JPH0697031A - Positioning method for projection aligner - Google Patents

Abstract

PURPOSE:To achieve that the base line amount of an alignment system and the amount of rotation of a reticle are measured with high accuracy in a projec tion aligner provided with an off-axis type alignment system. CONSTITUTION:A large-sized reference mark plate 17 is fixed to a part adjacent to a wafer W on a wafer stage 6. Reference marks 38A, 38B and a reference mark 39 are formed on the large-sized reference mark plate 17 at intervals at which TTR-type alignment systems 5A, 5B and a wafer alignment system 9 can be abserved simultaneously. The traveling direction of the wafer stage 6 is corrected so as to be along a coordinate system decided by a straight line connecting the reference marks 38A, 38B, and a reticle R is fixed by referring to the reference marks 38A, 38B.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a positioning method of a projection exposure apparatus for exposing a mask pattern onto a photosensitive material applied to a substrate such as a semiconductor wafer or a glass plate for a liquid crystal display device, and particularly to an off-axis method. The present invention relates to a positioning method for a projection exposure apparatus having a function of managing a so-called baseline amount of the alignment system with high accuracy.

[0002]

2. Description of the Related Art Conventionally, a projection exposure apparatus provided with an off-axis type alignment system is disclosed in Japanese Patent Laid-Open No. 53-5697.
As disclosed in Japanese Patent Laid-Open No. 5 and JP-A-56-134737, a reference mark plate is provided on a wafer stage that holds a wafer as a photosensitive substrate and moves two-dimensionally by a step-and-repeat method. The reference mark plate is fixed and the distance between the off-axis type alignment system and the projection optical system, that is, the so-called baseline amount is managed.

FIG. 13 shows a main part of a projection exposure apparatus having a conventional off-axis type alignment system. In FIG. 13, exposure light from a light source system (not shown) is condensed by a main condenser lens 1. As a result, the reticle R is illuminated with a uniform illuminance. The reticle R is held on the reticle stage 2, and the reticle stage 2 holds the reticle R in a state where the center Rc of the reticle R matches the optical axis AX of the projection optical system PL. Further, a pair of alignment reticle marks 3A and 3B are formed outside the pattern area on the lower surface of the reticle R, and mirrors 4A and 4B are provided above the reticle marks 3A and 3B, respectively.
TR (through the reticle) alignment system 5
A and 5B are arranged.

At the time of exposure, the pattern of the reticle R is projected and exposed onto each shot area of the wafer W on the wafer stage 6 via the projection optical system PL. A wafer mark for alignment is formed in each shot area of the wafer W. Further, a reference mark plate 7 having alignment marks 8 formed thereon is fixed near the wafer W on the wafer stage 6. When the wafer stage 6 is positioned so that the reference mark plate 7 is located at a position substantially conjugate with the reticle mark 3A or 3B in the projection field of the projection optical system PL, the alignment system 5A or 5B on the reticle R respectively causes the reticle mark 3A or 3B and the mark 8 are detected at the same time. The distance La between the reticle mark 3A (the same applies to 3B) and the center Rc of the reticle R is a value that is predetermined in design, and the reticle mark 3A on the image plane of the projection optical system PL (the surface of the reference mark plate 7).
The distance between the projection point of and the optical axis AX is La / M. This M
Is a projection optical system PL from the wafer W side to the reticle R side.
And the projection optical system PL is a 1/5 reduction projection optical system, M = 5.

An off-axis type wafer alignment system 9 is arranged outside the projection optical system PL. The optical axis of the wafer alignment system 9 is parallel to the optical axis AX of the projection optical system PL on the wafer stage 6. Then, the index plate 10 on which the index mark is formed is fixed inside the wafer alignment system 9, and the surface of the index plate 10 on which the index mark is formed is conjugate with the surface of the reference mark plate 8.

The baseline amount BL of the wafer alignment system 9 is defined as an interval between the optical axis on the wafer stage 6 of the wafer alignment system 9 and the projection point of the center Rc of the reticle R by the projection optical system PL as an example. To be done.
In order to measure the baseline amount BL, the wafer stage 6 is driven and, for example, the mark 8 on the reference mark plate 7 is first moved to a position A immediately below the wafer alignment system 9 to obtain an image of the mark 8. The amount of positional deviation from the index mark in the wafer alignment system 9 and the coordinates of the wafer stage 6 at that time are read. The coordinates of the wafer stage 6 are measured with high resolution by a laser interferometer. Thereby, the coordinates (X1, Y1) of the wafer stage 6 when the mark 8 is on the optical axis of the wafer alignment system 9 are obtained.

Next, the wafer stage 6 is driven to sequentially move the marks 8 on the reference mark plate 7 to the reticle marks 3A and 3B.
And the positions of the marks 8 and the reticle marks 3A and 3B, respectively, and the coordinates of the wafer stage 6 at that time are read. As a result, the mark 8 is located at the center between the reticle mark 3A and the reticle mark 3B, that is, the center Rc of the reticle R.
Coordinates of the wafer stage 6 (X2,
Since Y2) is known, coordinates (X1, Y1) and coordinates (X
2, Y2), the baseline amount BL is obtained. This baseline amount BL later reads the coordinates of the wafer mark on the wafer W by the wafer alignment system 9, and then positions each shot area of the wafer W within the exposure area of the projection optical system PL based on the read coordinates. It becomes the reference amount when doing.

That is, the distance between the center of a certain shot area on the wafer W and the wafer mark in the X direction is XP, and when the wafer mark coincides with the optical axis of the wafer alignment system 9, the wafer stage 6 moves in the X direction. Assuming that the position is X3 and the component of the baseline amount BL in the X direction is BLx, in order to match the center of the shot area designated by the wafer mark with the projection point of the center Rc of the reticle R, the wafer stage 6 is moved. It may be moved in the X direction by the amount of the following formula. X3-BLx-XP

The amount of movement in the Y direction can also be expressed by the same formula. The calculation formula is a calculation formula when the arrangement of FIG. 13 is used, and the calculation method differs depending on the arrangement of the reticle marks 3A and 3B or the arrangement of the wafer alignment system 9. No matter which calculation formula is used,
After the position of each wafer mark on the wafer W is detected in advance by using the wafer alignment system 9 of the axis system, each shot area on the wafer W is positioned within the exposure area of the projection optical system PL according to the detected position. By performing the exposure, the pattern of the reticle R is accurately superimposed and exposed on each shot area of the wafer W.

Further, even if the reticle R is mounted around the optical axis of the projection optical system PL while rotating from a reference angle, the overlay accuracy deteriorates. Therefore, the rotation amount of the reticle R (reticle R is as follows.・ Rotation) is measured. That is, in FIG. 13, the wafer stage 6 is driven to sequentially move the marks 8 on the reference mark plate 7 to the positions B and C, and the conjugate images of the marks 8 and the reticle marks 3B and 3A are respectively moved by the alignment systems 5B and 5A. Measure the relative positional relationship with. As a result, the wafer stage 6
The rotation amount of the reticle R based on the running direction of is measured. Then, conventionally, the reticle R is moved by a reticle fine movement mechanism (not shown) so that the rotation amount of the reticle R based on the running direction of the wafer stage 6 becomes a predetermined allowable value or less.
I was adjusting the amount of rotation.

[0011]

In the conventional projection exposure apparatus, since the baseline amount of the off-axis wafer alignment system 9 and the rotation amount of the reticle R are measured with the running direction of the wafer stage 6 as a reference, There is a disadvantage that the baseline amount and the rotation amount of the reticle R cannot be measured with high accuracy due to a measurement error of the coordinate measuring system of the wafer stage 6. That is, the influence of fluctuations that occur in the optical path of the laser interferometer that monitors the coordinates of the wafer stage 6, the error in the reset position of the initial coordinates of the wafer stage 6, or the setting position of the movable mirror that reflects the laser beam on the wafer stage 6. Due to an error or the like, an error is included in the measurement value of the coordinates of the wafer stage 6, and an error is also included in the baseline amount and the rotation amount of the reticle R based on the measurement value.

In view of the above problems, the present invention provides a projection exposure apparatus having an off-axis type alignment system,
It is an object of the present invention to provide a positioning method for measuring the baseline amount of the alignment system and the rotation amount of the reticle with high accuracy.

[0013]

A projection aligner positioning method according to the present invention positions a mask (R) having a pattern (12) to be exposed and a plurality of alignment marks (3A, 3B). And the mask stage (6) for holding the photosensitive substrate (W) on which a plurality of substrate marks are formed and positioning the photosensitive substrate (W), and the pattern of the mask (R) for the substrate stage. (6) A projection optical system (PL) for imaging and projecting on a region near each substrate mark on the photosensitive substrate (W) on the mask (R)
Image of the first mark detection means (9) for detecting the substrate mark on the photosensitive substrate (W) without passing through the reference mark (38A or 38B) for the mask arranged on the substrate stage (6). Stage of a projection exposure apparatus provided with a second mark detecting means (5A or 5B) for detecting a positional deviation amount between the mark (3A or 3B) of the mask (R) via the projection optical system (PL). The present invention relates to the positioning method of (6).

In the present invention, the substrate stage (6)
A first reference mark (39) provided above and detected by the first mark detection means (9) and a plurality of second reference marks (38A, 38) as reference marks for the mask.
The reference mark plates (17) arranged apart from each other so that B) are simultaneously detected by the first mark detecting means (9) and the second mark detecting means (5A, 5B), respectively, and the substrate stage (6). Substrate stage coordinate measuring means (18X, 18Y, 19X, 20) for measuring the moving coordinate and the rotation amount of the
X, 21Y), and a reference mark plate (17) and substrate stage coordinate measuring means (18X, 18Y, 19X, 2) in advance.
The inclination with respect to the coordinate axes (X, Y) of 0X, 21Y) is measured in advance, and the running coordinate system of the substrate stage (6) is determined with the reference mark plate (17) as a reference.

In this case, the amount of positional deviation between the conjugate image of the plurality of second reference marks (38A, 38B) on the reference mark plate (17) and the plurality of marks (3A, 3B) on the mask (R) is set to the second. Detected by the mark detection means (5A, 5B) of
It is desirable to calculate the inclination between the mask (R) and the coordinate axes of the substrate stage coordinate measuring means (18X, 18Y, 19X, 20X, 21Y) based on the detected positional deviation amount and the inclination measured in advance. .

The second mark detecting means (3A, 3
B) is arranged movably along a predetermined axis, and the second tilt is set along the coordinate axis of the substrate stage coordinate measuring means (18X, 18Y, 19X, 20X, 21Y) in which the calculated tilt is corrected. The mark detecting means (3A, 3B) may be moved.

[0017]

According to the present invention, the first fiducial mark (3
9) and a plurality of second reference marks (38A, 38B) are arranged separately so that they can be simultaneously detected by the first mark detecting means (9) and the second mark detecting means (5A, 5B), respectively. A large fiducial mark plate (17) is provided. Therefore, the first as an off-axis alignment system without moving the substrate stage (6)
The mark detecting means (9) of the first reference mark (3
9) to detect the position of the second mark detecting means (5A, 5A).
The position of the second fiducial mark (38A, 38B) is detected by (B), and the first fiducial mark (3
By correcting the detection results in the interval between the position of 9) and the position of the second reference mark (38A, 38B),
The baseline amount of the off-axis alignment system can be obtained with high accuracy and high speed.

Then, one of the running coordinate systems of the substrate stage (6) is set, for example, in parallel with a straight line passing through the plurality of second reference marks (38A, 38B) of the reference mark plate (17), so that the substrate stage The pitching error of (6) can be reduced.

Next, as described above, the plurality of second reference marks (38) are formed by the second mark detecting means (5A, 5B).
Since the positions (A, 38B) are detected, the rotation amount of the mask (R) when the reference mark plate (17) having a larger size is used as a reference can be obtained from these positional relationships. In this case, the rotation amount of the substrate stage (6) can be obtained without moving it.

Further, the reference mark plate (17) and the substrate stage coordinate measuring means (18X, 18Y, 19X, 20) are previously prepared.
Since the inclination with respect to the coordinate axis (X, Y) of (X, 21Y) is measured, the substrate stage coordinate measuring means (18) is obtained from the rotation amount and the inclination of the mask (R) obtained as described above.
X, 18Y, 19X, 20X, 21Y) coordinate axes (X,
The tilt between Y) and the coordinate system of the mask (R) is obtained. That is, the rotation error of the mask (R) corresponding to the reticle rotation can be measured easily and with extremely high accuracy.

Next, the second mark detection means (3A, 3
B) is arranged movably along a predetermined axis, and the second tilt is set along the coordinate axis of the substrate stage coordinate measuring means (18X, 18Y, 19X, 20X, 21Y) in which the calculated tilt is corrected. When the mark detecting means (3A, 3B) is moved, the second mark detecting means (3
Even if (A, 3B) is moved, mark detection is performed with high accuracy.

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows the configuration of the projection exposure apparatus of the present embodiment. In FIG. 1 in which parts corresponding to FIG. 13 are assigned the same reference numerals, a pattern in which a circuit pattern to be exposed on the reticle R is formed. Two reticle marks 3A, 3 for alignment so as to face the outside of the area 12
B is provided. TTR (through the reticle) above one of the reticle marks 3A via the mirror 4A.
A system alignment system 5A is arranged, and a TTR system alignment system 5B is also arranged above the other reticle mark 3B via a mirror 4B. The reticle R is held on the reticle stage 2 and
Allows the reticle R to be translated (X, Y directions) and rotated in a two-dimensional plane by a drive system (not shown).

At the time of exposure, the pattern of the reticle R is projected and exposed onto each shot area of the wafer W on the wafer stage 6 via the projection optical system PL. Further, in the one TTR type alignment system 5A above the reticle R,
The image of the reticle mark 3A and the image of the mark in the projection area of the projection optical system PL is formed on the image pickup surface of the image pickup element 14A for the Y direction and the image pickup surface of the image pickup element 15A for the X direction by the objective lens 13A via the mirror 4A. To be done. Further, part of the light from the objective lens 13A also enters the light receiving element 16A.
The light receiving surface of the light receiving element 16A is conjugate with the pupil surface (Fourier transform surface) of the projection optical system PL. Similarly, in the other TTR-type alignment system 5B, the image of the reticle mark 3B and the image of the mark in the projection area of the projection optical system PL are picked up by the objective lens 13B via the mirror 4B. And an image is formed on the image pickup surface of the image pickup device 15B for the X direction. Further, a part of the light from the objective lens 13B also enters the pupil-conjugated light receiving element 16B of the projection optical system PL.

In this embodiment, the X-direction laser interferometer 19X and the pitching measurement laser interferometer 20X are respectively connected to the X-direction end portions on the wafer stage 6 by X-direction.
The laser beam LB1 emitted parallel to the direction (see FIG.
(Refer to (a)) and the movable mirror 1 which reflects LB2 in the incident direction.
8X is attached, and at the end in the Y direction on the wafer stage 6, a moving mirror that reflects in the incident direction a laser beam LB3 emitted in parallel from the Y direction laser interferometer 21Y in the Y direction perpendicular to the X direction. Attach 18Y. In this case, the extension line of the laser beam LB1 and the extension line of the laser beam LB3 intersect at the optical axis AX of the projection optical system PL.

Further, the large fiducial mark plate 17 is fixed near the wafer W on the wafer stage 6 and inside the region where the movable mirror 18X and the movable mirror 18Y intersect. On the large fiducial mark plate 17, there are two fiducial marks 38A and 3A.
8B is formed along a straight line substantially parallel to the X direction, and the reference mark 39 is formed at a position separated in the Y direction along the perpendicular bisector of the reference marks 38A and 38B. The intervals between the reference marks 38A and 38B are set so that the reference marks 38A and 38B can be observed simultaneously by the TTR alignment systems 5A and 5B, respectively. Various other alignment marks are formed on the large reference mark plate 17 as described later.

The large reference mark plate 17 has a vapor deposition layer of chromium or the like formed on the surface of a transparent member having a low expansion coefficient such as a quartz plate.
A light transmitting portion having a shape of a reference mark or the like is formed on a part thereof by etching or the like. In addition to the reference marks illuminated from the projection optical system PL side, there are also large-scale reference mark plates 17 illuminated from the wafer stage 6 side, such as the reference marks 38A and 38B. Therefore, the illumination light IL in the same wavelength band as the exposure light used when illuminating the reticle R is guided into the inside of the wafer stage 6 via the light guide 22, and the illumination light IL causes a predetermined mark on the large-sized reference mark plate 17 to be guided. Is illuminated from the bottom side.

Reference numeral 9 denotes an off-axis wafer alignment system.
Of the reflection prism 23, the objective lens 24, the mirror 25, the main body 26, the X-direction image pickup device 27X and the Y-direction image pickup device 27Y, and the wafer alignment system 9 in the Y direction of the projection optical system PL. Attach to the side part.
The images of the marks in the areas facing the reflecting prism 23 are formed on the image pickup surfaces of the image pickup devices 27X and 27Y, respectively. In this case, in the present embodiment, the wafer stage 6 is driven to drive the reference marks 38A and 3A of the large reference mark plate 17.
8B are TTR type alignment systems 5A and 5 respectively.
The position of the reference mark 39 is set so that the reference mark 39 fits within the observation field of view of the wafer alignment system 9 when the reference mark 39 is placed at a position where the reference mark 39 can be observed. Therefore, in the present embodiment, the TTR type alignment systems 5A and 5B and the wafer alignment system 9 can simultaneously observe the corresponding fiducial marks on the large fiducial mark plate 17, respectively. Therefore, without moving the wafer stage 6,
The baseline amount of the off-axis wafer alignment system 9 can be measured.

An auxiliary wafer alignment system 11 is fixed to the side surface of the projection optical system PL in the X direction. This auxiliary wafer alignment system 11 also has a reflecting prism 28, an objective lens 29, and a mirror 3 facing the wafer stage 6.
0, a main body 31, an X-direction image pickup device 32X, and a Y-direction image pickup device 32Y. The auxiliary wafer alignment system 11 is a wafer alignment system 9
Similar to the above, it is used to observe the position of the fiducial mark on the large fiducial mark plate 17.

Further, in the present embodiment, the TTL for detecting the mark on the wafer W and the large reference mark plate 17 through only the projection optical system PL without passing through the reticle R.
A (through the lens) type wafer alignment system is provided separately for the X direction and the Y direction. The TTL type wafer alignment system for the X direction includes a projection optical system PL via a mirror 33X, an objective lens 34X and an objective lens 34X which are fixed to the outer peripheral portion in the Y direction between the reticle stage 2 and the projection optical system PL. It has a light receiving element 37X for receiving the light returned from the side. Furthermore, the objective lens 3
A laser step alignment type light-transmitting system (hereinafter, referred to as "LSA-type light source") for transmitting a slit-shaped light beam onto the wafer stage 6 in order to make alignment detection light incident on the projection optical system PL via 4X. "Transmitting system")
35X and a light transmitting system (hereinafter, referred to as "LIA system light transmitting system") for transmitting a coherent light beam of two light beams 36X
And are provided in a switching manner.

Similarly, the TTL type wafer alignment system for the Y direction is a reticle stage 2 and a projection optical system PL.
And a light receiving element 37Y that receives light returned from the projection optical system PL side via the objective lens 34Y and the objective lens 34Y. In addition, in symmetry with the TTL type wafer alignment system for the X direction, an LSA type light transmitting system 35Y for transmitting a slit-shaped light beam onto the wafer stage 6 and two coherent light beams are transmitted. A light transmission system 36Y of LIA system for switching is provided by a switching system.
Although the TTL type wafer alignment system for the Y direction is arranged in the positive direction of the X axis with respect to the projection optical system PL in FIG. 1, the wafer alignment system is actually the X direction.
It is located in the negative direction of the axis.

FIG. 2A shows the large-sized reference mark plate 1 in FIG.
7 shows the detailed mark shape and mark arrangement of FIG.
As shown in (a), the reference mark 3 for the reticle mark
8A and 38B are formed by arranging light-shielding multi-mark patterns for X-direction and Y-direction in a light-transmitting portion surrounded by a broken line, respectively, for an off-axis wafer alignment system 9. The reference mark 39 of the X direction and Y
It is a reflection type lattice pattern formed at a predetermined pitch in each direction. The distance between the reference marks 38A and 38B in the X direction is M. Reference mark 38A for reticle mark
, 38B are illuminated from the bottom by the illumination light guided through the light guide 22 of FIG. 1, and the reference marks 38A and 38B can be said to be luminescent marks.

Further, a reference mark 40Y for interference of two light beams, which is composed of a diffraction grating having a predetermined pitch in the Y direction, is arranged in the vicinity of the reference mark 38B in the X direction, and the reference mark 38A and 38B is centered at the center of the reference mark 38A and 38B. A reference mark 40X for two-beam interference formed of a diffraction grating having a predetermined pitch in the X direction is arranged at a position where the 40Y is rotated 90 ° counterclockwise. The X-direction position of the reference mark 40X for two-beam interference and the Y-direction position of the reference mark 40Y are detected by the two-beam laser beams emitted from the LIA type light transmission systems 36X and 36Y of FIG. 1, respectively. It Also,
With the reference marks 38A and 38B being observed by the TTR type alignment systems 5A and 5B of FIG. 1, respectively, the reference marks 40X and 40Y of the two light fluxes from the LIA type light transmitting systems 36X and 36Y of FIG. 1, respectively. Those fiducial marks 40X so that they are illuminated by the illumination light
And 40Y. However, as already explained,
The TTL type wafer alignment system including the LSA type light transmitting system 35Y and the LIA type light transmitting system 36Y of FIG.
Actually, it is arranged on the opposite side with respect to the projection optical system PL.

A cross mark 41 having a cross-shaped opening pattern is formed at a position near the reference mark 39 between the center of the reference marks 38A and 38B and the reference mark 39. This cross mark 41 is also illuminated from the bottom surface of the large reference mark plate 17 by the illumination light guided through the light guide 22 of FIG. Therefore, after that, the cross mark 41
Is referred to as a "light emitting cross mark 41". Further, in the vicinity of the emission cross mark 41 on the large-sized reference mark plate 17 in the X direction,
A reference mark 42Y for slit scanning for the Y direction is formed by arranging reflective patterns of several μm square in the X direction, and the point is located M / 2 away from the emission cross mark 41 in the negative direction of the X axis. As a reference mark 42 for slit scanning for the X direction, a reference pattern 42Y is formed by arranging a reflection type pattern of several μm square in the Y direction at a position rotated 90 ° counterclockwise.
Form X. The light from the light emitting cross mark 41 is transmitted through the projection optical system PL to the reticle mark 3A of the reticle R in FIG.
Mark 4 for slit scanning when illuminating
The reference marks 42X and 42Y are arranged so that 2X and 42Y are illuminated by the slit-like illumination light from the LSA type light transmitting systems 35X and 35Y, respectively.

For example, the LSA type light transmitting system 35Y of FIG.
2A, the wafer stage 6 is driven to move the large reference mark plate 17 in the Y direction in a state where the illumination light is emitted as a slit-shaped beam spot LSP1 in the vicinity of the reference mark 42Y. When the pattern LSP1 and the reference mark 42Y coincide with each other in the Y direction, strong diffracted light is emitted from the reference mark 42Y in a predetermined direction. By detecting this diffracted light,
The Y coordinate of the reference mark 42Y can be detected extremely accurately. Similarly, the X coordinate of the reference mark 42X for the X direction can be detected extremely accurately.

FIG. 2B shows the reticle mark 3A of FIG. 1 (the reticle mark 3B has the same shape),
The reticle mark 3A is configured by combining a double rectangular pattern 43 and a cross-shaped pattern 44. The rectangular pattern 43 is the TTR type alignment system 5A of FIG.
The cross-shaped pattern 44 is used for detecting the matching state between the light emitting cross mark 41 and the reticle mark 3A in FIG. 2A.

FIG. 3 shows the configuration of the TTR type alignment system 5A of FIG. 1. In FIG. 3, the mirror 4A is arranged above the reticle mark 3A of the reticle R with an inclination of 45 °, and is projected from the projection optical system PL. The objective lens 13A is directed in the direction in which the light transmitted through the reticle mark 3A is reflected by the mirror 4A.
To place. The light flux that has passed through the objective lens 13A is split into two light fluxes by the half mirror 45A.
The light reflected by is incident on the half mirror 47A via the imaging lens 46A, and the light beams divided into two by the half mirror 47A are respectively image pickup devices 15A and Y for the X direction, which are charge coupled image pickup devices (CCD). It is incident on the image pickup surface of the directional image pickup device 14A. Imaging elements 15A and 14A
Of the reticle R is conjugate with the surface on which the reticle mark 3A is formed. Therefore, the image pickup devices 15A and 14
The image pickup surface of A is also conjugate with the mark forming surface of the large reference mark plate 17. In addition, the image pickup device 15A for the X direction
The main scanning line (horizontal scanning line) of the reticle mark 3A
Is orthogonal to the X mark, and the direction of the main scanning line of the image sensor 14A for the Y direction is orthogonal to the Y mark of the reticle mark 3A.

At this time, the large fiducial mark plate 17 of FIG. 1 moves to move the reticle mark 3A shown in FIG.
Reference mark 38A of the large reference mark plate 17 shown in (b)
When the conjugate image 38A of the image pickup device substantially overlaps the conjugate image 38A of the image pickup device, the images shown in FIG. 4C are formed on the image pickup surfaces of the image pickup devices 15A and 14A. In this case, the image pickup device 15A for the X direction supplies the image pickup signal of the rectangular region 55X in the X direction of FIG. 4C to the image processing circuit 49A of FIG. 3, and the image pickup device 14A for the Y direction of FIG. The image pickup signal of the rectangular area 55Y in the Y direction of c) is supplied to the image processing circuit 48A of FIG. The image processing circuit 49A obtains the amount of positional deviation in the X direction between the reticle mark 3A and the conjugate image 38P of the reference mark 38A, and the image processing circuit 48A determines the reticle mark 3A.
The positional deviation amounts of the reference mark 38A and the conjugate image 38P in the Y direction are obtained, and the positional deviation amounts in the X direction and the Y direction are supplied to the main control system 50, respectively.

The main control system 50 includes a reticle mark 3A and a reference mark 3 via a drive system 51 for the reticle stage 2.
The position shift amount from the conjugate image 38P of 8A is set to be a predetermined value or less. Further, the main control system 50 includes a laser interferometer 19
The wafer stage 6 is positioned via the drive system 52 for the wafer stage 6 based on the measurement coordinates of X, 20X and the laser interferometer 21Y. Further, in FIG. 3, the light flux transmitted through the half mirror 45A is incident on the light receiving surface of the light receiving element 16A including a photomultiplier or the like via the relay lenses 53A and 54A. The light receiving surface of the light receiving element 16A is conjugate with the pupil plane of the projection optical system PL, and when the light emission cross mark 41 of the large fiducial mark plate 17 of FIG. 2A is at a position substantially conjugate with the reticle mark 3A, The light beam emitted from the light emitting cross mark 41 and passing through the projection optical system PL and the transmitting portion around the reticle mark 3A is photoelectrically converted by the light receiving element 16A.

The light receiving element 16A detects the amount of transmitted light that changes when the conjugate image of the light emitting cross mark 41 scans the reticle mark 3A (or 3B) and generates a photoelectric conversion signal SSD corresponding to the amount of transmitted light. This photoelectric conversion signal SS
The process of D is an up / down pulse output from the laser interferometers 19X and 21Y in accordance with the movement of the wafer stage 6 in FIG. 1 (for example, one pulse for each 0.01 μm movement amount).
The signal is sampled in synchronism with the above, is digitally converted and then stored in the memory. By specifically processing the photoelectric conversion signal SSD, the conjugate image of the light emission cross mark 41 is converted into the reticle mark 3A (or 3).
B) when the wafer stage 6 coordinates (X,
Y) is required. The TTR type alignment system 5B of FIG. 1 is also configured similarly to FIG. 3, but the main control system 50 and the drive systems 51, 52 are common.

FIG. 5 shows the structure of the TTL type wafer alignment system in the Y direction shown in FIG. 1. In FIG.
The red laser beam from the e-Ne laser light source 56 is used as the mark illumination light. Since the red laser beam has weak sensitivity to the resist layer of the wafer W, even during the exposure of the wafer W, the red laser beam can be used to detect the wafer mark near each shot area of the wafer W. It can be carried out. Further, this TTL type wafer alignment system includes an LSA type light transmitting system 35Y and LIA.
2 with different mark detection principle, which is composed of a light transmitting system 36Y
One alignment system is incorporated, and two alignment systems are selectively used. Such a configuration is disclosed in JP-A-2-272305 and JP-A-2-272305.
Since it is disclosed in detail in Japanese Patent No. 283011, a brief description will be given here.

The red laser beam from the He-Ne laser light source 56 is split by a beam splitter 57, and the two split laser beams go to shutters 58 and 59 which are alternately opened and closed. In FIG. 5, the shutter 58 is closed and the shutter 59 is open.
The laser beam that has passed through the optical path opened by 9 is LI
It is incident on the A (two-beam interference alignment) type light transmitting system 36Y. The LIA type light transmitting system 36Y divides the supplied laser beam into two laser beams and emits the two laser beams with a constant frequency difference using an acousto-optic modulator or the like. . In FIG. 5, LIA
Two laser beams emitted from the system light-transmitting system 36Y are arranged parallel to each other in the direction perpendicular to the paper surface of FIG.

The two laser beams are reflected by the mirror 60 and the half mirror 61, and then split into two light beams by the beam splitter 62. The two laser beams reflected by the beam splitter 62 intersect at the aperture of the diaphragm 65 on the surface conjugate with the wafer W by the objective lens 64, and the two laser beams passing through the diaphragm 65 are reflected by the mirror 33Y. And enters the projection optical system PL. The two laser beams emitted from the projection optical system PL intersect again, for example, in the vicinity of the reference mark 40Y for two-beam interference on the large reference mark plate 17 in FIG. These two
One-dimensional interference fringes are formed in the region where the two laser beams intersect, and the interference fringes are formed in the pitch direction of the interference fringes (Y in this example) at a speed according to the frequency difference between the two laser beams.
Direction). In this case, the reference mark 40Y in FIG. 2 is a diffraction grating parallel to the interference fringe, and the reference mark 4Y
From 0Y, the interference beat light whose intensity changes at the beat frequency according to the frequency difference between the two laser beams is reflected.

When the pitch of the diffraction grating of the reference mark 40Y and the pitch of the interference fringes are set in a certain relation, the interference beat light is reflected vertically upward from the large reference mark plate 17, and the interference beat light is 2 Along the optical path of the laser beam of the book, the projection optical system PL of FIG.
It returns to the beam splitter 62 through 5 and the objective lens 64. The interference beat light transmitted through the beam splitter 62 enters the light receiving element 37Y. The light receiving surface of the light receiving element 37Y is arranged on a surface that is substantially conjugate with the pupil surface of the projection optical system PL. Further, the light receiving element 37Y is configured by disposing a plurality of photoelectric conversion elements (photodiodes or the like) separated from each other, and the interference beat light is located at the center of the light receiving element 37Y (center of the pupil conjugate plane). Is received by. The photoelectric conversion signal becomes a sinusoidal AC signal equal to the beat frequency, and this AC signal is supplied to the phase difference measuring circuit 66.

Further, the two laser beams that have passed through the beam splitter 62 pass through the inverse Fourier transform lens 68 and the mirror 69 and cross as a parallel light flux on the transmissive reference grating plate 70. Therefore, on the reference lattice plate 70, 1
A three-dimensional interference fringe is formed, and this interference fringe flows in one direction at a speed according to the beat frequency. Either the interference light of the ± first-order diffracted light or the interference light of the zero-order light and the second-order diffracted light generated in parallel from the reference grating plate 70 enters the light receiving element 71.
The intensity of these interference lights also changes sinusoidally at a frequency equal to the beat frequency, and the light receiving element 71 photoelectrically converts the interference light to generate an AC signal having a frequency equal to the beat frequency.
This AC signal is supplied to the phase difference measuring circuit 66.

The phase difference measuring circuit 66 has a phase difference Δφ (−180 ° <φ ≦ 180) between the AC signal from the light receiving element 71 and the AC signal from the photoelectric conversion element at the center of the light receiving element 37Y.
°) is obtained, and information SSB of the positional deviation amount in the Y direction of the reference mark 40Y on the large-sized reference mark plate 17 corresponding to the phase difference Δφ is supplied to the main control system 50 in FIG. The resolution of the positional deviation amount is, for example, about 0.01 μm. The main control system 50 of FIG. 3 servo-controls the drive system 52 of the wafer stage 6 based on the positional deviation information SSB from such a TTL alignment system of the LIA system so that the reference mark 40Y of the large reference mark plate 17 is controlled. The wafer stage 6 is arranged so that the reference lattice plate 70 is always driven into a fixed positional relationship.
Can be servo-locked. Similarly, the LI of FIG.
When the A-system light-transmitting system 36X is used, the wafer stage 6 is so arranged that the reference mark 40X of the large reference mark plate 17 of FIG. 2A is always driven into a fixed positional relationship with respect to the reference grid plate for the X direction. Can be servo-locked.

Returning to FIG. 5, when the shutter 58 is opened and the shutter 59 is closed, the laser beam is incident on the LSA (laser step alignment) type light transmitting system 35Y. What is the laser step alignment method?
As disclosed in Japanese Patent No. 233011, this is a method of scanning a mark with a slit-shaped laser spot light extending in a direction orthogonal to the mark detection direction. A signal obtained by photoelectrically converting the diffracted light or scattered light generated from the mark is sampled in synchronization with the up / down pulse from the laser interferometers 19X and 21Y generated with the movement of the wafer stage 6 for scanning the mark. To be done.

The laser beam incident on the LSA type light-transmitting system 35Y is formed by the action of the internal beam expander and the cylindrical lens so that the beam cross section at the converging point is shaped into a slit extending in one direction and emitted. It The laser beam emitted from the LSA type light-transmitting system 35Y and having a slit-shaped cross section has a half mirror 61 and a beam splitter 6.
2, incident on the projection optical system PL via the objective lens 64, the diaphragm 65 and the mirror 33Y. The diaphragm 65 is He-Ne.
The laser beam is conjugate with the surface of the large-sized reference mark plate 17 (the surface of the wafer W) under the wavelength of the laser beam, and the laser beam is condensed in a slit shape at the opening of the diaphragm 65. LS of FIG.
The beam spot LSP1 of the laser beam generated by the A-system light-transmitting system 35Y and emitted from the projection optical system PL is
As shown in FIG. 2A, it is shaped like a slit extending in the X direction at a stationary position in the exposure area of the projection optical system PL.

At this time, when the wafer stage 6 is scanned in the Y direction and the reference mark 42Y for slit scanning on the large reference mark plate 17 crosses the beam spot LSP1, diffracted light or scattered light is emitted from the reference mark 42Y. Occur. The generated reflected light such as diffracted light or scattered light is incident on the light receiving element 37Y via the projection optical system PL, the mirror 33Y, the diaphragm 65, the objective lens 64 and the beam splitter 62 in FIG. The reflected light from the reference mark 42Y is photoelectrically converted by the photoelectric conversion element around the central photoelectric conversion element in the light receiving element 37Y, and this photoelectric conversion signal is supplied to the LSA processing circuit 67. LSA processing circuit 67
Of the photoelectric conversion signal is sampled according to the up / down pulse signal UDP from the laser interferometer 21Y for the wafer stage 6, digitally converted and written in the memory. Further, the LSA processing circuit 67 determines the beam spot LSP of FIG.
The Y-coordinate of the wafer stage 6 when the center point in the Y-direction of 1 and the center point in the Y-direction of the reference mark 42Y exactly match is calculated, and this Y-coordinate is used as the mark position information SSA in FIG. It is supplied to the control system 50. Mark position information S
SA is also used for drive control of the drive system 52 for the wafer stage 6.

Further, in FIG. 5, the LSA processing circuit 67 is provided.
The photoelectric conversion signal SS from the light receiving element 54A of FIG.
A memory for storing data obtained by sampling D in synchronization with the up / down pulse UDP and a circuit for high-speed arithmetic processing of the signal in the memory are provided, and the reticle mark 3A and the light emission cross of FIG. The coordinates of the wafer stage 6 when they match the conjugate image of the mark 41 are calculated as the projection position information SSC of the reticle mark 3A, and this projection position information SSC is supplied to the main control system 50. The alignment system including the LSA type light transmitting system 35X and the LIA type light transmitting system 36X of FIG. 1 is also configured in the same manner as in FIG.

FIG. 6 shows the configuration of the off-axis wafer alignment system 9 shown in FIG. 1. In FIG. 6, the illumination light EL includes a half mirror 72, a lens system 73, a mirror 74, and the like.
The surface of the large reference mark plate 17 (or the surface of the wafer W) is illuminated via the objective lens 24 and the reflection prism 23. Light reflected from the surface of the large-sized reference mark plate 17 enters the half mirror 72 through the prism mirror 23, the objective lens 24, the mirror 74, and the objective lens 73, and the light reflected by the half mirror 72 enters the index plate 75. To do. The index plate 75 and the surface of the large-sized reference mark plate 17 are conjugate with each other, and the image of the reference mark 39 shown in FIG. 2A is formed on the index plate 75 when the baseline amount is measured. The illumination light EL has a wavelength range in which the photosensitivity of the resist layer of the wafer W is extremely low.
It has a band of about nm.

As shown in FIG. 7, the index plate 75 is formed on a transparent glass plate with index marks 78A and 78B each consisting of four light-shielding line groups at predetermined intervals in the X direction, and is formed in the Y direction. Also has index marks 79A and 79B formed of four light-shielding line groups at predetermined intervals. Further, in FIG. 7, the large fiducial mark plate 17
The image 39P of the reference mark 39 of FIG. 2 is formed in the center of the index plate 75.
8A, 78B position shift amount in the X direction and its image 39P
The amount of positional deviation between the index marks 79A and 79B in the Y direction represents the amount of positional deviation between the reference mark 39 and the optical axis of the wafer alignment system 9.

In FIG. 6, the images of the index mark on the index plate 75 and the reference mark 39 (or the wafer mark on the wafer W) on the large-sized reference mark plate 17 are passed through the imaging lens 76 and the half mirror 77, respectively. X consisting of CCD
Image sensor 27X for direction and image sensor 27Y for Y direction
Is imaged on the image pickup surface of. In this case, the image pickup regions of the image pickup elements 27X and 27Y on the index plate 75 are set to the region 80X in the X direction and the region 80Y in the Y direction of FIG. 7, respectively. Further, the directions of the main scanning lines (horizontal scanning lines) of the image pickup devices 27X and 27Y are determined to be directions conjugate with the X direction and the Y direction of FIG. 7, respectively, and the image pickup signals of the image pickup devices 27X and 27Y are omitted in the drawing. The amount of positional deviation between the index plate 75 and the reference mark 39 (or the wafer mark on the wafer W) of the large-sized reference mark plate 17 can be obtained by performing the processing described in 1. The information SSE indicating the information on the amount of positional deviation is supplied to the main control system 50 in FIG.

In this example, an example of the detection center in the X direction on the index plate 75 of the off-axis wafer alignment system 9 is the center coordinate in the X direction of the index marks 78A and 78B in FIG. The center coordinate in the X direction of one index mark 78A may be the detection center. The same applies to the detection center of the wafer alignment system 9 in the Y direction. Similarly, an example of the position of the reference mark image 39P in the X direction on the index plate 75 is the average position of the detection positions of the line patterns in the X direction of the image 39P, and the average position of the reference mark image 39P in the Y direction. An example of the position is the average position of the detected positions of each line pattern of the image 39P in the Y direction. However, when the baseline amount of the wafer alignment system 9 is calculated, it is necessary to convert the coordinates of the detection centers into the coordinates of the conjugate point on the wafer stage 6.

Next, a method of positioning the wafer stage 6 of the projection exposure apparatus of this example will be described. In this regard, the reference of the running direction of the wafer stage 6 in the conventional projection exposure apparatus is the fixed mirrors inside the laser interferometers 19X and 21Y of FIG. 1 and the movable mirrors 18X and 18Y corresponding thereto, respectively.
The relative position with is used as the absolute coordinate. Therefore, there are inconveniences that an error is mixed in the absolute coordinates due to the air fluctuation, and the yaw error of the wafer stage 6 becomes large by performing the coordinate measurement at a position deviated from the optical axis AX of the projection optical system PL. Further, it is necessary to move the wafer stage 6 at the time of measuring the baseline amount, and if the measurement is repeated several times for averaging in order to improve accuracy, there is a disadvantage such as a decrease in throughput.

On the other hand, in the present example, the running direction of the wafer stage 6 is set with reference to the large fiducial mark plate 17 of FIG. This is performed by aligning the position of the reticle R with the position of the large fiducial mark plate 17 as a reference when checking the baseline amount of the off-axis wafer alignment system 9, and further by using the laser interferometer 19 that functions as a yawing sensor.
The difference between the measured values of X and the laser interferometer 20X is reset to zero.

First, the error in measuring the baseline amount (baseline check) will be specifically described with reference to FIG. In FIG. 8A, a pattern 8 indicated by a broken line
The position 1 is the reference mirror position of the movable mirrors 18X and 18Y in FIG. 1, and the position of the pattern 82 shown by the solid line is the actual position of the movable mirrors 18X and 18Y when the interferometer is reset. The reference mirror position is a virtual arbitrary position.
The inclination of the pattern 82 with respect to the pattern 81 in this state, that is, the yawing error when the interferometer is reset is set to θi. Laser beams LB1 and LB2 are emitted from the laser interferometers 19X and 20X of FIG. 1 to the movable mirror 18X, and a laser beam LB3 is emitted from the laser interferometer 21Y of FIG. 1 to the movable mirror 18Y. Then, the coordinate values in the X direction measured by the laser interferometers 19X and 20X are set as Lx0 and Lf0, respectively, and the coordinate values in the Y direction measured by the laser interferometer 21Y are set as Ly, and initial coordinate values are set as follows. . Lx0 = Lf0

Next, referring to FIG. 1, the large fiducial mark plate 17
8 to move the movable mirrors 18X and 18Y when performing the baseline check of the wafer alignment system 9.
The pattern 83 of FIG.
Of the yaw error relative to the pattern 81 of θ
a. Further, when the distance between the laser beams LB1 and LB2 in the Y direction is L, and the coordinate values in the X direction measured by the laser interferometers 19X and 20X are Lx and Lf, respectively, the following equation holds. Lf = (θa−θi) L + Lx

The difference (θa-θi) L of the coordinate values is stored as an offset A in the memory. As a result, the coordinate value Lx measured by the laser interferometer 19X and the coordinate value Lf measured by the laser interferometer 20X become (Lf = A + L
The state satisfying the relationship of x), that is, the yawing state of the movable mirrors 18X and 18Y at the time of baseline check can be treated as the reference mirror position thereafter. Also, FIG.
(C) Large reference mark plate 1 for baseline check
7 and the movable mirrors 18X and 18Y are shown in FIG.
As shown in (c), the moving mirror 18X
On the other hand, the large reference mark plate 17 has a rotation error of the angle θf. This rotation error θf can also be defined as the inclination with respect to the X axis defined by the movable mirrors 18X and 18Y, which is a straight line passing through the centers of the two reference marks 38A and 38B of the large reference mark plate 17.

Next, when the reticle R is aligned, as shown in FIG. 1, the two reference marks 38A and 38B of the large-sized reference mark plate 17 are connected to the reticle marks 3A and 3B by design. Place near the. As shown in FIG. 2A, the reference marks 38A and 38B.
Are illuminated from the bottom by the illumination light IL, so that the periphery of the reticle marks 3A and 3B on the reticle R are also illuminated by the illumination light IL. Therefore, first, the position and the rotation amount of the reticle R are set so that the reticle marks 3A and 3B are in a predetermined state with respect to the reference positions inside the alignment systems 5A and 5B of the TTR system, respectively. Fix 2 In this case, two-axis photoelectric microscopes may be provided inside the TTR alignment systems 5A and 5B, and the positions of the reticle marks 3A and 3B may be measured by a total of four-axis photoelectric microscopes.

However, the TTR type alignment system 5A
And the moving mirror 18 and a straight line connecting the reference positions of 5B.
Based on the inclination with the coordinate axis determined by X and 18Y, FIG.
As shown in (c), the reticle R has movable mirrors 18A and 18A.
There is a rotation error of an angle θr with respect to the coordinate axis defined by Y. Therefore, the rotation error θr and the magnification error of the projection optical system PL are calculated using the TTR type alignment systems 5A and 5B of FIG.

That is, as described with reference to FIG. 4C, in the TTR alignment system 5A, the X-direction error RAX1 and the Y-direction error RAY of the conjugate image of the reference mark 38A and the reticle mark 3A. Is required, TT
In the R type alignment system 5B, the reference mark 3
Error RA in the X direction between the conjugate image of 8B and the reticle mark 3B
The error RAθ in the X2 and Y directions is obtained. It is assumed that these errors RAX1 and the like are values converted into values on the wafer stage 6. And the reference marks 38A and 38B
When the distance between and is M, the rotation error Rr of the reticle R and the magnification error Rm of the projection optical system PL based on the reference marks 38A and 38B of the large reference mark plate 17 are as follows.

## EQU1 ## Rr = (RAθ-RAY) / M Rm = (RAX2-RAX1) / M

In this case, the rotation error θf of the large fiducial mark plate 17 with respect to the coordinate system defined by the movable mirrors 18X and 18Y is measured in advance and stored as a system error. Specifically, in order to measure the rotation error θf,
In FIG. 1, first, the reference mark 38B and the reticle mark 3 are shown.
A displacement amount Y1 in the Y direction with respect to B is measured by a TTR alignment system 5B, and then a displacement amount Y2 in the Y direction between the reference mark 38A and the reticle mark 3B is moved by moving the wafer stage 6 in the X direction. Is measured by the TTR type alignment system 5B. Since the distance between the reference marks 38A and 38B is M, the rotation error θf is as follows. θf = (Y1-Y2) / M

From FIG. 8C, using the rotation error θf and the rotation error Rr of (Equation 1), the moving mirrors 18X, 1
Rotation error θr of reticle R with respect to the coordinate system defined by 8Y
Is as follows. θr = Rr−θf

The moving direction of the wafer stage 6 is the moving mirror 1.
Since it is determined by the angles of the reflecting surfaces of 8X and 18Y, the rotation error θr can be considered as the rotation error between the running direction of the wafer stage 6 and the reticle R. It is assumed that the orthogonality correction and the bending correction of the movable mirrors 18X and 18Y have been performed before this measurement. Then, in this embodiment, the traveling direction of the wafer stage 6 is set according to the coordinate system in which the coordinate system of the movable mirrors 18X and 18Y is corrected by software by the rotation error θf. That is, as shown in FIG. 9, assuming that the coordinate systems X and Y indicated by broken lines are coordinate systems defined by the movable mirrors 18X and 18Y, the wafer stage 6 connects the reference marks 38A and 38B on the large-sized reference mark plate 17. The vehicle runs along a coordinate system having the straight line 84 as the X axis.

Next, with reference to FIG. 10, an example of the operation in the case of performing the alignment and the baseline check of the reticle R in this example will be described. First, in step 201, the laser interferometers 19X, 20X and 21Y are reset. Considering, for example, in the X direction, if there is a difference (dead path) between the optical path length to the fixed mirror and the optical path length to the movable mirror 18X in the laser interferometer 19X at the time of reset, the measured value is obtained only by changing the temperature of the atmospheric gas. Changes and a measurement error occurs. Therefore, the distance to a position where there is no dead path is obtained, and the measured values of the laser interferometers 19X, 20X, and 21Y are corrected so that the interferometer value with respect to temperature becomes zero. As a result, the measurement error caused by the temperature change can be minimized.

Next, at step 202, the wafer stage 6 is positioned by the laser step alignment (LSA) method. That is, the reference marks 42X and 42Y for slit scanning in FIG. 2A scan the beam spot from the LSA type light transmitting system 35X and the beam spot from the LSA type light transmitting system 35Y, respectively. And the Y coordinate of the reference mark 42Y are measured. Then, the large-sized reference mark plate 17 is positioned at the position at the time of the baseline check with the positions of the reference marks 42X and 42Y as references. Since the position variation of the beam spot is small in the laser step alignment method, when positioning the large fiducial mark plate 17 by this method, the interferometer reset or the TTR method alignment system 5 is used.
It is possible to suppress variations in the coordinates of the wafer stage 6 due to the setting errors of A and 5B. Therefore, the positioning operation can be performed with high accuracy and stability.

Then, in step 203, the TTR
Position setting (reset) of the alignment systems 5A and 5B of the method is performed. The reset of the alignment systems 5A and 5B may change the positional relationship between the position of the wafer stage 6 and the image processing regions 55X and 55Y of FIG. 4C. To avoid this, for example, the alignment system 5
Biaxial photoelectric microscopes may be provided in A and 5B, respectively, and the positions of the alignment systems 5A and 5B may be set with high precision using these photoelectric microscopes with the reference marks 38A and 38B on the large reference mark plate 17 as references. However, when the alignment systems 5A and 5B are fixedly used, step 203 is omitted.

Next, after setting the reticle R on the reticle stage 2 (step 204), the reticle R is aligned with the reference marks 38A and 38B of the large reference mark plate 17 as a reference, and the reticle R is aligned. Fix R (step 205). After that, from (Equation 1), the reference marks 38A, 3 of the large reference mark plate 17
A rotation error Rr of the reticle R and a magnification error Rm of the projection optical system PL based on 8B are calculated (step 20).
6).

Next, in step 207, the baseline amount in the X direction of the off-axis wafer alignment system 9 and the baseline amount in the X direction of the TTL type wafer alignment system including the LIA type light transmitting system 36X are measured. I do. Specifically, the positional deviation amounts obtained as a result of measuring the positions of the reference marks 38A and 38B of the large-sized reference mark plate 17 in the X direction by the TTR alignment systems 5A and 5B, respectively, are RAX1 and RAX2 and the large-sized reference mark plate 17, respectively. The amount of positional deviation when the position of the reference mark 39 in the X direction in the X direction is measured by the FIA.
X, the position shift amount when the position of the reference mark 40X for two-beam interference of the large-sized reference mark plate 17 is measured by the laser beam from the LIA type light transmission system 36X is defined as LIXA.
In this case, the magnification error Rm of (Equation 1) is used to calculate the baseline amount BE1x in the X direction of the wafer alignment system 9 and the baseline BE2x in the X direction of the alignment system including the LIA type light transmission system 36X as follows. Like
However, on the large fiducial mark plate 17, the fiducial marks 3
The difference in the X direction between the midpoint of 8A and 38B and the center of the reference mark 39 is L10, and the difference in the X direction between the midpoint of the reference marks 38A and 38B and the reference mark 40X is L20.

## EQU00002 ## BE1x = L10 + FIAX- (RAX2-Rm * M / 2) BE2x = L20 + LIAX- (RAX2-Rm * M / 2)

At this time, the wafer stage 6 is fixed in the X direction by applying servo so that the position of the reference mark 40X for two-beam interference becomes constant in the X-direction, and the reference mark 40Y for two-beam interference. Servo is fixed so that the position in the Y direction becomes constant in the Y direction. The same applies to the following step 208. Further, the average value Lx of the measurement values of the laser interferometer 19X at the time of the baseline check and the average value Lf of the measurement values of the laser interferometer 20X are calculated, and the difference A (= Lf−Lx) between them is calculated. This serves as a reference for obtaining the yawing error of the wafer stage 6 as described above.

Next, in step 208, the Y direction baseline amount of the off-axis wafer alignment system 9 and the Y direction baseline amount of the TTL type wafer alignment system including the LIA type light transmitting system 36Y are measured. I do. Similar to step 207, the positional displacement amounts obtained as a result of measuring the positions of the reference marks 38A and 38B in the Y direction by the TTR alignment systems 5A and 5B are RAY and RAθ, respectively, and the positions of the reference mark 39 in the Y direction are set. The amount of positional deviation measured by the wafer alignment system 9 is FIAY, and the position of the reference mark 40Y is LIA.
The amount of positional deviation when measured with the laser beam from the system light-transmitting system 36Y is defined as LIAY. In this case, using the rotation error Rr of (Equation 1), the baseline amount BE1y in the Y direction of the wafer alignment system 9 and the LIA type light transmission system 3 are used.
Baseline BE in the Y direction of the alignment system including 6Y
2y is as follows. However, on the large-sized reference mark plate 17, the difference in the Y direction between the midpoint of the reference marks 38A and 38B and the center of the reference mark 39 is L11, and the difference between the midpoint of the reference marks 38A and 38B and the reference mark 40Y in the Y direction. The difference is L21.

[Equation 3] BE1y = L11 + FIAY− (RAθ−Rr × M / 2) BE2y = L21 + LIAY− (RAθ−Rr × M / 2)

Next, in step 209, the baseline amount in the X direction of the wafer alignment system of the TTL method including the LSA light transmitting system 35X and of the laser step alignment method is measured. Specifically, when the wafer stage 6 is driven and the vicinity of the reticle mark 3A is scanned in the X direction by the light emission cross mark 41 of the large reference mark plate 17, the image of the light emission cross mark 41 is aligned with the reticle mark 3A.
Amount of positional deviation ISS from the design value of the position that matches the direction
X, and the positional deviation amount LSAX from the design value of the position where the reference mark 42X for slit scanning and the beam spot from the LSA type light transmitting system 35X match each other are obtained.

Further, in order to correct the Abbe error, the measurement value of the laser interferometer 19X is Lx, the measurement value of the laser interferometer 20X is Lf, and it is obtained in step 207 (Lf-Lx).
Is defined as A, the Abbe error Ab is defined by the following equation. Ab = Lf-Lx-A The difference in the X direction when this Abbe error Ab is averaged for a certain time before and after the scanning of the wafer stage 6 is AbX, and the difference in the X direction between the light emission cross mark 41 and the reference mark 42X is L3.
0, the position of the light emission cross mark 41 and the LSA type light transmission system 3
The difference in the Y direction from the beam spot position from 5X is ΔL
Letting SA be the distance between the laser beams LB1 and LB2 in the Y direction be FR, the baseline amount BE3x in the X direction of the wafer alignment system of the TTL method and the laser step alignment method is as follows.

## EQU00004 ## BE3x = L30 + LSAX-AbX.times..DELTA.LSA / FR- (ISSX-Rm.times.M / 2).

Similarly, the wafer stage 6 is driven to scan the vicinity of the reticle mark 3A in the Y direction with the light emission cross mark 41 of the large reference mark plate 17. Then, the positional deviation amount ISSY from the design value at the position where the image of the light emission cross mark 41 matches the reticle mark 3A in the Y direction, the reference mark 42Y for slit scanning, and the light transmitting system 35Y of the LSA system.
Based on the amount of positional deviation LSAY from the design value of the position where the beam spot from the position is matched, the baseline amount BE3y in the Y direction of the wafer alignment system of the TTL method and the laser step alignment method is as follows. However, the difference in the Y direction between the position of the light emitting cross mark 41 and the position of the reference mark 42Y is L31. In this case, there is no Abbe error, so the formula is simple. However, if the marks are arranged in the X direction under the same conditions as in the Y direction, it is not necessary to correct the Abbe error.

## EQU00005 ## BE3y = L31 + LSAY- (ISSY-Rr.times.M / 2)

Thus, the two-beam interference including the wafer alignment system 9 of FIG. 1 and the wafer alignment system of the laser step alignment system including the LSA type light transmitting systems 35X and 35Y and the LIA type light transmitting systems 36X and 36Y. A baseline amount for each type of wafer alignment system is determined. In FIG. 10, the operation when the reticle R is exchanged starts from step 204, and the operation during normal reticle alignment is step 2
The operation when starting from 05 and performing only the baseline check starts from step 206. Further, although steps 206 to 208 are separately executed in the above description, for example, the rotation error / magnification error and the baseline amount ((Equation 1) are performed by simultaneously performing the rotation / magnification measurement and the baseline measurement. (Equation 4) may be calculated at once.

Next, an example of the operation when the wafer W is aligned using the off-axis wafer alignment system 9 of FIG. 1 will be described with reference to FIG. Figure 1
1 is shot areas 86-1, 86-2, ... On the wafer W.
... and wafer marks formed corresponding to them are shown in FIG.
In FIG. 5, the wafer mark 87X in the X direction around the wafer mark 87X
Difference in the X direction between −5 and the center 88-5 of the shot area ΔX
Also, the difference ΔY in the Y direction between the wafer mark 87Y-5 in the Y direction and the periphery thereof and the center 88-5 of the shot area is set to a predetermined value by design.

Then, in FIG. 1, the wafer stage 6 is driven to move the shot area 86-5 of the wafer W to the lower portion of the wafer alignment system 9, and then the wafer alignment system 9 first detects an image of the wafer mark 87X-5. Is sandwiched between the index marks 78A and 78B of the index plate 75 of FIG. 7, and the X coordinate of the wafer mark 87X-5 is measured. Next, the image of the wafer mark 87Y-5 is sandwiched between the index marks 79A and 79B of the index plate 75 shown in FIG.
Measure the Y coordinate of 7Y-5. After that, when the pattern of the reticle R is exposed on the shot area 86-5 on the wafer W, the coordinates of the wafer marks 87X-5 and 87Y-5, the design differences ΔX and ΔY, and the above-described procedure are used for measurement. By driving the wafer stage 6 to the coordinates determined based on the baseline amount of the wafer alignment system 9, the center 88-5 of the shot area 86-5 is a conjugate point of the center of the reticle R, that is, the light of the projection optical system PL. Aligns with axis AX. As a result, exposure is performed with good overlay accuracy.

Next, the operation when the TTR type alignment systems 5A and 5B of FIG. 1 are arranged so as to be movable in the X direction will be described. First, the large reference mark plate 1 of FIG.
After the correction with reference to the reference marks 38A and 38B of FIG. 7, the running direction of the wafer stage 6 in the X direction is shown in FIG.
It is assumed that the direction is parallel to the straight line 84 in (a). Also, T
It is assumed that the loci of movement of the TR alignment systems 5A and 5B are 85A and 85B, respectively, and the alignment system 5A is fixed at two points on the locus 85A, for example. In this case, the width d2 of the positioning error in the Y direction and the width d1 of the positioning error in the X direction with respect to the straight line 84 of the alignment system 5A are both set to about several μm or less. In addition, as shown in FIG. 12B, at the first position of the locus 85A, T
The amount of deviation in the X direction between the reticle mark 3A observed in the TR alignment system 5A and the image 38AP of the reference mark, and the reticle mark 3A 'observed in the TTR alignment system 5A at the position next to the locus 85A. The amount of deviation in the X direction from the reference mark image 38AP 'may differ. However, if the widths d1 and d2 of the positioning error are each smaller than about several μm, the positional deviation amount in FIG. 12B can be accurately measured.

In the above embodiment, the large reference mark plate 1 is used.
The reference mark 39 of No. 7 is arranged on the bisector of the reference marks 38A and 38B. For example, the reference mark 39 is formed on a straight line connecting the reference marks 38A and 38B.
May be arranged. 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.

[0080]

According to the present invention, the first reference mark and the plurality of second reference marks serving as the plurality of reference marks for the mask are simultaneously detected by the first mark detecting means and the second mark detecting means, respectively. Since the reference mark plates that are arranged apart from each other so as to be detected are provided, the baseline amount of the off-axis alignment system can be measured with high accuracy without moving the substrate stage. Further, there is an advantage that the rotation amount of the reticle can be measured with high accuracy by using the plurality of second reference marks on the reference mark plate.

Further, when the rotation amount of the mask is obtained with reference to the plurality of second reference marks on the reference mark plate, the mask substrate is obtained from the rotation amount and the inclination between the reference mark plate and the coordinate system of the substrate stage coordinate measuring means. The rotation error of the stage coordinate measuring means with respect to the coordinate system can be measured with extremely high accuracy and at high speed.

Further, the second mark detecting means is movably arranged along a predetermined axis, and the second mark detecting means is arranged along the coordinate axis of the substrate stage coordinate measuring means in which the inclination calculated in the above procedure is corrected. When the second mark detecting means is moved, the baseline amount can be measured with high accuracy even when the second mark detecting means is moved.

[Brief description of drawings]

FIG. 1 is a partially cutaway perspective view showing a main part of a projection exposure apparatus according to an embodiment of the present invention.

2A is an enlarged plan view showing a large reference mark plate 17, and FIG. 2B is an enlarged plan view showing a reticle mark 3A or 3B.

FIG. 3 is a block diagram showing a configuration of a TTR type alignment system 5A.

FIG. 4A is a plan view showing a reticle mark 3A,
(B) is a plan view showing an image 38AP of the reference mark 38A,
(C) is a plan view showing an example of an observation screen of the alignment system 5A of the TTR system.

FIG. 5 is a block diagram showing a configuration of a TTL type wafer alignment system.

FIG. 6 shows an off-axis type wafer alignment system 9
3 is a block diagram showing the configuration of FIG.

7 is a diagram showing an index plate 75 of the wafer alignment system 9. FIG.

8A and 8B are explanatory views in the case of measuring a yawing error of a wafer stage in the embodiment,
(C) is an explanatory diagram for obtaining a rotation error between the large reference mark plate 17 and the reticle R.

FIG. 9 is an explanatory diagram of the corrected running of the wafer stage.

FIG. 10 is a flowchart showing an example of an operation at the time of baseline check in the embodiment.

FIG. 11 is an enlarged plan view showing the arrangement of shot areas on the wafer W of the embodiment.

FIG. 12 is an explanatory diagram when the TTR type alignment systems 5A and 5B are arranged so as to be movable.

FIG. 13 is a front view showing a main part of a projection exposure apparatus including a conventional off-axis type wafer alignment system.

[Explanation of symbols]

 R Reticle PL Projection optical system W Wafer 2 Reticle stage 5A, 5B TTR type alignment system 6 Wafer stage 9 Wafer alignment system 14A, 15A, 14B, 15B Imaging device 16A, 16B Light receiving device 17 Large fiducial mark plate 18X, 18Y Moving mirror 19X, 20X, 21Y Laser interferometer 23 Reflection prism 24 Objective lens 27X, 27Y Imaging device 35X, 35Y LSA type light transmitting system 36X, 36Y LIA type light transmitting system 38A, 38B Reference mark 39 Reference mark 40X, 40Y 2 light flux Reference mark for interference 41 Light emitting cross mark 42X, 42Y Reference mark for slit scanning 50 Main control system 51 Reticle stage drive system 52 Wafer stage drive system

Claims (3)

[Claims] 1. A mask stage for positioning and fixing a mask having a pattern to be exposed and a plurality of marks for alignment, and a photosensitive substrate having a plurality of substrate marks formed thereon for holding the photosensitive substrate. A substrate stage for positioning, a projection optical system for image-projecting the pattern of the mask onto a region near each substrate mark on the photosensitive substrate on the substrate stage, and a substrate on the photosensitive substrate without passing through the mask A first mark detecting means for detecting a mark; and a first mark detecting means for detecting a positional deviation amount between the conjugate image of the reference mark for the mask arranged on the substrate stage and the mark of the mask via the projection optical system. In the method of positioning the substrate stage of a projection exposure apparatus including two mark detection means, the first mark detection method is provided on the substrate stage. A first fiducial mark detected by a step and a plurality of second fiducial marks as the fiducial marks for the mask are separated so as to be simultaneously detected by the first mark detection means and the second mark detection means, respectively. And a reference mark plate that is arranged as a substrate, and a substrate stage coordinate measuring unit that measures the moving coordinates and the amount of rotation of the substrate stage, and measures the inclination between the reference mark plate and the coordinate axis of the substrate stage coordinate measuring unit in advance. A positioning method of the projection exposure apparatus, wherein the running coordinate axis of the substrate stage is defined with reference to the reference mark plate. 2. The position deviation amount between the conjugate image of the plurality of second reference marks on the reference mark plate and the plurality of marks on the mask is detected by the second mark detecting means, and the detected position is detected. 2. The positioning method of the projection exposure apparatus according to claim 1, wherein the tilt between the mask and the coordinate axis of the substrate stage coordinate measuring means is calculated from the amount of deviation and the tilt measured in advance. 3. The second mark detecting means is movably arranged along a predetermined axis, and the second mark detecting means is arranged along the coordinate axis of the substrate stage coordinate measuring means in which the calculated inclination is corrected. 3. The method for positioning a projection exposure apparatus according to claim 1, wherein the mark detecting means is moved.