JP3427141B2 - Pattern position measuring method and apparatus - Google Patents

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 the position of a pattern formed on a substrate, and more particularly to precisely measuring the position of a pattern formed on a mask, reticle or the like used in a projection exposure apparatus. A method and apparatus.

[0002]

BACKGROUND ART Generally, when measuring the position of a pattern formed on a substrate such as a mask or reticle, the substrate to be measured is supported on a stage substantially horizontally and the edge of the pattern is optically detected. To do. In this type of measurement method, since the substrate supported on the stage bends due to its own weight, it is necessary to correct the displacement amount due to the bending. Therefore, as disclosed in, for example, Japanese Patent Laid-Open No. 5-332761, the deflection shape of the reference substrate is detected in advance by measurement or simulation, and the displacement amount (lateral displacement amount) of the pattern on the substrate due to this deflection is calculated. is doing. Then, the displacement amount is used as a correction value and the actually measured pattern position is corrected to obtain the position of the pattern in the state without any bending.

[0003]

However, since the actually measured substrate has a slight error with respect to the reference substrate, it is difficult to correct it according to the theory. In particular, at the present time when the pattern formed on the substrate becomes higher in density and finer, the conventional method as described above cannot obtain sufficient pattern position measurement accuracy. Therefore, the present invention was created with the object of improving the position measurement accuracy of the pattern formed on the substrate.

[0004]

In the present invention, in order to solve the above-mentioned problems, the thickness of the measured substrate (or the surface height of the measured substrate mounted on the stage) is measured. The pattern measurement position is corrected in consideration of the thickness error with respect to the reference substrate. That is, the first correction value for the position of the pattern is obtained as a theoretical value based on the bending shape of the reference substrate. Next, the thickness (or surface height) is measured for each substrate to be measured, and the amount of thickness error with respect to the thickness of the reference substrate is detected. Based on this thickness error, the first correction value is corrected and the second correction value is calculated. Then, the pattern position of the substrate to be measured is corrected based on the second correction value.

The thickness of the substrate is one of the parameters that has a large specific gravity in detecting the bending of the substrate. Therefore,
If the thickness (or surface height) of the substrate to be measured is actually measured to obtain a thickness error and the pattern position is corrected based on this, the measurement accuracy of the pattern position is improved.

The thickness of the substrate to be measured can be measured by detecting the surface height of the substrate. Further, it is desirable that the height of the surface of the substrate to be measured be near the support point of the substrate.

In addition to the substrate to be measured having a thickness error in manufacturing as described above, a plurality of substrates to be measured having different thickness standards may be used to measure the pattern positions. The amount of displacement of the pattern due to bending is calculated in advance, the difference in thickness from the reference substrate is detected for each substrate to be measured, and the amount of displacement is corrected based on this difference in thickness, and then the measurement result of the pattern position is calculated. If corrected, the measurement accuracy can be similarly improved, and since it is not necessary to store the displacement amount for each thickness of the substrate to be measured, the stored data can be greatly reduced. in this case,
It is only necessary to obtain the displacement amount of the pattern due to the deflection of only one reference substrate, but when the difference in the thicknesses of the substrates to be measured is extremely large, the plurality of substrates having different thicknesses are subject to the respective deflections. It is desirable to find the amount of displacement of the pattern.

Although the amount of displacement of the pattern due to the bending of the reference substrate is stored, the bending shape of the reference substrate (specifically, the height position of the substrate surface) is stored. , A function (for example, a quartic function) having the position of the substrate in the horizontal plane as a variable may be approximated, and the coefficient (parameter) of the approximated function may be stored or used as a reference. You may make it memorize | store the gradient according to the bending shape in the board | substrate. Particularly in the latter case, it is desirable to divide the substrate into a plurality of areas and store the gradient for each area. When the bending shape (function) and the gradient are stored, the displacement amount of the pattern due to the bending is calculated based on the stored data.

[0009]

EXAMPLES The present invention will be described in detail below with reference to examples. FIG. 1 shows the overall configuration of a pattern position measuring apparatus embodying the present invention, and FIG. 3 shows the configuration of a control portion of the apparatus. The pattern position measuring apparatus according to the present embodiment detects the position of a fine pattern (not shown) formed on a substrate 10 such as a reticle, and includes an optical device 12 and an XY stage 15 on which the substrate 10 is placed. The objective lens 11 arranged on the upper part of the substrate 10, the movable mirrors 13a and 13b arranged on the upper end portion of the XY stage 15, and the XY stage 15
Interferometers 14a and 14 for detecting the positions of the X and Y directions, respectively.
b and the light receiving element 50 arranged around the objective lens 11.
a, 50b, 51a, 51b, a display device 21 that displays the measured pattern position, a drive device 150 that drives the XY stage 15, a storage device 22 that stores predetermined data, and a general device overview. Control device 20 for controlling
It has and. As shown in FIG. 2, the substrate 10 is placed on the XY stage 15 with the pattern forming surface facing upward.
The substrates are placed without being adsorbed on the substrate rests 6a, 6b, 6c provided at the places.

As shown in FIG. 3, the optical device 12 automatically adjusts the focal position by driving the laser light source 12a for irradiating the substrate 10 with a laser spot through the objective lens 11 and the objective lens 11 in the Z direction. Autofocus device 12
and b. The pattern formed on the substrate 10 is magnified by the objective lens 11 and imaged at a predetermined position in the optical device 12. Autofocus device 12
The focus detection method according to b is described in, for example,
The synchronous detection method disclosed in Japanese Patent No. 44325 can be used. The autofocus device 12b is configured to detect the surface height of the substrate 10 and send a signal according to the height to the control device 20.

The driving device 150 includes a motor (not shown), which moves the XY stage 15 in the XY two-dimensional directions. The interferometers 14a and 14b are movable mirrors 13.
The position of the XY stage 15 is detected by irradiating the reflecting surfaces of a and 13b and the reflecting surface of a fixed mirror (not shown) with a laser beam for length measurement and receiving the reflected light. The information is sent to the control device 20. The four light receiving elements 50a, 50b, 51a, 51b are the substrate 10
The scattered light or the diffracted light generated at the uneven edge portion of the upper pattern is received, and the detection signal is sent to the control device 20. Based on this edge detection signal, the control device 20 detects the edge position of the pattern. The details of such an edge detection method are disclosed in Japanese Patent Publication No. 56-25964.

A two-dimensional (X-direction, Y-direction in FIG. 1) displacement amount (reference correction amount) of a pattern position caused by bending of the substrate 10 due to its own weight at a plurality of locations is stored in the storage device 22 in advance. There is. Further, a plurality of areas are defined on the substrate 10 corresponding to the positions where such correction amounts are obtained, and the data thereof is stored in the storage device 22. For example, as shown in FIG. 4, assuming that the correction amounts at 25 positions on the substrate 10 are stored in the storage device 22, the control device 20 causes the correction amounts to be obtained at the respective positions (31a to 31n, 31p to 31).
25 regions are defined around z). The correction amount may be obtained in advance by actual measurement, or may be obtained by using a simulation technique such as the finite element method. The method of calculating the correction amount by actual measurement will be described in detail later.

The storage device 22 stores reference correction amounts (first correction amounts) Δx and Δy for a plurality of types of substrates having at least one of different materials, shapes and sizes. It should be noted that, for a plurality of substrates having different formation conditions (material, size, etc.) other than the thickness, the reference correction value is stored for only one type of thickness. This reference correction amount is modified according to the actual thickness of the substrate to be measured. For example, when the thickness of the reference substrate (ideal thickness) is t and the actual thickness of the measured substrate is t ', the difference Δt between the two is taken as the thickness error amount, and the reference correction amount To fix. As an example, if this corrected reference correction amount (second correction amount) is Δx ′, Δy ′, Δx ′ = Δx · | t / (t−Δt) | = Δx · t / T 'Δy' = Δy · | t / (t-Δt) | = Δy · t / t 'The amount of bending of the substrate depends on the thickness of the substrate, and thus the thickness is It is important to convert the reference correction value in consideration of the error amount.

Next, the overall operation of the pattern position measuring device configured as described above will be described with reference to the flowchart of FIG. Here, the control device 20 stores the reference correction value in advance (step 1). X with the substrate to be measured (substrate 10) with the pattern formation surface facing up
It is set on the substrate mounting portions 6a, 6b, 6c on the Y stage 15 (step 2). After that, the XY stage 15 is moved to adjust any of the substrate mounting portions 6a, 6b, and 6c so as to be positioned on the optical axis of the optical device 12, and the position of the surface of the substrate 10 is adjusted by the autofocus device 12b. (Height) is detected (step 3).

Here, the detection of the in-focus position by the autofocus device 12b will be briefly described. First, the laser light from the laser light source 12a is imaged in a spot shape (or a slit shape) on the substrate 10 via the objective lens 11, and the reflected light from the substrate 10 is re-imaged via the objective lens 11. At this time, the position of the pinhole (or slit) is simply oscillated in the optical axis direction (Z-axis direction) around the predetermined focusing surface, and the light transmitted through the pinhole (or slit) is not shown in the photodetector. Receives light. This transmitted light is photoelectrically converted by this photodetector, and the output signal obtained from the photodetector by this photoelectric conversion is synchronously detected (synchronous rectification) at the frequency of simple vibration. As a result, an S curve signal in which the voltage value with respect to the position in the Z direction changes in an S shape as shown in FIG. 6 is obtained.

The S-curve signal has a characteristic that the defocus amount d and the voltage value V have linearity in a small section before and after the in-focus position d0, and the voltage value V becomes zero at the in-focus position d0. Therefore, the height of the surface of the substrate 10 in the Z direction with respect to the in-focus position d0 can be easily obtained based on the S curve signal, that is, an ideal moving horizontal plane of the XY stage 15 that moves the substrate 10 two-dimensionally. The distance between the (reference plane) and the substrate 10 can be detected.

The controller 20 controls the position of the substrate 10 in the Z-axis direction, the position of the upper surface of the XY stage 15 which is stored in advance, the distance between the XY stage 15 and the substrate 10 (the substrate mounting portions 6a, 6b, 6c). Substrate height from the three elements)
The thickness of 0 is calculated (see FIG. 2 (b)). Then, the thickness (theoretical thickness) of the reference substrate stored in the storage device 22 under substantially the same forming conditions (except the thickness) as the substrate 10.
t is read, and the difference between this value t and the actual thickness t ′ of the substrate 10 is calculated as the thickness error amount Δt (step 4).
Next, the reference correction values (theoretical correction values) Δx and Δy are changed based on the thickness error amount Δt, and the changed reference correction amounts Δx ′ and Δy ′ are stored in the storage device 22. (Step 5). At this time, the controller 20 adjusts the correction amount Δ for each of the plurality (25) of areas on the substrate as shown in FIG.
Calculate x'and Δy '.

Thereafter, in response to a measurement start command from an input device (not shown), the control device 20 monitors the stage position signals from the interferometers 14a and 14b, and controls the drive device 150 so that the XY stage 15 comes to the initial position. To do. Next, the control device 20 sequentially moves the XY stage 15 from the initial position, and causes the spot light from the optical device 12 to relatively scan the substrate surface. When the spot light hits the pattern edge on the substrate surface, scattered light is generated there, and this is detected by the light receiving elements 50a, 50b, 51a, 51b. When the light receiving elements 50a, 50b, 51a, 51b detect the scattered light, the light receiving elements 50a, 50b, 51a, 51b output an edge detection signal corresponding thereto to the control device 20.

When the controller 20 receives the edge detection signal, the controller 20 determines the position of the pattern edge at that time by the interferometer 14a,
While reading from 14b (step 6), the correction amount (Δx ′, Δy ′) of the area including the pattern position is read from the storage device 22. The correction amount at this time is revised according to the thickness error Δt of the substrate as described above.

Thereafter, the control device 20 corrects the position of the detected pattern edge in the XY two-dimensional directions based on the correction value read from the storage device 22 (step 7).
Then, the control device 20 obtains the pattern position and the line width based on the corrected edge position, and stores these in the storage device 2
It is stored in memory 2 and displayed on the display device 21 (step 8). The control device 20 stores the corrected edge position in the storage device 22, and according to the operator's instruction from an input device (not shown), the pattern interval on the substrate is determined based on the stored edge position. The calculated intervals are displayed on the display device 21.

In the above description, with respect to a plurality of substrates having different forming conditions other than the thickness, the reference correction amount is stored in the storage device 22 for only one type of thickness. If the formation conditions are substantially the same but the thicknesses are significantly different, it is preferable to obtain and store the reference correction amounts for a plurality of reference substrates having different thicknesses. At this time, the reference correction amount for the reference substrate having the thickness closest to the thickness of the measured substrate is read from the storage device 22, and the read reference correction amount is the difference in thickness between the measured substrate and the reference substrate. Will be corrected in.

The amount of displacement of the pattern due to the bending of the reference substrate is stored in the storage device 22 as the reference correction amount. For example, the bending shape of the reference substrate (specifically, the substrate surface height It is also possible to approximate the position) with a function (for example, a quartic function) having the position of the substrate in the horizontal plane as a variable, and store the coefficient (parameter) of the approximated function. Particularly in the latter case, it is desirable to divide the substrate into a plurality of areas and store the gradient for each area. When the bending shape (function) and the gradient are stored, the displacement amount of the pattern due to the bending is calculated based on the stored data.

Next, a basic method (principle) of obtaining the reference correction amount stored in the storage device 22 will be briefly described with reference to FIG. FIG. 7 shows a cross section of the substrate when it is placed on the three substrate placement parts on the stage 15. The solid line shows the state where the substrate is not bent by its own weight, and the broken line shows the state where the substrate is bent by its own weight. First, in the state of (a), the bending shape due to the own weight of the substrate is detected. Next, (a)
From this state, the substrate is turned upside down as shown in (b), and the bending shape due to the weight of the substrate at this time is detected. The bending shape of the substrate can be easily obtained by measuring the height of the substrate surface from the reference plane at a predetermined interval using the autofocus device 12b.

When the data of the bending shape detected in the state of FIG. 7B is mirror-inverted to the left and right, the bending shape as shown in FIG. 7C is obtained. Further, when the average data of the bending shape data in (a) and the bending shape data in (c) is illustrated, the bending shape is as shown in (d).
The bending shape of the substrate in (d) shows the bending shape of the ideal substrate without deformation. That is, from the above-described average data of the bending shape, it is possible to obtain the bending shape only by the weight of the substrate without being affected by the deformation (waviness) of the substrate. Then, the gradients at a plurality of points on the substrate surface are calculated from the average data, and the two-dimensional displacement amounts of the pattern position caused by these gradients are calculated at a plurality of points and stored in the storage device 22.

In the above embodiment, the substrate 10 is placed on the XY stage 15 with the pattern surface facing upward, but generally, the substrate 10 is used while being supported with the pattern surface facing downward. There are many cases where Therefore, as disclosed in, for example, Japanese Unexamined Patent Publication No. 6-18220, if the deflection of the substrate when the substrate 10 is supported with the pattern formation surface facing downward is also taken into consideration and corrected, the actual substrate is obtained. It is possible to measure the pattern position in a state closer to the state in which is used. Note that the bending shape when the substrate 10 is supported with the pattern forming surface facing downward is obtained in advance outside the apparatus, for example, by using a substrate having the same material and shape as the substrate 10.

The basic method (principle) of obtaining the reference correction amount has been described above with reference to FIG. 7, but here, first, when the substrate 10 is supported with the pattern surface facing upward, The details of the flexure shape and the method of obtaining the reference correction value based on the flexure shape will be described. When the XY stage 15 is at the initial position, the height measurement point 31 a shown in FIG. 8 comes on the optical axis of the objective lens 11 of the optical device 12. Control device 20
By reading the output voltage before the focus function of the autofocus device 12b of the optical device 12 operates,
The height H31 of the pattern forming surface is measured, and the height measurement point 31
It is stored together with the position of the XY stage 15 corresponding to a.

The control device 20 sequentially sets the heights H31b to H31n and H31p to H31z of the pattern forming surface at the height measuring points 31b to 31n and 31p to 31z on the substrate 10.
Is stored together with the position of the XY stage 15 at each height measurement point. Next, height measurement points 31a arranged in the X direction
Based on the relationship between the height of 31e and the position of the XY stage 15, the bending shape in the line 32a in the X direction can be calculated as follows.
It is approximated by the following formula. ^{z = a1 · X 4 + a2} · X 3 + a3 · X 2 + a4 · X + a
5 In the above equation, since the unknowns a1 to a5 are five for the five data of z and X, the fourth-order approximation equation is uniquely determined.

In this way, height measurement points 31f to 31j, 31k to 31n and 31p and 31q in the X direction are sequentially measured.
Also for 31u and 31v to 31z, a fourth-order approximate expression of the bending shape is obtained. Furthermore, the bending shape of the line 32b in the Y direction is approximated by the following quartic equation. ^{z = a1 · Y 4 + a2} · Y 3 + a3 · Y 2 + a4 · Y + a
5

Similarly, height measurement points 31 in the Y direction are sequentially obtained.
b-31w, 31c-31x, 31d-31y, 31e
A quaternary approximation formula of the flexure shape is obtained for .about.31z. As a result, the bent shape of the pattern formation surface of the substrate 10 is obtained as shown in FIG.

Next, the controller 20 controls the XY stage 15
After returning to the initial position, the drive device 150 is controlled to
The Y stage 15 is sequentially moved from the initial position. When the substrate 10 is relatively scanned with the spot light from the optical device 12,
Edge detection signals are output from the light receiving elements 50a, 50b, 51a, 51b. The controller 20 controls the XY stage 1 when the edge detection signal is output from the position signals of the interferometers 14a and 14b when the edge of the pattern is detected.
Read position 5.

For example, assume that edge detection signals are output at the pattern edge positions (hereinafter, referred to as pattern positions) 33a and 33b in FIG. Pattern positions 33a, 3
Coordinate values corresponding to the position of 3b are read from the interferometers 14a and 14b and stored. The control device 20 recognizes the X coordinate value of the pattern position 33a, and of the quaternary approximation formulas previously obtained, the height measurement points 33c and 33d on the approximation formulas adjacent to the pattern position 33a are in the X direction. Slope θX
3, θX4 is calculated. These gradients θX3, θX4
Can be obtained by differentiating the previously calculated fourth-order approximation formula and substituting the X coordinate value.

The pattern position 33a and the height measuring point 33c,
When the positional relationship with 33d is as shown in FIG. 8, the gradient θX1 in the X direction of the pattern position 33a is calculated by the following formula by proportional distribution. θX1 = (L2 · θX3 + L1 · θX4) / (L1 +
L2) The gradient θX2 in the X direction at the position 33b of the other pattern is calculated in the same manner.

The gradients θY1 and θY2 in the Y direction are calculated in the same manner. Next, pattern positions 33a, 33
b correction amount (t · θX1) / 2, (t · θY1)
/ 2, (t · θX2) / 2, and (t · θY2) / 2 are calculated. Note that t is the thickness of the substrate 10 (reference thickness). Using this correction amount, the coordinate value corresponding to the position of the pattern in the state without bending is obtained from the coordinate value corresponding to the position of the pattern detected by the interferometers 14a and 14b. Here, the bending shape in the line 32c in the X direction where the pattern positions 33a and 33b are located is the bending shape of the pattern forming surface 33 in FIG. 9A, and is considered to be an arc shape with the point 0 as the center.

The correction amount is such that the neutral surface 10 'does not expand or contract,
Since the dimensional change amount of the substrate 10 due to the deformation of the neutral surface 10 'is very small, it can be ignored and can be immediately obtained from the gradient. The correction amount in the Y direction can be similarly considered.

When the positions of the patterns are measured on a plurality of substrates to be measured which have substantially the same formation conditions other than the thickness, first, the autofocus device 12b is used on the first substrate to be measured as described above. The flexure shape (approximate function) is calculated to calculate the gradient, and the displacement amount (correction amount) of the pattern is calculated based on the gradient and stored in the storage device 22. Then, the pattern position is corrected on the first measured substrate based on the calculated correction amount, and the thickness difference Δt ( = T
-T ') is detected, the previously stored correction amount is corrected based on the thickness difference Δt, and the pattern position is corrected based on the corrected correction amount.

Heretofore, the calculation of the correction amount used for converting the substrate 10 supported with the pattern forming surface facing upward to a state without bending has been described. Here, a method of converting into a state in which the pattern forming surface is supported downward (actual use state) will be further described.

The control device 20 uses the above-described proportional distribution from the fourth-order approximation formula stored in advance, and uses the proportional distribution described above to obtain the gradients θX1 ', θX at the position of the pattern in the state without any bending.
2 ′, θY1 ′, θY2 ′ are calculated, and the correction amount (t ·
θX1 ′) / 2, (t · θY1 ′) / 2, (t · θX
2 ′) / 2 and (t · θY2 ′) / 2 are calculated.
Using these correction amounts, the position 3 of the pattern in the second bending state, which occurs when the pattern forming surface is supported downward from the coordinate value corresponding to the position of the pattern in the state without bending, 3
The coordinate values corresponding to 3a and 33b are calculated. It should be noted that the line 32 in the X direction where the pattern is located in the state without any bending
The flexure shape in c is the pattern forming surface 3 in FIG.
It is assumed that the flexure shape of 3 is an arc shape with the point 0 ′ as the center.

The position of the pattern thus obtained is
It is close to the pattern position when the substrate 10 is supported with the pattern forming surface facing downward. The distance between the pattern positions 33a and 33b in the flexed state shown in FIG. 9A is larger than that in the case where the substrate 10 is in the ideal plane state.
This means that the error of t · (θX1-θX2) / 2 is included. However, as shown in FIG. 9, θX1 and θX2 are positive when the inclination of the substrate 10 rises to the right and negative when the inclination of the substrate 10 rises to the left. In this case, the distance between the pattern positions 33a and 33b is measured long if the gradient difference (θX1−θX2) is positive, and short if it is negative. Also,
Even if the substrate 10 is tilted with respect to the horizontal plane, the error is θX1.
Since it is calculated from the difference between θX2 and θX2, the inclination is canceled.

The distance between the positions 33a and 33b of the pattern in the bent state shown in FIG. 9B is t.multidot. (. Theta.X1 ') as compared with the position of the pattern in the state where the substrate 10 is not bent.
This means that it includes an error of −θX2 ′) / 2. However, θX1 ′ and θX2 ′ are as shown in FIG.
When the slope of is rising to the right, it is positive, and when it is rising to the left, it is negative.
In this case, the distance between the pattern positions 33a and 33b is measured longer if the gradient difference (θX1′−θX2 ′) is positive, and shorter if it is negative. Also,
Even if the substrate 10 is tilted with respect to the horizontal plane, the error is θX.
Since it is calculated from the difference between 1'and θX2 ', the inclination is canceled.

Therefore, the pattern position 3 in the second flexure shape when the substrate 10 is supported with the pattern formation surface facing down
The distance between 3a and 33b is {t (θX1-θX2)
/ 2}-{t (θX1'-θX2 ') / 2}. After that, by adding the thickness error Δt to the ideal thickness t of the substrate by the method described above, more accurate pattern position data based on the thickness t ′ actually possessed by the substrate 10 can be obtained.

Although the embodiment of the present invention has been described above, the present invention is not limited to this embodiment, and various design changes can be made within the scope of the idea of the invention indicated by the claims. Is possible. For example, the height of the surface of the substrate 10 may be detected by reading the vertical movement amount of the objective lens 11 by means of an encoder, an interferometer, a potentiometer, or the like, without using the autofocus device 12b. Further, not only the vertical movement amount of the objective lens 11 but also the vertical movement in the Z direction on the XY stage 15
A stage may be provided and the vertical movement amount of the Z stage may be read.

[0042]

As described above, according to the present invention, the thickness error of the substrate to be measured (manufacturing error, difference in standard thickness).
Therefore, there is an effect that the measurement accuracy is improved as compared with the case where the pattern position formed on the substrate is corrected only by the theoretical correction value by the reference substrate.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram showing a pattern measuring apparatus according to the present invention.

2A and 2B show arrangements of substrates used in the pattern measuring apparatus of the present invention, where FIG. 2A is a plan view and FIG. 2B is A in FIG.
It is a sectional view of the -A 'direction.

FIG. 3 is a block diagram showing a configuration of a control portion of the pattern measuring apparatus of the present invention.

FIG. 4 is an explanatory diagram for explaining the operation of the present invention.

FIG. 5 is a flowchart showing the operation of the present invention.

FIG. 6 is a graph showing the action of the present invention.

FIG. 7 is an explanatory view showing the operation of the present invention.

FIG. 8 is an explanatory view showing the operation of the present invention.

FIG. 9 is an explanatory view showing the operation of the present invention.

[Explanation of symbols]

10 ... Substrate 11 ... Objective lens 12 ... Optical device 12a ... laser light source 12b ... autofocus device 14a, 14b ... Interferometer 15 ... XY stage 20 ... Control device 21 ... Display device 22 ... Storage device 150 ... Drive device 50a, 50b, 51a, 51b ... Light receiving element

─────────────────────────────────────────────────── ─── Continued Front Page (58) Fields surveyed (Int.Cl. ⁷ , DB name) G01B 11/00-11/30 G01B 21/00 G06T 1/00-7/00

Claims (13)

(57) [Claims] 1. A pattern position measuring method for measuring the position of a pattern formed on a substrate, comprising: a first step of obtaining a displacement amount of the pattern due to bending of a predetermined reference substrate as a first correction value; A second step of measuring a thickness error amount of the substrate to be measured with respect to; and a third step of obtaining a second correction value by correcting the first correction value based on the thickness error amount; A fourth step of measuring the position of the pattern formed on the substrate to be measured; and a fifth step of correcting the position of the measured pattern based on the second correction value. Pattern position measurement method. 2. The pattern position measuring method according to claim 1, wherein in the second step, the thickness error amount is obtained by measuring a surface height of the substrate to be measured. 3. The pattern according to claim 2, wherein the substrate to be measured is placed on a plurality of supporting portions on a stage; and the surface height is measured in the vicinity of the supporting portions. Position measurement method. 4. The first correction value obtained in the first step is obtained based on two bending amounts when the reference substrate is supported on the front side and the back side, respectively. The described pattern position measuring method. 5. A method for measuring the position of a pattern of a substrate to be measured placed on a stage substantially horizontally, wherein the reference is determined based on a difference between the thickness of the substrate to be measured and a predetermined reference substrate. A first step of correcting the displacement amount of the pattern caused by the deflection of the substrate due to its own weight; a step of measuring the position of the pattern of the substrate to be measured, and correcting the measured pattern position based on the corrected displacement amount. A pattern position measuring method comprising: two steps. 6. A displacement gradient of the reference substrate is stored in advance, a displacement amount of the pattern is calculated based on the stored gradient at a measurement point of the pattern position, and the calculated displacement amount. The method according to claim 5, wherein is corrected with the thickness error. 7. A pattern position measuring device for measuring the position of a pattern formed on a substrate, and a storage means for storing a displacement amount of the pattern due to bending of a predetermined reference substrate as a first correction value; A thickness measuring means for measuring a thickness error amount of the substrate to be measured with respect to;
Calculating means for obtaining a second correction value by modifying the first correction value based on the thickness error amount; position measuring means for measuring the position of the pattern formed on the substrate to be measured. A pattern position measuring device comprising: a correction unit that corrects the position of the measured pattern based on the second correction value. 8. The pattern position according to claim 7, wherein the thickness measuring means measures a surface height of the substrate to be measured and detects the thickness error amount based on the measured value. measuring device. 9. The substrate to be measured is supported on a stage by being supported substantially horizontally by a plurality of supporting portions; and the thickness measuring means includes the surface near the supporting portion on which the substrate to be measured is mounted. The pattern position measuring device according to claim 8, wherein the height is measured. 10. The pattern position measuring device according to claim 8, wherein the thickness measuring means is an autofocus device. 11. An apparatus for measuring the position of a pattern of a substrate to be measured placed substantially horizontally on a stage, wherein a difference in thickness between the substrate to be measured placed on the stage and a predetermined reference substrate. Detecting means for detecting a pattern displacement amount, which is caused by the deflection of the reference substrate due to its own weight, in accordance with the detected difference in thickness, and the pattern position is measured based on the corrected displacement amount. A pattern position measuring device comprising: a correction unit that corrects a result. 12. The correction unit calculates a displacement amount of the pattern based on a storage unit that stores a gradient due to the bending of the reference substrate and the stored gradient at a measurement point at the pattern position, The apparatus according to claim 11, further comprising a calculation unit that corrects the calculated displacement amount with the thickness error. 13. A focus sensor for measuring the surface height of a substrate to be measured placed on the stage, wherein the detection means obtains the difference in thickness based on the output of the focus sensor. The device according to claim 11 or 12, characterized in that.