By scanning a laser microscope, the width, height, and the like of the pattern are obtained from each movement amount, and the shape of the pattern is measured.

However, in recent years, the pattern formed on a semiconductor wafer has begun to be extremely miniaturized, and the fine pattern has a size equal to or smaller than the above-mentioned wavelength of the measuring light. Have been. When measuring such a fine pattern using a measuring device such as a conventional laser microscope, there is a problem that the shape of the pattern cannot be accurately measured due to the influence of a diffracted wave or a cutoff frequency generated in the pattern on the substrate. is there.

[0007] Further, a measuring device such as a scanning electron microscope (SEM) or a scanning tunnel microscope (STM) is used.
The fine pattern described above can also be measured. However, measurement of the shape of a fine pattern or the like is an important inspection step performed many times in a semiconductor manufacturing process, for example, and an apparatus such as an SEM is very expensive,
Since the use procedure and apparatus scale are large, there is a problem that arranging them in the manufacturing process causes an increase in manufacturing cost and a decrease in throughput. Further, there is a problem that nondestructive inspection / measurement cannot be performed in the SEM.

The present invention has been made in view of such circumstances, and provides a shape measuring method and a device capable of accurately measuring a fine pattern with a relatively inexpensive device. Aim.

[0009]

The present invention has the following configuration in order to achieve the above object. That is, the invention according to claim 1 is a shape measuring method for measuring the shape of a pattern formed on a substrate, wherein a step of condensing predetermined measurement light output from a light source on the substrate, The step of detecting a light distribution that is a luminous intensity distribution in each direction in space, which is caused by the collected measurement light being reflected on the substrate, and the step of detecting the collected measurement light and the substrate. A step of relatively moving the measurement light so as to cross the pattern, a step of storing a detected light distribution data group that is a set of the light distributions sequentially detected with the movement, and a pattern of a known shape. A step of obtaining a reference light distribution data group, which is a set of the obtained light distributions, for each of a plurality of patterns having a known shape; and, from the plurality of reference light distribution data groups, approximating the detected light distribution data group. Reference light distribution day Obtaining a group, and determining a shape of a pattern formed on the substrate based on a known shape of a pattern corresponding to a reference light distribution data group approximating the detected light distribution data group. It is characterized by the following.

According to a second aspect of the present invention, there is provided a shape measuring apparatus for measuring a shape of a pattern formed on a substrate,
A light source that outputs a predetermined measurement light, a light collecting unit that collects the measurement light output from the light source on the substrate, and the measurement light that is collected by the light collection unit is reflected on the substrate. Light distribution detecting means for detecting light distribution which is a luminous intensity distribution in each direction in a space generated by the light source, and the measurement light and the substrate condensed by the light condensing means so that the measurement light crosses the pattern. Moving means for moving relatively to the;
A storage unit for storing a detected light distribution data group, which is a set of the light distributions sequentially detected by the light distribution detection unit along with the movement by the moving unit, and a set of the light distribution obtained from a pattern of a known shape A reference light distribution data group, and a reference light distribution acquisition unit that obtains each of a plurality of patterns of known shapes, and from among the plurality of reference light distribution data groups that are obtained by the reference light distribution acquisition unit, the storage unit Approximate data calculation means for obtaining reference light distribution data approximate to the detected light distribution data group stored in the means, based on the known shape of the pattern corresponding to the reference light distribution data obtained by the approximate data calculation means, Determining means for determining the shape of the pattern formed on the substrate.

According to a third aspect of the present invention, in the shape measuring apparatus according to the second aspect, a polarizing means for polarizing the measuring light is provided in the optical path from the light source to the light distribution detecting means.

According to a fourth aspect of the present invention, in the shape measuring apparatus according to the third aspect, the light distribution detecting means detects a P-polarized light distribution based on the measurement light P-polarized by the polarizing means. And S-polarized light distribution based on the measurement light S-polarized by the polarizing means.

[0013]

The operation of the first aspect of the invention is as follows. The measurement light output from the light source is collected on the substrate. The collected measurement light is reflected on the substrate. The luminous intensity distribution in each direction in the space due to the reflection, that is, the light distribution depends on the shape of the pattern. The collected measurement light and the substrate are relatively moved so as to cross the pattern formed on the substrate, and a detected light distribution data group, which is a set of light distribution from the substrate, is stored. A reference light distribution data group approximating the detected light distribution data group is obtained from the plurality of reference light distribution data groups obtained for each of the plurality of known shape patterns. Based on the shape of the pattern corresponding to this reference light distribution data,
The shape of the pattern formed on the substrate is determined. The pattern formed on the substrate is measured as a pattern formed in the shape determined here.

According to the second aspect of the present invention, the measuring light output from the light source is focused on the substrate by the focusing means. The measurement light is reflected on the substrate, and a light distribution, which is a light intensity distribution in each direction in a space depending on the shape of the pattern, is detected by the light distribution detecting means. A detected light distribution data group, which is a set of light distributions sequentially detected by the light distribution detecting means by the relative movement of the measurement light and the substrate by the moving means, is stored in the storage means. The reference light distribution obtaining means obtains a reference light distribution data group, which is a set of light distributions obtained for a pattern having a known shape, for a plurality of patterns having a known shape. The approximate data calculation means obtains a reference light distribution data group that approximates the detected light distribution data group from the plurality of reference light distribution data groups obtained by the reference light distribution acquisition means. The determination means determines the shape of the pattern formed on the substrate based on the shape of the pattern corresponding to the reference light distribution data.

According to the third aspect of the present invention, the polarizing means provided in the optical path from the light source to the light distribution detecting means polarizes the measuring light sent from the light source. The light distribution detecting means detects a light distribution based on the polarized measurement light.

According to the fourth aspect of the present invention, the light distribution detecting means detects the P-polarized light distribution based on the measurement light P-polarized by the polarizing means, and detects the S-polarized light based on the S-polarized measurement light. Detect light distribution.

[0017]

Embodiments of the present invention will be described below with reference to the drawings. <First Embodiment> FIG. 1 shows a shape measuring apparatus according to a first embodiment of the present invention. The shape measuring device includes an optical system unit 1 that irradiates a predetermined pattern on a fine pattern formed on a substrate 100 such as a semiconductor wafer, a glass substrate, or a metal substrate, and a control system unit that controls the optical system unit 1. 2 is provided.

The optical system unit 1 has a laser output section 11 for outputting a laser beam as a predetermined measuring beam. This laser output unit 11 corresponds to a light source in the present invention. The first polarizing plate 12a provided so as to face the laser output unit 11 converts the laser light (indicated by a chain line arrow in FIG. 1) output from the laser output unit 11 into, for example, P-polarized light.

The half mirror 13 reflects a part of the P-polarized laser light passing through the first polarizing plate 12a downward. The laser light reflected by the half mirror 13 enters the objective lens 14. This objective lens 14 corresponds to the light collecting means in the present invention. The objective lens 14 is provided so that its focal point 14a is located on the substrate 100, and the laser beam incident on the objective lens 14 is
The light is focused on the focal point 14a on the zero. The substrate 100 is mounted on a mounting table (not shown).

The laser light focused on the focal point 14a on the substrate 100 is reflected on the substrate 100 near the focal point 14a. The luminous intensity of the laser beam reflected on the substrate 100 to the space above the substrate 100 varies according to the shape of the pattern formed on the substrate 100 near the focal point 14a.
That is, the light distribution, which is the luminous intensity distribution in each direction in the space, changes according to the shape of the pattern.

The objective lens 14 captures light distribution generated by reflection on the substrate 100. This light distribution passes through the half mirror 13 while maintaining the state, and
The light is again P-polarized by the polarizing plate 12 b and received by the light distribution detection unit 15. The light distribution detection unit 15 includes, for example, a plurality of light receiving elements formed in a line shape. Since each of the light receiving elements receives the laser light, the light distribution of the laser light can be detected as the whole light receiving element. The light distribution of the laser light detected by the light distribution detection unit 15 is collected on the control system unit 2 side as detected light distribution data. In addition,
The first polarizing plate 12a and the second polarizing plate 12b correspond to the polarizing means in the present invention. The light distribution detecting unit 15 corresponds to a light distribution detecting unit in the present invention.

The control system unit 2 controls the control system unit 2 in a centralized manner and performs a predetermined arithmetic processing.
U21, a memory 22 for storing detected light distribution data sequentially collected by the CPU 21, a mouse 23 for instructing various settings, a monitor 24 for displaying the shape of a pattern formed on the substrate 100, and a CPU 21. And a drive unit 25 including a drive mechanism such as a three-axis drive type servomotor that drives the optical system unit 1 up, down, left and right in response to the instruction. Note that the CPU 21
Corresponds to the reference light distribution acquisition unit, the approximate data calculation unit, and the determination unit in the present invention, the memory 22 corresponds to the storage unit in the present invention, and the driving unit 25 corresponds to the moving unit in the present invention.

Hereinafter, processing performed by the control system unit 2 in the entire shape measuring apparatus will be described with reference to the flowchart of FIG. In this embodiment, a case will be described in which a predetermined fine pattern is formed on the substrate 100 by a photolithography process or the like, and then the cross-sectional shape of the fine pattern is measured.

Step S1 (Display Initial Screen) When the shape measuring device is activated, the CPU 21 proceeds to FIG.
An initial screen 40 as shown in (a) is displayed on the monitor 24, and the optical system unit 1 is set in an initial state. In the initial screen 40, a process desired by the operator is displayed on the CPU 21.
Instruction, a shape setting instruction 42, and a measurement start instruction 4
3 is displayed. Further, on the initial screen 40, a mouse cursor 5 for indicating each of the instructions 41 to 43 is also displayed. The mouse cursor 5 is the mouse 2
By moving on the screen in conjunction with the movement of 3 and clicking a button (not shown) provided on the mouse 23,
This is for instructing a predetermined command in a portion where the mouse cursor 5 is overlaid.

Step S2 (Initial Setting) The operator places the calibration substrate on a mounting table (not shown) of the shape measuring apparatus. Further, the operator moves the mouse cursor 5 while observing the initial screen 40, and instructs the calibration command 41.
When the calibration command 41 is instructed, the CPU 2
1 outputs a laser beam from the laser output unit 11 and detects a light distribution of the laser beam obtained by being reflected by the calibration substrate. The detected light distribution is collected as calibration detection light distribution data, and calibration is performed based on the calibration detection light distribution data.

The operator instructs a model setting command 42 on the initial screen 40 to select a model corresponding to the shape of the pattern formed on the substrate 100 (FIG. 4A).
reference). When the model setting instruction 42 is instructed, the CPU 21 displays a model setting screen 45 shown in FIG.
4 is displayed.

On the model setting screen 45, models 1 to n are displayed. The models 1 to n correspond to the cross-sectional shape of a fine pattern presumed to be formed on the substrate 100 by a photolithography process or the like. Although only the models 1 to 4 are displayed on the model setting screen 45 of FIG.
Is moved up and down, the display in the model setting screen 45 can be scrolled up and down, and any one of the models 1 to n can be displayed on the model setting screen 45.

For example, the model 1 displayed on the model setting screen 45 indicates that the pattern formed on the substrate 100 has a sectional shape having a step having an angle t and a height h. The model 3 indicates a groove-shaped cross-sectional shape having a width w, a depth h, and an angle t. further,
Model 4 indicates a convex cross-sectional shape having a width w, a height h, and an angle t. The width w, height or depth h, and angle t are elements for specifying the pattern shape. In this step S2, the sectional shape of the pattern formed on the substrate 100 is specified.

When it is estimated that a pattern having the cross-sectional shape shown in the model 3 is formed on the substrate 100, the operator designates the model 3 with the mouse cursor 5. When the model 3 is instructed, the CPU 21 displays the initial screen 40 again. Further, the operator instructs a measurement start command 43 to measure the cross-sectional shape of the pattern.

Step S3 (Measurement of Light Distribution Data) When the measurement start command 43 is instructed, the CPU 21 instructs the optical system unit 1 to start measurement, and the driving unit 2
5 is instructed to move the optical system unit 1. The optical system unit 1 focuses a laser beam on the substrate 100, and the light distribution detection unit 15 detects a light distribution generated by reflection on the substrate 100. At this time, the optical unit 1 is moved by a predetermined distance at a predetermined speed by the driving unit 25. CPU 21
Stores the light distribution detected by the light distribution detector 15 as detected light distribution data in the memory 22 at predetermined intervals.

For example, when a groove-shaped cross-sectional pattern as shown in FIG. 5 is formed on the substrate 100, the focal point 14a where the laser light is focused is moved at a predetermined distance with the movement of the optical system unit 1. Is moved between the measurement position X1 and the measurement position Xn. The measurement position X1 is set, for example, at a position of a distance 1 from the center of the groove-shaped cross-sectional pattern. Note that this measurement position X1 is, for example,
An alignment mark formed at 0 or an end face of the substrate 100 may be used as a reference.

The CPU 21 determines that the focal point 14a is at the measurement position X1.
During the movement from the measurement position Xn to the measurement position Xn, light distribution generated by reflection near the focal point 14a is collected at predetermined intervals as detected light distribution data. As a result, the measurement positions X1, X2,.
, Xn-1 and Xn-1 can be collected (hereinafter, simply referred to as "detected light distribution data group Rr" when indicating a plurality of detected light distribution data as a whole). Here, the light distribution generated near the focal point 14a includes the light distribution generated at the focal point 14a and the light distribution generated near the focal point 14a.

FIG. 6 shows each detected light distribution data detected at each of the measurement positions X1 to Xn. Each detected light distribution data shown in FIG. 6 is represented by a light distribution curve. This light distribution curve is a curve showing, as a function of the direction, the luminous intensity distribution in each direction in the space around the focal point 14a of the laser light reflected by the focal point 14a, and is a polar coordinate having the focal point 14a as the origin. Is represented. Hereinafter, the detected light distribution data collected in the memory 22 will be described as a light distribution curve represented by polar coordinates. Note that, in addition to the light distribution curve represented by the polar coordinates described above, the luminous intensity distribution captured by the entire light distribution detection unit 15 is detected by the horizontal axis at the position of each light receiving element and the vertical axis by the light receiving element. It can also be represented by a coordinate system corresponding to the luminosity to be measured.

The light distribution curve 60 based on the detected light distribution data at, for example, the measurement position X1 shown in FIG. 6 has a substantially semicircular shape. This indicates that the focal point 14a is a mere plane having no uneven pattern on the substrate 100.
The measurement position X3 is a position where the focal point 14a is close to the pattern. The light distribution curve 61 (see FIG. 6) generated by reflection near the focal point 14a at the measurement position X3 is
The light distribution curve 61 is deformed in accordance with the shape of the pattern formed on the light emitting portion 00 and the position of the focal point 14a. Similarly, the light distribution curves at the measurement positions X3 to Xn-1 are as follows.
The shape of the light distribution curve changes according to the shape of the pattern and the position of the focal point 14a. Further, the measurement position Xn
Then, since the focal point 14a moves on a mere plane,
Similar to the measurement position X1, the light distribution curve 62 is substantially semicircular.

Step S4 (determining the pattern shape) The processing performed in step S4 is shown in step T shown in FIG.
This will be described with reference to 1 to T3.

Step T1 (Generates theoretical light distribution data group) The CPU 21 assigns a predetermined value to each of the elements (width w, depth h, angle t) for specifying the model 3 specified in step S2. give. For example, (width w, depth h, angle t) = (0.01 to 0.50, 0.01 to 0.50, 9
0 to 180). Thus, a plurality of types of groove-shaped models 3 are determined. That is, the width w of the model 3
At a pitch of 0.01 μm from 0.01 to 0.50 μm, and a depth h of 0.01 to 0.1 μm at a pitch of 0.01 μm.
The angle t is changed to 50 μm, and the angle t is 90 to 1 at 1 ° pitch.
By changing it to 80 °, a plurality of types of groove-shaped models 3 having different widths w, depths h and angles t are determined. Step T1 corresponds to the function of the reference light distribution acquisition means in the present invention.

Specifically, first, the CPU 21 calculates (width w, height h, angle t) = (0.01, 0.01, 90)
The model 3 of the sectional shape specified by is determined. CPU 21
When the pattern 3 of the model 3 is assumed to be formed on the substrate 100, the same focus 14a of the laser beam as in step S3 is set to the virtual measurement positions Y1 to Yn corresponding to the measurement positions X1 to Xn described above. To move virtually. With the movement of the focal point 14a, the virtual measurement positions Y1 to Y
The light distribution detected by each of n is theoretically obtained as theoretical light distribution data. Hereinafter, when the entire theoretical light distribution data detected at each of the virtual measurement positions Y1 to Yn is indicated, it is simply referred to as a “theoretical light distribution data group”, and the plurality of theoretical light distribution data obtained by the model 3 is used. The whole is referred to as a theoretical light distribution data group Rs1. Note that the theoretical light distribution data and the theoretical light distribution data group correspond to the reference light distribution data and the reference light distribution data group in the present invention, respectively.

Each of the theoretical light distribution data detected at the virtual measurement positions Y1 to Yn determines the focused laser light, the cross-sectional shape of the pattern formed on the substrate 100, the positional relationship between the focal point 14a and the pattern, and the like. It can be theoretically obtained by Maxwell's equation.

Next, the CPU 21 determines a model 3 having a cross-sectional shape specified by (width w, height h, angle t) = (0.02, 0.01, 90), and sets a theoretical light distribution data group. Find Rs2. Similarly, (width w, height h, angle t) = (0.
03, 0.01, 90), (0.04, 0.01, 9)
0),..., (0.50, 0.50, 180) are determined, and the theoretical light distribution data groups Rs3 to Rsn corresponding to each model 3 are determined. . The theoretical light distribution data group R obtained in step T1
FIG. 7 shows s1 to Rsn.

Step T2 (Specify theoretical light distribution data group) Each theoretical light distribution data group Rs1 to R obtained in step T1
sn and the detected light distribution data group Rr obtained in step S3
Is calculated, and a theoretical light distribution data group that is the closest to the detected light distribution data group Rr is obtained. Step T2
Corresponds to the function of the approximate data calculation means in the present invention.

Specifically, first, an error between the detected light distribution data at the measurement position X1 and the theoretical light distribution data at the virtual measurement Y1 is calculated. Similarly, measurement positions X2 to X
n, the detected light distribution data at n and the virtual measurement positions Y2 to Yn
The respective errors are calculated with respect to the theoretical light distribution data in. The amount of error is calculated by calculating the sum of the calculated errors. The calculation of the error amount is also performed for each of the theoretical light distribution data groups Rs1 to Rsn.

The CPU 21 sets each of the theoretical light distribution data groups Rs1 to Rs1 to
The error amount of Rsn is calculated, and, for example, a theoretical light distribution data group Rsi that minimizes the error amount is specified (see FIG. 7). If the measurement position X1 of the detected light distribution data group Rr is shifted from the virtual measurement position Y1 of the theoretical light distribution data group, the error amount may not be obtained accurately. In this case, the measurement positions X1 to Xn of the detected light distribution data of the detected light distribution data group Rr are appropriately shifted, and the shifted detected light distribution data group Rr and the theoretical light distribution data Rs1 to Rsn are shifted.
Is calculated, and a theoretical light distribution data group Rsi that minimizes the error is calculated.

Step T3 (Determine the Element of the Sectional Shape) The CPU 21 determines the element (width w, height h, angle t) of the model 3 of the sectional shape based on the theoretical light distribution data group Rsi = (wi,
hi, ti). Further, the CPU 21 calculates (width w, height h, angle t) = (wi, hi, wi) in accordance with the error amount between the theoretical light distribution data group Rsi and the detected light distribution data group Rr.
ti) is calculated by using the elements (width w, height h, angle t) of the cross-sectional shape model 3 based on the theoretical light distribution data group Rsi + 1 = (wi + 1,
hi + 1, ti + 1) or the element (width w, height h, angle t) of the model 3 of the cross-sectional shape based on the theoretical light distribution data group Rsi-1 = (wi-1, hi-1, ti) Interpolate with each value of -1).
This interpolated element (width w, height h, angle t) = (w
I, hI, tI) are finally determined as elements of the model 3. Step T3 corresponds to the function of the determination means in the present invention.

Step S5 (display the determined shape) The CPU 21 displays the measurement result screen 80 shown in FIG.
4 is displayed. In the initial screen 80 of the measurement result, the elements of the model 3 (width w, height h, angle t) = (wI, hI, t
The cross-sectional shape specified by I) and the value wI of each element,
hI and tI are displayed respectively. The operator grasps the cross-sectional shape of the fine pattern formed on the substrate 100 by observing the measurement result screen 80.

Step S6 (End?) Further, when measuring the cross-sectional shape of another pattern,
Steps S2 to S6 described above are repeatedly performed. If another pattern is not measured, the process ends.

According to the above-described apparatus, by detecting the luminous intensity distribution of the laser light (measurement light) reflected on the substrate 100 in each direction in the space, the sectional shape of the fine pattern formed on the substrate 100 is detected. Is measured. In other words, by using a measurement method that is completely different from a measurement method such as a laser microscope, which is a conventional shape measurement device, it has a simpler configuration than the conventional device and has a finer structure that cannot be measured by the conventional device. It can measure various patterns. In addition, from a plurality of theoretical light distribution data groups theoretically obtained from a plurality of known shape patterns, a theoretical light distribution data group most similar to a detected light distribution data group obtained by measurement is obtained. The element for specifying the shape of the pattern based on the theoretical light distribution data group is interpolated by the element for specifying the shape of the pattern based on the theoretical light distribution data near the calculated theoretical light distribution data.
The shape of the pattern formed in the pattern can be obtained more accurately.

<Second Embodiment> Next, a second embodiment will be described with reference to FIG. FIG. 9 is a block diagram illustrating a shape measuring apparatus according to the second embodiment. Portions common to the first embodiment described above are denoted by the same reference numerals, and description thereof is omitted. As shown in FIG. 9, the shape measuring apparatus according to the second embodiment is different from the shape measuring apparatus according to the first embodiment in that, instead of the second polarizing plate 12b, a polarization beam that polarizes measurement light into P-polarized light and S-polarized light. It has a splitter 18.

The polarization beam splitter 18 serves to P-polarize and S-polarize the light distribution of the laser beam passing through the half mirror 13. For example, the laser light that passes through the polarization beam splitter 18 and enters the first light distribution detection unit 15a is P-polarized. On the other hand, the laser light that is reflected in the polarization beam splitter 18 and enters the second light distribution detection unit 15b is S-polarized. First polarizing plate 12a
The polarization beam slitter 18 corresponds to a polarization unit in the present invention, and the first light distribution detecting unit 15a and the second light distribution detecting unit 15b correspond to a light distribution detecting unit in the present invention.

The first light distribution detecting section 15a outputs a P-polarized P
Detect the polarized light distribution. The second light distribution detection unit 15b detects S-polarized S-polarized light distribution.

The CPU 21 collects the P-polarized light distribution sequentially detected by the first light distribution detecting unit 15a as the P-polarized light detection light distribution data in accordance with the movement of the focal point 14a.
The S-polarized light distribution sequentially detected by the light distribution detector 15b is collected as S-polarized light distribution data. The collected P-polarized light distribution data group and the S-polarized light distribution data group are stored in the memory 2.
Stored in 2.

For example, the memory 22 stores the measurement positions X1 to X
The two data of the P-polarized light detection light distribution data and the S-polarization detection light distribution data at each of the n measurement positions are stored as the detected light distribution data at each of the measurement positions X1 to Xn.

When obtaining the theoretical light distribution data at step T1 in the first embodiment, the CPU 21 calculates the theoretical P-polarized light detection light distribution data obtained by theoretically obtaining the light distribution of the P-polarized laser light and the S-polarized light distribution data. Theoretical light distribution data including theoretical S-polarization detection light distribution data obtained by theoretically calculating the light distribution of the obtained laser light. Hereinafter, the shape of the pattern formed on the substrate 100 will be obtained as the detected light distribution data and the theoretical light distribution data of the first embodiment.

The shape measuring apparatus according to the second embodiment detects and distributes P-polarized light distribution data based on P-polarized light distribution and S-polarized light distribution data based on S-polarized light distribution. Since optical data is used, when calculating the amount of error between the detected light distribution data group based on all these data and the theoretical light distribution data group, the number of error elements to be calculated increases, and it is possible to obtain a more appropriate error amount. it can.

The detected light distribution data and the theoretical light distribution data (see FIGS. 6 and 7) detected in each of the above-described embodiments are:
The shape of the light distribution curve changes depending on the material of the substrate 100 to be measured, the state of the measurement light, the atmosphere around the substrate, and the like. Therefore, the present embodiment is merely an example, and the present invention is not limited to this.

The present invention can be modified as follows. (1) In the first embodiment described above, the case where the measurement light is P-polarized has been described. However, the present invention is not limited to this, and is also applicable to the case where the measurement light is S-polarized perpendicular to the P-polarized light. be able to.

(2) In each of the above embodiments, the substrate 100
Although the cross-sectional shape of the pattern formed in was described,
The three-dimensional shape of the pattern can be measured by moving the focal point 14a where the measurement light is converged over the entire surface of the substrate 100.

(3) In each of the above embodiments, the substrate 100
Is considered as a single material, but for example, in the initial setting (step S2), a material composed of two or more types of materials may be selected together with the models 1 to n.

(4) In each of the above-described embodiments, laser light was used as the measurement light.
For example, a mercury lamp, a halogen lamp, a xenon lamp, or the like may be used. However, it is preferable that the light is monochromatic light using an optical filter or the like.

(5) In each of the above embodiments, the first polarizing plate 12a, the second polarizing plate 12b, or the polarizing beam splitter 18 is used as the polarizing means. , Half mirror 13.

(6) In each of the above-described embodiments, the light distribution detecting section 15 is constituted by a line type light receiving element. However, for example, the light distribution detecting section 15 is constituted by a matrix type light receiving element.
A surface sensor such as a CCD may be used.

(7) In each of the above embodiments, the focal point 14a
Although the optical system unit 1 is configured to move to move the substrate, for example, the substrate 100 may be configured to move.

(8) In each of the above-described embodiments, the first polarizing plate 12a is provided at a position facing the laser output unit 11. For example, the laser light (measurement light) output from the laser output unit 11 is already provided. When polarized, the first polarizing plate 12a can be removed.

(9) In each of the above-described embodiments, step T
In FIG. 1 (FIG. 3), the theoretical light distribution data and the theoretical light distribution data group are obtained as the reference light distribution data and the reference light distribution data group by the theoretical operation by the CPU 21, but a plurality of shapes whose shapes are actually confirmed are obtained. By preparing the patterns and performing the process of step S3 (FIG. 2) for each pattern, the measurement data may be acquired as the reference light distribution data and the reference light distribution data group.

[0064]

As is apparent from the above description, according to the first aspect of the present invention, the measurement light condensed on the substrate is reflected and directed in each direction in the space depending on the shape of the pattern. Since the shape of the pattern formed on the substrate is measured based on the light distribution that is the luminous intensity distribution, even if the pattern shape is not sufficiently large compared to the wavelength of the measurement light, the shape of the pattern can be measured. It can be measured accurately.

According to the invention described in claim 2, according to claim 1
And the configuration of the device can be simplified as compared with the conventional method, so that the device can be manufactured at low cost.

According to the third aspect of the present invention, since the light distribution based on the measurement light polarized by the polarizing means is detected by the light distribution detecting means, the shape of the pattern formed on the substrate can be further improved. It can be measured accurately.

According to the fourth aspect of the present invention, both the P-polarized light distribution based on the P-polarized measurement light and the S-polarized light distribution based on the S-polarized measurement light are detected. The shape of the pattern formed on the substrate can be measured more accurately.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a shape measuring apparatus according to a first embodiment.

FIG. 2 is a flowchart illustrating a flow of a process performed by the shape measuring device.

FIG. 3 is a flowchart showing a flow of a process performed in step S4.

FIG. 4 is a diagram showing an initial screen and a model setting screen displayed on the shape measuring device.

FIG. 5 is a diagram showing how a pattern is measured by a shape measuring device.

FIG. 6 is a diagram showing a detected light distribution data group detected by the shape measuring device.

FIG. 7 is a diagram showing a theoretical light distribution data group theoretically obtained by the shape measuring device.

FIG. 8 is a diagram showing a measurement result screen.

FIG. 9 is a block diagram illustrating a shape measuring apparatus according to a second embodiment.

FIG. 10 is a block diagram showing a conventional shape measuring device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Optical system unit 2 ... Control system unit 11 ... Laser output part 12a ... 1st polarizing plate 12b ... 2nd polarizing plate 13 ... Half mirror 14 ... Objective lens 15 ... Light distribution detection part 15a ... First light distribution detection part 15b ... Second light distribution detecting section 21... CPU 22... Memory 23.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2F065 AA54 BB02 BB18 CC17 FF41 GG04 HH04 HH08 LL32 MM01 QQ24 QQ39 RR08 4M106 BA05 CA19 CA39 DB02 DB08 DB12 DB14 DJ17 DJ20 DJ21

Claims (4)

[Claims] 1. A shape measuring method for measuring a shape of a pattern formed on a substrate, comprising: condensing predetermined measurement light output from a light source on the substrate; The process of detecting light distribution, which is a luminous intensity distribution in each direction in a space, which is caused by light reflecting off the substrate, and the measurement light converging the condensed measurement light and the substrate to form the pattern. A step of relatively moving so as to cross, a step of storing a detected light distribution data group which is a set of the light distribution sequentially detected with the movement, and a step of storing the light distribution obtained from a pattern of a known shape. A step of obtaining a set of reference light distribution data groups for each of a plurality of patterns of known shapes; and a reference light distribution data group approximating the detected light distribution data group from the plurality of reference light distribution data groups. The process of seeking Determining a shape of a pattern formed on the substrate based on a known shape of a pattern corresponding to a reference light distribution data group approximating the detected light distribution data group. Measuring method. 2. A shape measuring device for measuring the shape of a pattern formed on a substrate, comprising: a light source for outputting predetermined measuring light; and a measuring light output from the light source condensed on the substrate. Light collecting means, and light distribution detecting means for detecting light distribution, which is a luminous intensity distribution in each direction in a space, generated by the measurement light collected by the light collecting means being reflected on the substrate, Moving means for relatively moving the measurement light condensed by the light condensing means and the substrate so that the measurement light crosses the pattern; and sequentially detecting the light distribution detecting means with the movement by the moving means. Storage means for storing a detected light distribution data group, which is a set of light distributions, and a reference light distribution data group, which is a set of light distributions obtained from a pattern of a known shape, is converted into a plurality of patterns of a known shape. About each of Reference light distribution obtaining means, and a plurality of reference light distribution data groups obtained by the reference light distribution obtaining means, the reference light distribution data approximating the detected light distribution data group stored in the storage means. Approximation data calculating means to be obtained, and determining means for determining the shape of the pattern formed on the substrate based on the known shape of the pattern corresponding to the reference light distribution data obtained by the approximate data calculating means. A shape measuring device characterized by the above-mentioned. 3. The shape measuring apparatus according to claim 2, further comprising: a polarizing unit that polarizes the measuring light in an optical path from the light source to the light distribution detecting unit. 4. The shape measuring apparatus according to claim 3, wherein the light distribution detecting means detects a P-polarized light distribution based on the measurement light P-polarized by the polarizing means, and S-polarized light by the polarizing means. A shape measuring device for detecting S-polarized light distribution based on the measured measuring light.