JPH05160000A - Exposure apparatus - Google Patents

Abstract

PURPOSE:To measure the error in a magnification or the distortion amount of a projection optical system by means of a simple constitution and always stably. CONSTITUTION:The title exposure apparatus is provided with the following: a plurality of slit patterns 20, 22 which have been formed in advance on a reticle R; and a photodetection face which has been formed on a wafer stage 10 so as to be nearly the same height as a substrate W. In addition, it is provided with the following: a photoelectric sensor 11 which detects the interval in a measuring direction (the X-direction) between images by means of a projection optical system PL of the slit patterns 20, 22; and a main control system 18 which finds the magnification and the distortion in the measuring direction of the projection optical system PL by means of the detected interval.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an exposure apparatus for transferring a pattern on a first surface onto a substrate on a second surface via a projection optical system. Such an exposure apparatus is used, for example, in a lithography process when manufacturing a semiconductor element.

[0002]

2. Description of the Related Art In a photolithography process for manufacturing a semiconductor element, a circuit pattern or the like formed on an original plate which is usually called a photomask or reticle (hereinafter referred to as "reticle") is coated on a semiconductor wafer. The resist layer thus formed is baked, and then photoetching is performed using the resist pattern formed by the development process as a mask. An exposure apparatus having a projection optical system with high resolution is used in the step of transferring the pattern of the reticle onto the resist layer on the wafer. In this type of exposure apparatus, it is required to accurately transfer the pattern of the reticle onto the wafer at a predetermined magnification or reduction ratio, and the imaging characteristics of the projection optical system, particularly magnification error or distortion. There is provided a mechanism for correcting In order to perform the correction, it is first necessary to accurately measure the error in magnification or the amount of distortion.

In order to perform such measurement in the conventional exposure apparatus, an optical member having a slit-shaped window is mounted on an XYZ stage which forms a specific pattern on a reticle and positions a wafer three-dimensionally. I put it. Then, the image of the specific pattern on the reticle by the projection optical system and the slit-shaped window are converted into an enlarged image of an appropriate size through the optical system, and the image of the pattern on the reticle is converted from this enlarged image. The distance between the slit-shaped window and the slit-shaped window is measured, and the error in the magnification of the projection optical system or the amount of distortion is obtained from the deviation of this distance from the design value.

[0004]

However, the conventional measuring method has a disadvantage that the XYZ stage and the measuring optical system are complicated. In addition, since the distance between the pattern on the reticle and the slit-shaped window on the XYZ stage is enlarged by the optical system, there is an error in the measurement result due to the influence of the aberration of this optical system and the fluctuation of the characteristics of the optical system. There was a risk of mixing. The present invention has been made in view of the above problems, and an object of the present invention is to provide an exposure apparatus having a simple configuration and capable of always stably measuring an error in magnification or a distortion amount of a projection optical system.

[0005]

An exposure apparatus according to the present invention is, for example, as shown in FIG. 1, an exposure apparatus for transferring a pattern on a first surface R onto a substrate W on a second surface via a projection optical system PL. In, a plurality of specific patterns (20, 22) formed in advance on the first surface R and a light receiving surface installed on the second surface at substantially the same height as the substrate W are provided. A predetermined measurement direction (for example, X direction) of the image of the plurality of specific patterns (20, 22) by the projection optical system PL.
There is provided a light receiving means (11) for detecting the interval and a computing means (18) for obtaining the magnification and distortion of the projection optical system PL in the predetermined measurement direction from the detected interval.

In this case, for example, as shown in FIG. 4, a pattern Q having a predetermined width g is added to the specific pattern (20, 22).
The light receiving means (11) including 1 detects the width of the image of the pattern Q1 having the predetermined width g, and the calculating means (18) detects the width of the detected image from the light receiving surface of the light receiving means (11). Alternatively, the focus position of the projection optical system PL in the optical axis direction may be obtained.

Further, as shown in FIG. 4, for example, a pair of lines arranged in the specific pattern (20, 22) obliquely with respect to the predetermined measurement direction (X direction) and non-parallel to each other. As shown in, for example, FIG. 5, the light receiving surface of the light receiving means (11) including the portions Q1 and Q2 has a pair of images (2) formed by the projection optical system PL of the pair of line segments Q1 and Q2.
0I) may have non-parallel regions. Further, for example, as shown in FIG. 11A, the specific pattern (20, 22) includes a line segment (32) arranged non-parallel to the predetermined measurement direction (X direction), and The light receiving surface of the light receiving means (11) is, for example, as shown in FIG.
As shown in (b), the line segment (30) has a pair of regions (33, 34) arranged obliquely to the image formed by the projection optical system PL and non-parallel to each other. Good.

[0008]

According to the present invention, the intervals in the X direction, which is the measurement direction, of the plurality of specific patterns (20, 21) formed on the first surface R are measured in advance, and the second surface is measured. The X-direction interval between the images of the plurality of specific patterns projected by the projection optical system PL is detected on the light receiving surface that is installed on the upper surface at substantially the same height as the substrate W. The magnification error in the X direction of the projection optical system can be obtained from the detected interval and the original interval (the measured value or the design value). Further, as shown in FIG. 4, for example, the plurality of specific patterns are two patterns (19A, 19A) arranged in a direction perpendicular to the X direction.
In the case of B), the distortion of the projection optical system can be detected by detecting the positional shift of the images of these two patterns on the second surface in the X direction.

In this case, since the images of the specific patterns (20, 21) are directly detected by the light receiving surface of the light receiving means (11), the influence of aberrations of the optical system other than the projection optical system is affected. In addition, it has a simple structure such as the second surface.
Further, a pattern Q1 having a predetermined width g is included in the specific pattern (20, 22), and the projection optical system P of this pattern Q1 is included.
When the image of L is projected onto the second surface, the light receiving means (1
When the light receiving surface of 1) is focused on the projection optical system PL, the width of the image becomes the narrowest. By using this, the in-focus position of the light receiving surface in the optical axis direction of the projection optical system PL can be obtained.

Next, the specific pattern (20, 22)
, A pair of line segments Q1 and Q2 arranged obliquely with respect to the predetermined measurement direction (X direction) and non-parallel to each other.
Including the case where the light receiving surface of the light receiving means (11) has regions that are respectively non-parallel to the pair of images (20I) of the pair of line segments Q1 and Q2 formed by the projection optical system PL. The action and effect of will be described. Projection optical system PL
Assuming that the magnification of is equal to the design value, the image (20I) of the pair of line segments Q1 and Q2 and the light receiving surface of the light receiving means (11) are formed on the second surface, for example, as shown in FIG. It intersects at points P1 and P2. However, when the magnification deviates from the design value, the image (20I) moves to the position (20J) which is deviated by Δx in the measurement direction, and the image at the position (20J) and the light receiving surface of the light receiving means (20) are separated. The intersection moves to points R1 and R2.

The distance between intersection points P1 and P2 is y1, and the intersection point R1
Letting the distance between R2 and R2 be y2, Δy (= y1-y2) is measured on the light receiving means (20) side. And the image (2
The angle that one side of (0I) makes with the measurement direction (X direction) is θ
Then, there is a relationship of Δy = 2 · Δx · tan θ. Therefore, if θ> 27 ° holds, Δy
> Δx is established. This is Δx which is the actual displacement
It means that it can be magnified and observed. Also, the image (20
If I) is oblique to the measuring direction, the light receiving means (1
The displacement within a certain range can be measured at one place without scanning the light receiving surface of 1) in the measurement direction.

Next, the specific pattern (20, 22)
Including a line segment (32) arranged non-parallel to the predetermined measurement direction (X direction), the light receiving surface of the light receiving means (11) is an image of the line segment (32) by the projection optical system PL. The case of having a pair of regions (33, 34) arranged obliquely with respect to each other and non-parallel to each other reverses the roles of the specific pattern and the light receiving surface of the light receiving means to the above invention. It was done. Even in this case, the actual displacement can be magnified and observed, and the displacement within a certain range can be measured at one place.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an exposure apparatus according to the present invention will be described below with reference to the drawings. In this example, the present invention is applied to a stepper that prints a pattern on a reticle on a wafer by a step-and-repeat method.

FIG. 1 shows the overall structure of the stepper of this embodiment. In FIG. 1, the g-line or i from the mercury lamp 1 is used.
The exposure light such as a line is condensed by the elliptical mirror 2 and reflected by the mirror 3, and then passes through a shutter 4 for controlling the exposure amount and an optical integrator 5 for uniformizing the illumination light, and then a mirror 6.
Incident on. The exposure light reflected by the mirror 6 is converted into a substantially parallel light flux by the main condenser lens 7 to illuminate the reticle R with a substantially uniform illuminance. PL represents a projection optical system of both sides (or one side) telecentric, and if the orthogonal coordinates in a plane perpendicular to the optical axis of the projection optical system PL are represented by the X axis and the Y axis, fine movement in the X axis and the Y axis directions is performed. I can and X
The reticle R is held on the reticle stage 8 which can be minutely rotated in the Y plane, and the reticle alignment system 9 positions the reticle R with respect to the device (the optical axis of the projection optical system PL).

The reticle R of this example is a test reticle for measuring the error in magnification and the amount of distortion of the projection optical system PL, and the pattern area PA on the back surface of the reticle R is
Triangular wave slit patterns 20 and 22 extending in the Y direction perpendicular to the paper surface of FIG. 1 described later are formed. These slit patterns 20 and 22 are patterns for measuring magnification errors in the X direction, and although not shown, slit patterns for measuring magnification errors in the Y direction are also formed on the reticle R.

Reference numeral 10 denotes a wafer stage, and W denotes a wafer. The wafer stage 10 is an XY stage which moves in the XY plane, a Z stage which moves in the Z axis direction which is the optical axis direction of the projection optical system PL, and The wafer W is mounted on the wafer stage 10 which is composed of a θ stage or the like that makes a minute rotation in the XY plane. Reference numeral 11 denotes a photoelectric sensor that is a charge coupled (CCD) line sensor that is long in the Y direction, and the light receiving surface of the photoelectric sensor 11 and the exposure surface of the wafer W have the same height in the optical axis direction of the projection optical system PL. Thus, the photoelectric sensor 11 is fixed on the wafer stage 10 next to the wafer W. The wafer stage 10 is driven in the optical axis direction so that the image forming surface of the projection optical system PL and the light receiving surface of the photoelectric sensor 11 coincide with each other when measuring the magnification and the like. Then, the projection optical system PL of the slit pattern 20 or 22 of the reticle R
Image is projected onto the light receiving surface of the photoelectric sensor 11.
The photoelectric sensor 11 is a sensor for measuring a magnification error or the like in the X direction, and a photoelectric sensor (not shown) for measuring the magnification error or the like in the Y direction is provided so as to be orthogonal to the photoelectric sensor 11. ing.

The photoelectric conversion signals output from the series of light receiving elements of the photoelectric sensor 11 are amplified by the amplifier 13 to obtain the mark signal S, and the mark signal S is supplied to the signal processing unit 14. The signal processing unit 14 processes the mark signal S to obtain the coordinates in the Y direction of the intersection of the light receiving surface of the photoelectric sensor 11 and the image of the slit pattern 20 or 22 on the reticle R. Further, the signal processing unit 14 sends a command to the stage controller 15 to move the wafer stage 10 in a predetermined measurement direction. Reference numeral 16 denotes an XYθ-axis drive system, 17 denotes a Z-axis drive system, the XYθ-axis drive system 16 drives the XY stage and the θ stage of the wafer stage 10, and the Z-axis drive system 17 drives the Z stage of the wafer stage 10. .. Further, the XYθ-axis drive system 16 is provided with two sets of laser interferometers for measuring the position of the wafer stage 10 in the X-axis and Y-directions, and the Z-axis drive system 17 is provided with the Z-axis of the wafer stage 10. An encoder is arranged to measure the directional position. Stage controller 15
Recognizes the coordinates on the wafer stage 10 corresponding to, for example, the optical axis of the projection optical system PL and the height of the wafer stage 10 with respect to the projection optical system PL from the position information from the drive systems 16 and 17, and the coordinates and the height. The wafer stage 10 is positioned via the drive systems 16 and 17 such that

Reference numeral 18 denotes a main control system for controlling the entire operation, and the main control system 18 is supplied with the coordinates of the intersections obtained by the signal processing unit 14 and the position information of the wafer stage 10 obtained by the stage controller 15. To do. The main control system 18 comprehensively controls the imaging characteristics of the projection optical system PL, especially the measurement sequence of magnification error and distortion amount,
At the time of exposure, the opening / closing timing and opening time of the shutter 4 are controlled via a shutter controller (not shown), and finally the magnification error and distortion amount of the projection optical system PL are obtained. Further, although omitted in FIG. 1, the projection optical system PL of this example is divided into a plurality of lens groups, and by adjusting the interval between predetermined adjacent lens groups and the inclination amount of the predetermined lens group with respect to the optical axis. , The imaging characteristics can be changed. Then, the main control system 18 determines the required projection optical system P.
The distance between the lens groups of the projection optical system PL or the inclination amount of the lens groups is adjusted so that the magnification error of L or the like falls within the allowable value. Further, the imaging characteristic of the projection optical system PL can be changed by sealing the space between the predetermined adjacent lens groups of the projection optical system PL and changing the pressure of gas (air or the like) in the closed space. You can The reticle is projected onto the projection optical system P.
By moving in the optical axis direction of L or tilting with respect to the XY plane, it is configured to correct the imaging characteristics of the projection optical system PL, particularly the distortion aberration (barrel type, pincushion type distortion). Alternatively, in the present embodiment, the mechanism for correcting the image forming characteristic may have any configuration.

Next, the pattern arrangement of the reticle R of this example will be described with reference to FIG. In FIG. 2, the reticle R
Cross-shaped reticle alignment marks RM1, RM2, and RM3 are formed in the vicinity of the three sides, and the reticle R is positioned by detecting those marks RM1 to RM3 by the reticle alignment system 9 in FIG. Reticle R based on those three marks RM1 to RM3
A rectangular pattern area PA is formed at the center of the. And
In this example, a triangular wave slit pattern 20 having a structure in which the pattern unit 19A having a length of 2H in the Y direction is repeated in the Y direction is formed at the right end in the X direction of the pattern area PA, and the pattern is formed at the left end of the pattern area PA. Unit 21A
Is formed in the Y direction to form a triangular wave slit pattern 22. The slit patterns 20 and 22 are line-symmetric with respect to a straight line passing through the center of the reticle R and parallel to the Y axis. The distance between the center line of the slit pattern 20 and the center line of the slit pattern 22 in the X direction is set to a predetermined value and stored in the memory in the main control system 18. A set of slit patterns 2 whose longitudinal direction is the Y direction
0 and 22 measure the magnification error of the projection optical system PL in the X direction and the like.

Similarly, at the upper end of the pattern area PA in the Y direction, the pattern unit 23A having a length of 2H in the X direction is arranged in the X direction.
Triangular wave slit pattern 2 with repeated structure
4 is formed, and a triangular wave slit pattern 26 having a structure in which the pattern unit 25A is repeated in the X direction is formed at the lower end of the pattern area PA. Slit patterns 24 and 2
6 is line symmetric with respect to a straight line passing through the center of the reticle R and parallel to the X axis. Further, the distance between the center line of the slit pattern 24 and the center line of the slit pattern 26 in the Y direction is also set to a predetermined value and stored in the memory in the main control system 18. A pair of slit patterns 24 and 26 having the X direction as the longitudinal direction measures a magnification error of the projection optical system PL in the Y direction and the like. The shape of the slit pattern 20 and the like is not limited to the triangular wave shape as described later, and an edge pattern or the like can be used in addition to the slit pattern 20 and the like.

FIG. 3 shows a detailed structure of the photoelectric sensor 11 shown in FIG. As shown in FIG. 3, the photoelectric sensor 11 includes light receiving elements D1 to Dn arranged in the Y direction. The length of each light receiving element D1 in the Y direction, that is, the position resolution in the Y direction is, for example, several μm, and the entire length in the Y direction is projected on the length of the slit pattern 20 on the reticle R in the Y direction. It is about the length multiplied by the magnification (or reduction rate) of the optical system PL. However, when the method of moving the wafer stage 10 in the Y direction is used, the length of the photoelectric sensor 11 in the Y direction is determined by the Y of the pattern unit 19A on the reticle R.
It may be about the length obtained by multiplying the length 2H in the direction by the magnification of the projection optical system PL. Further, a photoelectric sensor 27 including light receiving elements E1 to En arranged in the X direction is arranged so as to be orthogonal to the photoelectric sensor 11. However, photoelectric sensor 1
Each of 1 and 27 may be composed of a two-dimensional image sensor or the like.

FIG. 4 shows the detailed shapes of the pattern unit 19A on the reticle R of FIG. 2 and the same pattern unit 19B adjacent thereto. In the pattern units 19A and 19B of FIG. 4, a light-shielding pattern 28 having a triangular wave shape at the left end and a light-shielding pattern 29 having a triangular wave shape at the right end are arranged in the X direction with a predetermined separation. Usually, it is formed by removing the chromium in the portion corresponding to the pattern unit 19 by etching when manufacturing the reticle. As a result, those light shielding patterns 28
The light transmitting triangular slit pattern 20 is formed between the light shielding pattern 29 and the light shielding pattern 29. Further, in the pattern unit 19A, the slit pattern 20 has an angle θ (0 ° ≦ θ) in the clockwise direction with respect to the X direction which is the measurement direction.
Slit Q1 with width g intersecting at <90 °) and slit Q1 with width g intersecting counterclockwise with respect to the X direction at angle θ
It is configured by connecting 2 and. The lengths of the slits Q1 and Q2 in the Y direction orthogonal to the X direction are both H.

If the reticle R of this example is used, it is possible to measure not only the X-direction but also the Y-direction magnification error of the projection optical system PL, etc. Will be described.

FIG. 5A shows the slit pattern 2 of FIG.
0 is projected onto the vicinity of the photoelectric sensor 11 on the wafer stage 10 via the projection optical system PL, and the slit pattern image 20I with high illuminance is the image of the slit pattern 20. In this case, when the mark signals S obtained from the series of light receiving elements of the photoelectric sensor 11 are plotted corresponding to the position in the Y direction, as shown in FIG.
Among them, only the value of the mark signal S obtained from the light receiving element at the portion intersecting the slit pattern image 20I becomes large.
Then, Y of two adjacent peak points of the mark signal S
The interval y in the direction can be regarded as the interval in the Y direction between adjacent intersections of the slit pattern image 20I and the photoelectric sensor 11. As described above, according to this example, it is possible to more easily measure the interval in the Y direction at the intersection of the slit pattern image 20I and the photoelectric sensor 11 than the mark signal.

Next, with reference to FIG. 6, a change in the measurement result when the relative position between the slit pattern image 20I and the photoelectric sensor 11 changes due to a magnification error of the projection optical system PL or the like will be described. In FIG. 6A, the coordinate of the photoelectric sensor 11 in the X direction is x1, and the slit pattern image when the magnification of the projection optical system PL is as designed is 20I. As shown in FIG. 6B, the slit pattern image 20
Since the mark signal S has a peak at the intersection of I and the photoelectric sensor 11, the position of the intersection in the Y direction can be detected. Then, the interval in the Y direction between the adjacent intersections P1 and P2 between the slit pattern image 20I and the photoelectric sensor 11 is defined as y1. When the X coordinate of the center line of the slit pattern image 20I is x1, the interval y1 is equal to Hβ, where β is the magnification of the projection optical system PL. On the other hand, when the magnification of the projection optical system PL changes, the slit pattern image 2
It is assumed that the image 20J is formed by offsetting 0I by Δx in the X-axis direction. The distance in the Y direction between adjacent intersections R1 and R2 between the misaligned image 20J and the photoelectric sensor 11 is y2.
Then, in this example, the interval y1 and the interval y2 can be measured.

Then, the measured positional deviation in the Y direction is Δ
If y, then the following equation holds.

In this case, the angle formed by the images 20I and 20J with the X direction is θ.
Therefore, the actual positional deviation Δx in the X direction can be obtained from the measured positional deviation Δy using the following equation.

[Expression 2] Δx = (Δy / 2) / tan θ

According to this (Equation 2), if θ> 27 °, Δx <Δy holds. This is because if the angle θ formed by the slit pattern 20 of FIG. 4 with respect to the X direction, which is the measurement direction, is set to be larger than 27 °, the positional deviation Δx of the slit pattern image in the X direction due to a magnification error is expanded to Δy. Means that it can be measured. For example, if θ = 79 ° is set, Δy = 10Δx, and the positional deviation Δx
Can be observed up to about 10 times. Therefore, according to this example, it is possible to directly magnify and observe the relative positional relationship between the photoelectric sensor 11 and the image 20I of the slit pattern as the reference mark without using the magnifying optical system. There is.

However, since only the interval y2 in FIG. 6 (a) is actually measured for a certain projection optical system PL, as y1 in the above (Equation 1), the center X coordinate is x1. It is also conceivable to use Hβ, which is the value of the ideal slit pattern image that is assumed to match. As a result, Δy can be regarded as a displacement in the X direction from the ideal state.

In addition, the magnification of the projection optical system PL in the Y direction may deviate from β, but the local error can be detected as follows. That is, FIG.
In (a), the misaligned slit pattern image 20
If the third intersection point following the intersection points R1 and R2 between J and the photoelectric sensor 11 is R3, the distance between the intersection points R1 and R3 is Y.
It is 2H · β in the state where there is no magnification error in the direction. On the other hand, if the actually measured interval is y3, the local magnification error Δβ _{Y in the} Y direction can be obtained by the following equation.

[Mathematical formula-see original document] [Delta] [beta] _Y = (y3-2H. [Beta]) / (2H. [Beta]) By correcting the [Delta] _y obtained from the above (Formula 1) in consideration of the magnification error thus obtained, The position shift in the X direction can be accurately obtained.

Next, with reference to FIG. 7, an example of a method for obtaining a magnification error in the X direction of the projection optical system PL from the positional deviation Δx in the X direction obtained from (Equation 2) will be described. For convenience of explanation, it is assumed that there is no magnification error in the Y direction. In FIGS. 7A and 7B, images 20I and 22I indicated by solid lines are slit patterns 2 on the reticle of FIG. 2 when the magnification of the projection optical system PL in the X direction is equal to the design value (this is β).
It is a virtual projection image of 0 and 22. Further, the images 20J and 22J indicated by the two-dot chain line are assumed to be the actual projected images of the slit patterns 20 and 22 by the projection optical system PL, respectively. In this case, the main control system 18 of FIG. 1 moves the wafer stage 10 in the X direction via the stage controller 15 so that the X coordinate of the photoelectric sensor 11 is x3.
And x4 are measured at two locations. The photoelectric sensor 11 whose X coordinate is x4 is represented by a photoelectric sensor 11A.

Slit pattern 2 on reticle R in FIG.
When the distance between the center line of 0 and the center line of the slit pattern 22 is LX, the distance between the X coordinates x3 and x4 is selected to be LX · β. In this case, the X coordinates of the center lines of the virtual slit pattern images 20I and 22I can be regarded as x3 and x4, respectively. Then, in the vicinity of the coordinate in the Y direction of FIG. 7 near YA, the distance y4 between the two intersections of the photoelectric sensor 11 and the actual slit pattern image 20J and the two intervals of the photoelectric sensor 11A and the actual slit pattern image 22J. The intersection interval y5 is measured. Also,
The interval between two intersections of the virtual slit pattern image 20I and the photoelectric sensor 11 and the virtual slit pattern image 2
The distance between the two intersections of 2I and the photoelectric sensor 11A is H
・ Β.

Then, by substituting Δy obtained by substituting the interval y4 and H · β into (Equation 1) into (Equation 2), the positional deviation Δx1 in the X direction of the slit pattern image 20J is obtained, and the interval y5 Δ obtained by substituting H and β into (Equation 1)
By substituting y into (Equation 2), the positional deviation Δx4 of the slit pattern image 22J in the X direction can be obtained. Therefore, assuming that the error of the magnification in the X direction near the Y coordinate of YA is Δβ _X1 , Δβ _X1 is as follows.

## EQU4 ## Δβ _X1 = (LX · β + Δx1 + Δx4) / LX-β = (Δx1 + Δx4) / LX

Similarly, even when the Y coordinate is near YB,
The positional deviation Δx2 of the actual slit pattern image 20J and the positional deviation Δx5 of the slit pattern image 22J are obtained, and even when the Y coordinate is in the vicinity of YC, the positional deviation Δx3 of the actual slit pattern image 20J and the positional deviation of the slit pattern image 22J. Δx6 is obtained. Then, by the same calculation as in (Equation 4), the Y coordinates are YB and Y, respectively.
Magnification errors Δβ _X2 and Δβ _X3 in the X direction near C
Is required. Further, for example, the difference between the magnification errors Δβ _X1 and Δβ _X2 and the difference between the magnification errors Δβ _X2 and Δβ _X3 can be regarded as local distortions of the projection optical system PL.

As described above, according to this example, the photoelectric sensor 1
1 is moved to two positions of coordinates x3 and x4 in the X direction,
The magnification error in the X direction and the amount of distortion of the projection optical system PL can be quickly measured only by measuring the coordinates of the intersections with the slit pattern images 20J and 22J. Therefore, since it is not necessary to scan the photoelectric sensor 11 in the X direction, the measurement time can be significantly shortened. Further, the angle θ formed by the slit patterns 20 and 22 on the reticle R in the X direction
, The slit pattern image 2
The positional deviation in the X direction between 0J and 22J can be enlarged and measured, and the magnification error and the like can be measured with higher accuracy. In this example, since the photoelectric sensor 11 has one light-receiving portion in the X direction, the slit patterns 20 and 22 are not provided.
In order to detect the image of the
It is necessary to move to a location (coordinates x3 and X4). However, by providing two light receiving portions of the photoelectric sensor 11 in the X direction, the images of the slit patterns 20 and 22 can be detected at one location in the X direction. That is, in the former case, the position data of the interferometer is required, but in the latter case, the measurement is performed on the basis of the pixel of the photoelectric sensor 11.

On the reticle R of FIG. 2, for example, the slit pattern 20 is formed by repeating the pattern unit 19A many times, and the slit pattern 22 is formed by repeating the pattern unit 21A many times. However, even if only the pattern unit 19A and the pattern unit 21A are formed on the reticle R, the magnification error of the projection optical system PL in the vicinity of the Y coordinate of YA in FIG. 7 can be measured. Further, when only the pattern units 19A and 19B of FIG. 2 are formed on the reticle R, the positional deviations Δx4 and Δx5 of FIG. 7 can be measured, and therefore the distortion of the projection optical system PL in the vicinity thereof can be determined from the difference between them. Can be measured. Therefore,
On the reticle R of FIG. 2, the magnification error of the projection optical system PL can be measured to some extent only by the existence of a plurality of the same patterns as the pattern unit 19A regardless of the arrangement.

Next, using the slit pattern 20 on the reticle R of FIG. 2, the wafer W for the projection optical system PL is used.
The focusing can be performed, which will be described with reference to FIGS. 8 and 9. FIG. 8A shows the photoelectric sensor 11
1 shows a state in which the light receiving surface of is aligned with the image forming surface of the projection optical system PL in FIG. 1, that is, a state in which the light receiving surface is focused on the projection optical system PL. In this case, the line width of the image 20I of the slit pattern 20 on the reticle R is the narrowest.
The mark signal S obtained from the photoelectric sensor 11 is shown in FIG.
As shown in (b), it has a pulse shape in the width y6 in the Y direction of the intersection region between the slit pattern image 20I and the photoelectric sensor 11.

On the other hand, when the light receiving surface of the photoelectric sensor 11 is displaced upward or downward from the in-focus position with respect to the projection optical system PL, as shown in FIG. 9A, the photoelectric sensor of the slit pattern 20 of the reticle R is formed. The width of the image 20I on 11 becomes wider. Accordingly, as shown in FIG. 9B, the width y7 of the pulse portion of the mark signal S obtained from the photoelectric sensor 11 also becomes wider than the width y6 of FIG. 8B. Therefore, the wafer stage 10 is arranged so that the width in the Y direction of the intersecting region between the photoelectric sensor 11 and the slit pattern image 20I is minimized.
The light receiving surface of the photoelectric sensor 11 can be focused on the projection optical system PL by adjusting the height of the.

In this case, if the width of the slit pattern 20 on the reticle R in the Y direction is g _Y and the magnification of the projection optical system PL is β, the Y direction most obtained from the mark signal S in FIG. 8B is obtained. The narrow pulse width y6 is about g _Y · β. Further, by moving the wafer stage 10 up and down from the in-focus state in advance and measuring the variation amount of the pulse width thereof, conversely, from the current pulse width (for example, the width y7 in FIG. 9B), the focus is changed. It is also possible to calculate the amount of positional deviation between the current Z-axis coordinate and. Then, the focusing speed can be increased by moving the wafer stage 10 in the Z direction by the amount of the positional deviation. By using the slit pattern 20 and the like of this example, it is possible to focus on the projection optical system PL in addition to the magnification error of the projection optical system PL. In this case, focusing is directly performed from an image obtained by projecting the pattern on the reticle onto the light receiving surface of the photoelectric sensor 11 via the projection optical system PL, so that it may change with time or due to exposure light absorption or the like. There is an advantage that the true focusing point (best focus position) can be accurately detected even when the focusing characteristics (best focusing surface) are displaced in the Z direction due to changes in the imaging characteristics of the projection optical system PL. At this time, the projection optical system P
If the best focus position at each of a plurality of points in the L image field is obtained, the best image plane of the projection optical system PL can also be obtained.

Although not shown in FIG. 1, FIG.
Such an exposure apparatus can be provided with an indirect focusing mechanism. Unlike the direct method in which the actual image plane of the projection optical system PL is detected and the exposure surface of the wafer W is aligned with this image plane, in the indirect method, the coordinates in the Z direction of the image plane are obtained in advance by the direct method. After that, the wafer stage 1 is adjusted so that the Z-direction coordinates of the exposure surface of the wafer W become the obtained coordinates.
Controls the height of 0. When the wafer W is in the image field of the projection optical system PL, focusing by the direct method is relatively difficult, and thus focusing by the indirect method is performed. As an indirect optical system, for example, a projection optical system P
A light beam that is obliquely incident on the optical axis of L and that is obliquely reflected from the wafer W is received by a position detection type light receiving element (P
An optical system that receives light by SD etc. is used.

When performing such indirect focusing, it is necessary to calibrate the position of the image plane of the projection optical system PL by the direct method. Since the focusing method using the slit pattern 20 of this example is a direct method,
It is suitable for regularly performing the calibration of the focus of such an indirect focusing mechanism. An example of the configuration of the indirect focus mechanism and its calibration operation is disclosed in, for example, Japanese Patent Laid-Open No. 60-168112, and therefore the description thereof is omitted here.

Next, another embodiment of the present invention will be described with reference to FIG. FIG. 10A shows a pattern on the reticle R of this embodiment. In FIG.
Reference numeral 8 is a light-shielding pattern in which the right edge is triangular. A triangular wave-shaped edge pattern 29 is formed by repeating a pair of adjacent edge patterns Q3 and Q4 at the right end of the light shielding pattern 28 in the Y direction. By projecting the image of the edge pattern 29 by the projection optical system PL on the light receiving surface of the photoelectric sensor 11 of FIG. 3 and detecting the Y coordinate of the boundary between the bright portion and the dark portion of the edge pattern image, the width of FIG. The widths corresponding to y1 and y2 are determined. Therefore, a magnification error in the X direction of the projection optical system PL can be measured.

In the example of FIG. 10A, the adjacent edge patterns Q3 and Q4 are both inclined with respect to the X axis, but one edge pattern may be parallel with the X axis, for example. .. FIG. 10B shows a pattern on the reticle R in such a case, and in FIG.
0 is a light-shielding pattern in which a right triangle is repeated in the Y direction. One pattern Q5 of a pair of adjacent edge patterns of the light shielding pattern 30 is parallel to the X axis, and the other pattern Q6 intersects the X axis at an angle θ, and the pair of edge patterns Q5 and Q6. Is repeated in the Y direction to form a sawtooth edge pattern 31. By obtaining the coordinate in the Y direction of the boundary between the bright portion and the dark portion of the image of the edge pattern 31, the magnification error in the X direction of the projection optical system PL can be similarly measured.

In the example of FIG. 10B, the relationship between the positional deviation Δx in the X direction and the actual positional deviation Δy in the Y direction is as follows.

## EQU00005 ## .DELTA.x = .DELTA.y / tan .theta. When this is compared with (Equation 2), the detection sensitivity is 1/2 in this embodiment, but if .theta.> 45.degree., .DELTA.x <
Δy. Therefore, by setting the crossing angle of the other edge pattern Q6 with respect to the X axis to 45 ° or more, the positional deviation can be enlarged and measured.

In the above-described embodiment, the pattern on the reticle R has a triangular wave shape or a sawtooth shape, and the light receiving surface of the photoelectric sensor 11 has a linear shape. However, it is obvious that the opposite relationship may be possible. In the case of the opposite relationship, a pattern having the X direction on the reticle R as the measurement direction is a straight line substantially perpendicular to the light transmissive X direction between the light shielding patterns as shown in FIG. 11A. The pattern 32 is formed. On the other hand, the shape of the light-receiving surface of the photoelectric sensor mounted on the wafer stage 10 is, for example, as shown in FIG. 11B, the line sensor 33 and the Z sensor that intersect in the X direction clockwise at an angle θ.
Line sensor 34 that crosses the direction counterclockwise at an angle θ
The basic unit consisting of is repeated in the Y direction. It is obvious that the magnification error in the X direction of the projection optical system PL can be measured even with such a combination.

Further, the pattern on the reticle R is a triangular wave slit pattern in the Y direction as shown in FIG. 4, and the shape of the light receiving surface of the photoelectric sensor is also a triangular wave pattern in the Y direction as shown in FIG. 11B. It can also be In this case, the phase difference between the image of the slit pattern on the reticle R and the pattern of the light receiving surface of the photoelectric sensor is set to 180 °. As a result, the above-mentioned relational expression of (Equation 2) becomes as follows, and the positional deviation in the X direction can be further doubled and measured.

(6) Δx = (Δy / 4) / tan θ

In the above embodiment, the case where the image forming characteristics (magnification, distortion, focus position, etc.) of the projection optical system applied to the exposure apparatus are obtained has been described, but it is also applicable to projection optical systems other than the exposure apparatus. The same effect can be obtained by applying the present invention. That is, the projection optical system to be inspected may be applied to any device, and the type thereof is a refraction optical system including a plurality of lens elements, a reflection optical system including a mirror, or a combination thereof. It may be a system.

Although the test reticle is used in the embodiment, a slit pattern may be formed on a part of the device reticle and the measurement may be performed in the same manner. This is particularly effective for a multi-die reticle (one in which a plurality of circuit patterns are drawn in the pattern area of the reticle). Further, for example, the unit pattern 19A of FIG. 2 is provided at each grid point obtained by dividing the pattern area of the reticle into a grid at a fine pitch.
Alternatively, a set of slit patterns of 25A and 25A may be formed. This makes it possible to measure the magnification in the X and Y directions at almost the same position. At this time, it is desirable to form a plurality of light receiving surfaces on the sensor 11 at corresponding positions. This is because there is no need to start and scan, and the measurement time can be shortened. Further, the slits on the sensor 11 may be a set of T-shaped patterns which are orthogonal to each other, but in this case, it is necessary to move the sensor 11 within the image field of the projection optical system. Furthermore,
A slit pattern composed of, for example, the unit patterns 19A, 25A, 21A, and 23A of FIG. 2 may be formed at each grid point obtained by dividing the reticle into a grid shape. This improves the measurement accuracy.

Further, when detecting the focal position of the projection optical system, marks may be arranged so that the measurement direction coincides with the sagittal (S) or meridional (M) direction in the image field. As a result, astigmatism can be obtained. Further, the sensor 11 may be other than the CCD. As described above, the present invention is not limited to the above-described embodiments, and various configurations can be taken without departing from the gist of the present invention.

[0049]

According to the present invention, the intervals between a plurality of specific patterns on the first surface are directly detected by the light receiving surface set at substantially the same height as the substrate during measurement.
There is an advantage that the magnification error or the amount of distortion of the projection optical system can be measured accurately and stably. Furthermore, since no magnifying optical system is used, the structure is simple and the maintenance is excellent. When the specific pattern includes a pattern of a predetermined width, the position of the second surface is adjusted so that the image width of the pattern of the predetermined width is the narrowest, so that the projection optical system is adjusted. Direct focusing can also be performed.

Further, when the specific pattern includes a pair of line segments which are arranged obliquely and non-parallel to each other with respect to a predetermined measuring direction, the light receiving surface of the light receiving means is arranged in the measuring direction. The positional shift amount of the image of the pattern can be measured only by positioning at one point. When positioning and measuring at one point in this way, the measurement time can be greatly shortened compared to the scanning method, and an average time is obtained using a charge storage type imaging device (CCD). A stable and stable signal can be obtained. Furthermore, by increasing the inclination angle of the line segment with respect to the measurement direction, it is possible to expand and measure the positional deviation in the measurement direction.
On the contrary, even when the light receiving surface of the light receiving means includes a pair of regions arranged obliquely and non-parallel to each other with respect to the predetermined measurement direction, the light receiving surface of the light receiving means is set to 1
The positional deviation amount of the image of the specific pattern can be measured only by positioning at the point. Furthermore, by increasing the inclination angle of the pair of regions with respect to the measurement direction, it is possible to expand and measure the positional deviation in the measurement direction.

[Brief description of drawings]

FIG. 1 is a configuration diagram including a partial cross-sectional view showing an embodiment of an exposure apparatus according to the present invention.

FIG. 2 is a plan view showing a pattern of a reticle R of the embodiment.

FIG. 3 is a plan view showing a photoelectric sensor 11 and the like shown in FIG.

FIG. 4 is an enlarged plan view showing a detailed configuration of a part of the slit pattern 20 of FIG.

5A is an enlarged plan view showing an image of a slit pattern formed on the photoelectric sensor 11, and FIG. 5B is a waveform diagram showing a mark signal S output from the photoelectric sensor 11 correspondingly. ..

6A is a plan view showing a case where the slit pattern image 20I is displaced with respect to the photoelectric sensor 11; FIG.
FIG. 6 is a waveform diagram showing a change in the mark signal S corresponding to the above.

FIG. 7 is a diagram for explaining how to obtain a magnification error or a distortion amount of a projection optical system.

8A is an enlarged plan view showing a slit pattern image 20I when the light receiving surface of the photoelectric sensor 11 is focused on the projection optical system, and FIG. 8B is a waveform diagram showing a corresponding mark signal S. Is.

9A is a slit pattern image 20 when the light receiving surface of the photoelectric sensor 11 is out of the focusing surface of the projection optical system. FIG.
FIG. 3B is an enlarged plan view showing I, and FIG. 3B is a waveform diagram showing the corresponding mark signal S.

10A is an enlarged plan view of an essential part showing a pattern on a reticle R of another embodiment, and FIG. 10B is an enlarged plan view of an essential part showing a modified example of the pattern of FIG. 10A.

11A is an enlarged plan view showing a part of a pattern on a reticle R according to still another embodiment of the present invention, and FIG. 11B is a part of the shape of the light receiving surface of the photoelectric sensor according to the embodiment. It is an enlarged plan view shown.

[Explanation of symbols]

 R reticle PL Projection optical system 10 Wafer stage 11 Photoelectric sensor 14 Signal processing unit 15 Stage controller 16 XYθ axis drive system 17 Z axis drive system 18 Main control system 19A, 21A Pattern unit 20, 22 Slit pattern 20I, 20J Slit pattern 20 Image 22I, 22J Image of slit pattern 22

Claims (4)

[Claims] 1. An exposure apparatus for transferring a pattern on a first surface onto a substrate on a second surface via a projection optical system, comprising: a plurality of specific patterns previously formed on the first surface; Light-receiving means having a light-receiving surface installed on the second surface at substantially the same height as the substrate, and detecting a gap in a predetermined measurement direction between the images of the plurality of specific patterns formed by the projection optical system; An exposure apparatus comprising: arithmetic means for obtaining a magnification and a distortion of the projection optical system in the predetermined measurement direction from the detected interval. 2. The specific pattern includes a pattern having a predetermined width, the light receiving means detects the width of the image of the pattern having the predetermined width, and the computing means detects the width of the image based on the width of the detected image. The exposure apparatus according to claim 1, wherein a focus position of a light-receiving surface in the optical axis direction of the projection optical system is obtained. 3. The specific pattern includes a pair of line segments arranged obliquely to each other with respect to the predetermined measurement direction and non-parallel to each other, and a light receiving surface of the light receiving means has a pair of line segments. 3. The exposure apparatus according to claim 1 or 2, wherein the exposure apparatus has areas that are respectively non-parallel to a pair of images formed by the projection optical system. 4. The specific pattern includes a line segment that is arranged non-parallel to the predetermined measurement direction, and the light receiving surface of the light receiving unit is with respect to an image of the line segment by the projection optical system. 3. A pair of regions, which are arranged obliquely and non-parallel to each other, respectively.
The exposure apparatus described.