JPS60186845A - Aligning device of exposure device - Google Patents

Abstract

PURPOSE:To improve the superposing accuracy of a pattern transferred onto a substrate to be exposed, by correcting both reticle rotation and chip rotation. CONSTITUTION:Two reticle alignment microscopes 16 and 17 are provided to align a reticle R with the main body of an exposure device. Moreover, two laser step alignment systems 18 and 19 are provided to detect a prescribed mark on a wafer W with a laser light by a through-the-lens system and make the alignment between the wafer W and main body of the device. A die-by-die alignment system 20 observes the mark Rs on the reticle R and the mark on the wafer W under a superposed condition. In addition to the above, two wafer alignment microscopes of an off-axis system are provided at prescribed intervals from a projecting lens 1.

Description

DETAILED DESCRIPTION OF THE INVENTION (Technical Field of the Invention) The present invention relates to an exposure apparatus for manufacturing semiconductor devices such as ICs and LSIs, and particularly relates to an exposure apparatus for manufacturing semiconductor elements such as ICs and LSIs, and in particular, for manufacturing semiconductor elements such as ICs and LSIs. The present invention relates to a step-and-repeat exposure apparatus that superimposes a circuit pattern of a mask on a circuit pattern of a semiconductor element using a step-and-repeat method for exposure.

(Background of the Invention) In recent years, so-called step-and-repeat exposure equipment, which is equipment that repeatedly exposes circuit patterns of masks on semiconductor wafers (hereinafter referred to as wafers), has been used in large numbers at sex shops to manufacture semiconductor devices. It has a great effect. In particular, reduction projection type exposure equipment that transfers the circuit pattern of a mask onto a wafer via a reduction projection lens projects the circuit pattern of the mask in reduction, so minute foreign matter adhering to the mask may be affected by the 'plIj & force of the reduction projection lens. Furthermore, since the circuit patterns on the mask are superimposed on each of the multiple circuit patterns formed on the wafer and exposed, an accurate camera can always be maintained even when the wafer expands and contracts during various processes. It also has the advantage of being able to be combined,
It can be said that this is an extremely productive exposure device. This type of exposure apparatus commands a two-dimensional moving stage on which a substrate is placed, and performs exposure by moving this stage at a constant pitch according to a fixed coordinate system relative to the circuit pattern image of the mask. This is repeated.For this reason, circuit patterns are arranged in a matrix on the wafer.

The array coordinates of the circuit patterns on this wafer follow the coordinate system of the stage during exposure, but when the wafer is replaced, a rotational shift occurs between the array coordinates on the wafer and the coordinate system of the stage.

Although this rotational deviation can be corrected by rotating the wafer, a small rotational deviation still remains.

This minute rotational deviation between the array coordinates on the wafer and the coordinate system of the stage is called wafer rotation. In order to achieve more precise fiber alignment exposure, the present applicant disclosed a method for correcting wafer rotation in Japanese Patent Laid-Open Publication No. 102823/1983, which was filed earlier. The miniature projection type exposure apparatus disclosed in this publication uses a step-and-repeat method in which a reticle (corresponding to 4 masks) on which a 0x enlarged pattern of the circuit pattern on the top is drawn is scanned once. During exposure, the reticle was mounted without any rotational deviation relative to the coordinate system of the stage.

However, when the size of the circuit pattern to be transferred onto the wafer becomes small and the line width becomes about 1 μm, minute rotational deviations of the reticle cannot be ignored. If the reticle is slightly rotated, not only will the accuracy of alignment with each circuit pattern on the wafer deteriorate, but when the first layer of circuit patterns are transferred onto the wafer, the arrangement coordinates on the wafer will However, the drawback is that it is not possible to obtain semiconductor devices satisfying the characteristics of the transfer stage with high productivity.

(Object of the Invention) The present invention solves the above-mentioned drawbacks, minimizes the rotational deviation between the circuit pattern on the mask (reticle) and each circuit pattern on the exposed substrate (wafer), and improves the rotational deviation between the circuit patterns on the mask (reticle) and the circuit patterns on the exposed substrate (wafer). 14 It is an object of the present invention to obtain a step-and-repeat exposure device that realizes l1lx overlapping exposure.

(Summary of the Invention) The present invention includes a mask holding means (reticle stage 2) for holding a mask (reticle R) having a first pattern (circuit pattern or alignment mark);
A second pattern that can be aligned with the Rotation error detection means (LSA system 19, DDA system 2) that detects a rotation error (Δθ)
0.20'), a rotation means (wafer holder 10) for rotating either the reticle R or the wafer W so that the rotation error is corrected, the first pattern whose rotation has been corrected, and a plurality of 8g2 The technical point is to provide a control means (gromonocer 60) for controlling the wafer stage so as to sequentially align each of the patterns.Furthermore, the present invention holds the reticle R and moves it to a predetermined orthogonal coordinate XY. The reticle stage 2 and the second pattern described in IIJ move along the array coordinates αβ and θ[
A wafer stage for holding a plurality of wafers W formed at regular intervals (pinch) and moving the wafers two-dimensionally along the orthogonal coordinates XY, and a rotation error detection means (KrJ).
LSA system 19, DDA system addition, 20'), the rotation means (wafer holder lO), and the rotation error (Δθ) when the wafer stage is positioned according to the arrangement pinch of the second pattern on the wafer W. A positional deviation detector 79 (LSA system 18,
19 or DDA system 20) and a control means (glomonoser 60) for moving either the wafer stage or the reticle stage 2 so as to correct the positional deviation.

(Example) Next, a schematic configuration of a reduction projection type exposure apparatus to which an example of the present invention is applied will be described based on FIGS. 1 and 2.

A reticle (mask) R on which a certain pattern is drawn by the transparent part and the light-shielding part is illuminated with a uniform intensity of light having a wavelength that sensitizes the photosensitizer.Thereby, the pattern area P r of the reticle R The optical image is projected onto the wafer W by a reduction projection lens (hereinafter simply referred to as a projection lens) 1.The reticle R is placed on a reticle stage 2 as shown in FIG. projects the light that has passed through 'r onto the pattern area of reticle R through lens 1.
An opening 2a is provided for allowing the light to enter the reticle R, and a holding portion 2b for vacuum suctioning the peripheral portion of the reticle R is provided. The deflected drive unit 3 moves the reticle stage 2 slightly in the X direction, and the drive unit 4 moves the reticle stage 2 slightly in the Y direction perpendicular to the X direction.
Perform dimensional alignment.

On the other hand, the Y stage 6 is moved in the Y direction by the drive unit 5.
An X stage 8 that moves in the X direction on this Y stage 6 by a drive unit 7, two stages 9 that can move up and down in two directions with respect to this X stage 8, and , a two-dimensional moving stage (hereinafter referred to as wafer stage) includes a wafer holder 10 that vacuum-chucks the wafer W and rotates minutely by a drive unit 11.
is configured. The second stage 9 is moved downward by a drive section (not shown) provided on the X stage 8. A light wave interferometer (hereinafter referred to as a laser interferometer) 12.14 using laser light is provided to detect the position of the wafer stage. The mirror 13 is provided on one side of the two stage 9 with its reflection plane extending in the Y direction. In addition, the mirror 15 extends its reflection plane in the X direction to form a second stage 9.
It is located on one side of the.

Therefore, the laser interferometer 12 irradiates the mirror 13 with a laser beam Bx having an optical axis parallel to the coordinate axis X, and
A fixed mirror provided inside the laser interferometer 12 is also irradiated with a laser beam, the reflected beam from the mirror 13 and the reflected beam from the fixed mirror are caused to interfere with each other, and changes in the interference fringes that occur are detected by optical distortion. By this, the X of the wafer stage
Detect the position of the direction. Similarly, the laser interferometer 14 also has a laser beam BY which has an optical axis M parallel to the coordinate axis Y on the mirror 15.
, and the reflected light flux from the internal fixed mirror and mirror 1
The position of the wafer stage in the X direction is detected by interference with the five reflected light beams. The projection lens 1 and the laser interferometers 12 and 14 are arranged so that the laser beam Bx and the laser beam By are orthogonal to each other in the same plane, and the optical axis AX passes through the intersection.

Now, this exposure apparatus is basically provided with four alignment optical systems and a detection system. One is to move the reticle R to the reticle body (as shown in Figure 1).
For example, two reticle alignment microscopes (hereinafter referred to as
(referred to as R-MIC) 16 , 17 . R-MIC
16 is an area on the reticle R where a predetermined pattern is drawn (
Observe the reticle mark Rxy provided around the pattern area Pr (hereinafter referred to as pattern area), and the R*M I C17 aligns the reticle mark RXy, which is different from the reticle mark Rxy, with the index in the R-MIC 16 around the pattern area pr. Positioning of the reticle R in the X direction and the Y direction is achieved by operating the drive units 3 and 4 as shown in FIG. The reticle R is moved by operating the drive units 3 and 4 to match the index inside.
rotational positioning is achieved. These R-M I
C16 and C17 may be used for visual alignment (alignment), but they may also be used for automatic position 1g alignment by photoelectrically detecting the observed images of the reticle marks Rxy and Rθ.

In addition, the optical system for alignment and the other detection thread are @
As shown in Figure 1, two lasers are used to detect a predetermined mark (Unino mark) on the wafer W using a through-the-lens (TTL) method with laser light and align the Unino W with the main body of the device. Stamp alignment system (hereinafter referred to as LSA system) 18 , 19 . LSA system 18 uses light of a wavelength that sensitizes a photosensitizer (photoresist), that is, a laser beam (for example, helium neon laser) that does not sensitize the photoresist at a wavelength different from the exposure light.
Mirror 18a is aligned perpendicularly to the optical axis AX of the projection lens l.
Project towards. mirror], 8a is the laser beam L
AX upward, that is, the bottom surface of the reticle R (projection lens 1
(the surface facing the surface). , the mirror 18b has a reflective surface parallel to the lower surface of the reticle R, is placed a predetermined distance from the lower surface of the reticle R, and directs the laser beam from the mirror 18a toward the center of the entrance pupil of the projection lens 1.2. reflect. In this embodiment, the projection lens 1 is assumed to be a non-telecenter link optical system on the object side (reticle R side) and a telecenter link optical system on the image side (wafer W side). Therefore, the optical axis of the laser beam LAx is
By providing the mirror 18b around the reticle R, that is, at a position away from the optical axis AX of the projection lens l, the mirror 18b is tilted at a predetermined angle with respect to the optical axis AX. However, on the U/'W side of the projection lens l, the optical axis of the laser beam LAX is parallel to the optical axis AX. LSA system 19, mirror 19a,
19b is similarly provided at a position spatially rotated by 90 degrees around the optical axis AX with respect to the LSA thread 18 and mirrors 18a and 18b, and the laser beam LAy is incident on the projection lens 1. Then, these two LSA systems 18 and 19 detect the alignment mark in the X direction and the alignment mark in the Y direction provided on the sea urchin ISW without going through the reticle R. Local region alignment is performed.

Now, the other optical system for alignment and detection system is die-pie-die (1) ie by which superimposes mark R3 on reticle R and mark on Unino 1W through projection lens 1 and observes them. Die) alignment system (hereinafter referred to as
(referred to as DDA system) 20. 1) The DA system 2o irradiates the mark Rs provided around the pattern region Pr of the reticle H with illumination light having the same wavelength as the exposure light, which is bent at right angles by the mirror 20a. The illumination light from the mirror 20a not only illuminates the mark Rs, but also passes through the projection lens l and illuminates a minute area including the mark on the Unino W. Then, the image of the mark on Unino W that is back-projected by the projection lens 1 and the image of the mark R8 of the reticle R are reflected by the mirror 20a and formed by the DDA system 20, and are formed on the pattern area Pr of the reticle R and the wafer W. The DDA system 20 detects the state of alignment with one local area to be exposed.The DDA system 20 is also not aligned, and the mark image is photoelectrically detected using a slit scanning photoelectron microscope or an image pickup tube (element). It is configured so that automatic alignment can be performed.

In the die Φ-by-die alignment method, when the projected image of the pattern region Pr of the reticle R and one local region on the wafer W are accurately overlapped, the mark on the wafer W and the mark RS of the reticle R are aligned in a predetermined positional relationship and observed by DDA system addition. For this reason, DD
After the wafer stage and reticle stage 2 are slightly moved so that they are aligned in the A system, the exposure operation can be started immediately. Here, it is assumed that the DDA system of the water device detects the positional deviation of the observed mark image of the reticle R1 wafer W in the X direction.

As shown in FIG. 2, the final alignment optical system and detection system consists of two off-axis wafer alignment microscope mirrors (hereinafter referred to as W -MIC) There are 21.22 te. This (02) W-M I C21.

22 were determined in advance at predetermined intervals and were provided at two representative locations on the wafer W soil. This is for detecting marks and performing global alignment of the wafer W with respect to the apparatus.

−“　　,　.

By the way, in FIG. 1, the alignment of the reticle R, the global alignment of the wafer W, and the alignment of the wafer W are shown.
- Alignment for each local region of
A reference mark plate 30 is provided on the wafer boulder 19, and is used for adjustment of each optical system and detection system.

This reference mark plate 3() is made by providing a light-reflecting chromium layer on the surface of a glass substrate, and etching necessary marks on the chromium layer.

Figure 2 shows the laser desso alignment system and W-M I.
FIG. 3 is an optical layout diagram showing a specific configuration of C21 and C22.

Although only LSA yarn 18 is shown here, LSA yarn 1
The same is true for 9. A laser beam emitted from a laser light source 40 such as a helium neon laser has a beam cross section expanded to a predetermined size by a beam expander 41.
The cylindrical lens 42 shapes the light beam into a light beam with a rectangular cross section. The shaped laser beam is reflected by a mirror 43 and reaches the mirror 18a via a lens 44, a beam sinter 45, and a lens 46. Mirror 18a
The reflected laser beam passes through the field stop 18c, reaches the mirror 18b, is further reflected by the mirror 18b, and proceeds toward the center of the projection lens 1 (iJila).

The field stop 18c is arranged at the imaging position of the laser beam reflected by the mirror 18a. At this imaging position, the laser beam is converged into a band-shaped spot light extending in a direction perpendicular to the plane of the paper in FIG. That is, due to the action of the cylindrical lens 42, the mirror 18a is removed from the lens 46.

The laser beam that passes through 18b and enters the projection lens 1 once reaches the field diaphragm within the plane of the paper in FIG. ) 18 It becomes a light flux that converges (images) at c and then diverges, and becomes a parallel light flux in a plane perpendicular to the plane of the paper. Then, this laser beam passes through the center of incidence -1a of the projection lens 1, and is imaged on the imaging plane FP of the projection lens 1 as a strip-shaped and elongated spot light.

The light intensity distribution of this spot light is elongated in the horizontal direction within the plane of the paper of FIG. 2, and becomes short in the direction perpendicular to the plane of the paper.

Now, the mirrors 18b are provided at predetermined intervals on the lower surface of the reticle R, but their positions are from the maximum projection area of the reticle, that is, from the optical axis AX of the projection lens 1 in this embodiment as shown in FIG. Reticle mark R at the farthest position
It is determined so that the projection optical path of xy, Rθ (indicated as light ray ELm in FIG. 2) is not blocked. Furthermore, when the reticle R is illuminated with exposure light, the image of the pattern region Pr of the reticle R is reduced and formed on the imaging plane FP of the projection lens 1 like the light beam EL. In FIG. 2, the surface of the fiducial mark plate 30 and the imaging plane FP coincide, and furthermore, this fiducial mark plate 30 is
The arrangement is shown so that it is irradiated with a laser beam from the SA system 18. As shown in FIG. 3, a lattice-shaped mark FM is formed between the reference mark plates and is irradiated with a spot light SP of the laser beam LAX to generate diffracted light. FIG. 3(a) is a plan view showing the two-dimensional positional relationship between the mark FM and the spot light SP, and FIG. 3(b) is a cross-sectional view of the mark FM. The mark FM is formed by regularly arranging minute linear elements Seg in a line, and the arrangement direction thereof is determined to coincide with the longitudinal direction (extension direction) of the spot light sP. When the spot light sP overlaps with the mark FM, as shown in FIG. 3(a), depending on the arrangement pitch of the linear elements Seg and the wavelength of the laser beam, there is a Higher-order diffracted lights such as second-order diffracted light ±D and second-order diffracted light ±Dt are generated. These diffracted lights. .

The beam splinter 45 receives the diffracted light 1) from the mark FM.

+I)+ ID2 is reflected and the diffracted light is directed to a detection system consisting of a spatial filter 47, a condensing lens 48, and a photoelectric detector 49. The spatial filter 47 is the projection lens 1
is placed at a position conjugate with the incident pi 1 a, and the diffracted light D
o 11) 0th order diffraction light of l tDt. That is, the specularly reflected light is blocked (IE)iL, and the other first-order diffracted light D1 and second-order diffracted light are transmitted. A condensing lens 48 condenses these high-order diffracted lights that have passed through the spatial filter 47 onto a photoelectric detector 49 .

By the way, as shown in FIG. 1, another set of laser beam alignment systems is provided as an LSA system 19. Therefore, the positional relationship between the projected image and the spot light SP on the imaging plane FP of the projection lens l will be explained with reference to FIG. In FIG. 4, a circular area if is an image field that can be projected to the maximum extent by the projection lens 1, and a rectangular area Pr' within the area if is a pattern area Pr of the reticle R in FIG.

This is the projection area. It is assumed that the optical axis AX of the projection lens 1 passes through the center of the circular area 100, and that the center of the rectangular area Pr' coincides with the optical axis AX. Then, the moving coordinate system of the wafer stage, that is, the x of the XY coordinate system, is set such that this center is the origin.
, the y-axis is set, the spot light SPx formed by the LSA system 18 is a long and narrow spot on the Y-axis, and the spot light SPy formed by the LSA system 19 is
It becomes a long and thin spot on the axis.

Spot light SPx Distance L in the Y direction from the OX axis
y and the distance Lx of the spot light SPy from the Y1 pongee in the X direction are predetermined at the time of manufacturing the device. Moreover, the distances Ly and Lx are rectangular (rectangular or square) areas Ar where the spot lights SPx and spy are inscribed in the circular area if.
It is determined to be located outside of the area if and inside the area if.

The area Ar' represents the maximum area that the projection lens 1 can project as a rectangle. Now, SPx detects the mark FM extending in the Y direction on the fiducial mark plate 30 or the mark extending in the Y direction on the wafer W, and detects the position IK deviation of the fiducial mark plate 30 or the wafer W in the X direction. Spot light S
Py detects marks extending in the X direction on the fiducial mark plate 30 and the wafer W, and detects positional deviation of the fiducial mark plate 30 and the wafer W in the Y direction.

In addition, the off-axis type W-MIC2 shown in Figure 2
1 and 22, for example, as disclosed in Japanese Unexamined Patent Publication No. 57-19726, a lattice-like mark formed on a wafer W is irradiated with an elongated strip-like spot of a laser beam, and the diffracted light from the mark is beamed. 'Turtle detected, W-MI of wafer W
This is to detect positional deviations regarding C21 and C22.

The band-shaped spot lights of the W-MICs 21 and 22 are adjusted in advance so as to form an image on the imaging plane FP of the projection lens 1.

Now, Figure 5 shows the off-axis W-M I
3 is a plan view showing the arrangement of band-shaped spot lights C21 and C22 and an image field if of the projection lens 1. FIG. The origin of the moving coordinate system XY of the two-dimensional moving stage is projected by the projection lens l.
When the W-MIC 21 is aligned with the optical axis AX of the
-The band-shaped spot light WYS of MIC21 is directed in the X direction (
Formed with a long mouth.

The spot light wys is arranged so that the center of the spot light wys in the X direction is separated by an interval. The irradiation position of the spot light WθS in the Y direction can be adjusted so that the line segment connecting the spot lights WYS and WO3 is parallel to X11+11 or at a predetermined angle with the y axis. Specifically, as shown in FIG. 6, the W-MIC 22 emits a parallel laser beam l. is incident, and the luminous flux lo is converged by the objective lens 22a and shaped into a spot light WθS.

At this time, a laser beam l enters the objective lens 22a.

is called parallel flat glass (hereinafter referred to as )・-ping glass)
22b to shift within a predetermined angle range.

This results in a luminous flux l. is deflected as a luminous flux 4.4, and the spot light WθS is displaced in the Y direction within a range of ΔY.

Now, in FIG. 5, the interval between the spotlights WYS and WO3 is determined so that marks provided at two representative locations on the wafer W can be detected simultaneously. In other words, two marks are provided on the wafer W separated by an interval so that the marks on the wafer W can be detected simultaneously by the W-1vilC21°22. These two marks each have a lattice-like pattern extending in the X direction, and the marks correspond to the urchin 2,
Used to detect positional deviation in the WΩ°C direction.

Now, FIG. 7 is a plan view showing the mark pattern of the reference mark plate 30 provided on the wafer holder 10. The reference mark plate 30 is provided with grid-like marks FMM and FMx parallel to each X-axis and Y-axis of the coordinate system XY, and xllll, y! Linear mark 30Y parallel to I+h
, 30X are provided in a crossed manner.

Marks FMX and FMY generate diffracted light by irradiation with spot lights SPx and SPy of I and SAAlO249 and spot lights wys and wθS of W-MIC21 and 22, and are used for LSA systems 1.8 and 19 and W- Used for adjustment, calibration, and various measurements of MIC21°22. Marks 30X and :30Y are used for adjustment, calibration, and measurement of R-MIC16, 17, DDA system marks, etc.

On the other hand, FIG. 8 is a plan view showing the mark arrangement of the reticle R suitable for this embodiment. In FIG. 8(a), the center of the pattern area Pr of the reticle R is set as RC (reticle center), and a coordinate system XY is determined so that the center RC becomes the origin. Assuming that the reticle R is not rotated with respect to this coordinate system XY, the mark Rxy is located on XQMII as a linear cross pattern as shown in FIG. IK is performed as a linear pattern that extends to the top.

Of course, the distances of Rxy and Rθ from the center RC are values determined in advance in the design of the reticle R. And Mark R
There is a mark R8711 on the side facing the center RC of xy.
The mark R8 is simply provided with a rectangular light-transmitting window in the light-shielding part, and as shown in Figure 8 (Q), the mark R8 is provided on the Edge R8I, R
82 is used for alignment or positional deviation detection depending on whether it is a DDA system. In Figure 8 (C), mark R is added to the DDA system.
8 and the mark 30Y of the fiducial mark plate 30, and when the interval between the edge R8I and the mark 30Y becomes equal to the interval between the edge R82 and the mark 30Y, the fiducial mark plate 30, In other words, alignment between the Unino stage and the reticle R has been achieved.

Note that the center lines of the amplifiers R8I and R82 of the mark RS are set to coincide with the X axis.

FIG. 9 is a circuit block diagram of the control system of this device. A processor (hereinafter referred to as CPU) 60 such as a microcomputer or a minicomputer centrally controls the operation of the entire apparatus via an interface circuit (hereinafter referred to as IFC) 61. A laser stamp alignment system processing circuit (hereinafter referred to as LSAC) 62 receives the photoelectric signal from the LSA optical detector 49 shown in FIG. 2, the laser interferometer 12,
Based on the position information from 14, for example, the amplifier pulse and down pulse generated for each unit movement amount (0°02 μm) of the two-dimensional movement stage, the intensity distribution of the diffracted light from the mark etc. on the wafer W is adjusted to the position. The position of the mark relative to the spot lights SPx and SPY is detected. However, the photoelectric signal corresponding to the diffracted light generated from the mark by the irradiation of the spot light SPx is caused by laser interference.
The photoelectric signal corresponding to the diffracted light generated from the mark by the irradiation of the spot light SPy is sampled in response to the up-down pulse of the laser interferometer 14.

A reticle alignment system processing circuit (hereinafter referred to as R-ALG) 63 works together with the R-MIC 16° 17 shown in FIG.
Mark Rxy for the index in C16, 17.

The amount of deviation of one line in the 170th place is calculated, and the drive units 3 and 4 of the reticle stage 2 are controlled in a feedback manner so that the amount of deviation becomes zero. The sweet R-MICs 16 and 17 are used to print marks 30 on the reference mark plate 30 through the projection lens 1.
X, 30Y (in some cases marks FMx, FMy) can also be detected, and at this time R-ALG63 is R-MIC
Measure the positional deviation of marks 30X and 30Y with respect to the indexes in 16 and 17, and send the results to the CPU 6 via the IFC 61.
Send to 0. Die・Pi・↓ Die alignment system processing circuit (hereinafter referred to as DDA) 64
moves together with the DDA system 20 in FIG. 1, observes the mark on the wafer W or the mark 30Y on the reference mark plate 30, and the mark R8 on the reticle R, and captures both mark images with light tq. This is to detect a relative positional shift between both marks in the Y direction.

Wafer alignment processing circuit (hereinafter referred to as WAC)
Reference numeral 65 detects the position iff of the mark on the wafer W in the Y direction with respect to the bond lights wys and wθS based on the optical signals from the W-MICs 21 and 22 shown in FIG. Incidentally, the WAC 65 also includes a function for rotating the no-ping glass 22b shown in FIG. 6 in response to a command from the CPU 60.

Now, the CPU fiO controls the wafer stage's X-direction drive unit 7 (hereinafter referred to as X-ACT7) and Y-direction drive unit 5 (hereinafter referred to as Y-ACT5) via the IFC 61.
, and a drive unit 11 (hereinafter referred to as θ-ACT 11) for rotating the wafer holder 19, respectively, output a drive command with a constant nozzle. The number of drives is also determined according to the calculations of the cPU 60 and the amount of positional deviation detected by various alignment processing circuits. Terminal #6 is also used by the operator to give instructions and display the progress of the device's operation.
6 is provided, and this terminal device 66 operates as a man-machine interface with the CPU 60. The IF 61 is provided with an X counter (not shown) that reversibly counts the amplifier down pulses from the laser interferometer 12 and a Y counter that reversibly counts the amplifier down pulses from the laser interferometer 14. By reading the values on both hands of the counter, the 2'-dimensional position (coordinate value) in the coordinate system Performs position detection and positioning.

Next, the operation of this embodiment will be explained based on the schematic flowchart shown in FIG. FIG. 10 shows the basic operations from alignment of the reticle R1 wafer W to the exposure trench. Below, the process is shown in step 200.
213 will be described, but first, the case where a first layer circuit pattern is transferred onto UNINO-W will be described.

Step 200: Measure the projected position of the detection center (index) of the R-MIC 16 in the coordinate system XY on the reticle R placed on the reticle stage 2 as shown in FIG. This situation will be explained based on FIG. 11.

FIG. 11 shows an image 16a of the index of the R-MICs 16 and 17 projected into the image field if of the projection lens 1.

17a and the arrangement of the reference mark plate 30 are shown.

It should be noted that a projected image Pr' of the pattern region Pr and a projected image Rs/ of the mark RS are shown in the image field controller when the reticle R is aligned without any positional deviation including rotational errors. The index image 16a of the R-MI C16 is a cross-shaped mark 30X on the reference mark plate 30.
, 30Y or the cross-shaped mark RXy of the reticle R, and the index image 17a of the R-MIC 17 sandwiches the mark Rθ of the reticle R or the mark 30Y of the reference mark plate 30. It is shaped like this. Now, in this step 200, the index image 1 of R-MIC16
The position of 6a in the Y direction is measured by a laser interferometer 14. In addition, in order to simplify the following explanation, the laser interferometer 12
, 14 is the origin of the coordinate thread XY when the intersection of the marks 30X and 30Y on the reference mark plate 30 is positioned to pass through the optical axis AX of the projection lens 1. shall be. That is, in that state,
The count value of the X counter for the laser interferometer 12 provided in the IFC 61 and the count value of the Y counter for the laser interferometer 14 are set in advance so that both become zero. The center of the index image 16a is not always located on the X axis as shown in FIG. 11 due to long-term drift and setting accuracy. Therefore, the CPU 60 moves the wafer stage to the origin position (where the intersection of the marks 30X and 30Y matches the optical axis AX).
After moving in the X direction from ffO, R-A L G 63
Marked by 30X.

30Y and the index image 16a are aligned.
Drive ACT7 and Y-ACT5. After the alignment is achieved, the CPU 60 reads and stores the Y coordinate value of the Unino stage at that time, that is, the value of the Y counter for the laser interferometer 14. Let the Y coordinate value be Y.

Step 201 Next, the CPU 60 loads the reticle R onto the reticle stage 2, vacuum-chucks it, and then loads the reticle R onto the reticle stage 2.
Align the reticle R by slightly moving the . The reticle stage 2 is moved by the drive units 3 and 4 via the R-ALG 63 so that the index image 16a and the mark Rxy of the reticle/l/R overlap, and the index image 17a and the mark Rθ of the reticle R overlap. This is achieved by

After achieving this reticle alignment, the reticle stage 2 is vacuum-adsorbed to a base (not shown). Also, the 1st Il!
As shown in FIG. 12, the eye reticle R is provided with marks Wy, Wθ, marks Sx, Sy, and Sy' used for wafer alignment, laser step alignment, etc. during electrical alignment of the second and subsequent layers. ing.

The marks Wy and Wθ are grid-like patterns, macroscopically linear marks extending in the X direction, and are provided on two sides of the reticle R parallel to the X axis. Mark SX is Y1
It is a lattice pattern extending over llll, and the marks sy,
Both sy' are lattice patterns extending on the X axis. The marks Sy% Sy' are provided on both sides of the center RC on the X-axis.

Step 202 Next, the c P u 60 measures the residual rotation error of the reticle R with respect to the coordinate system XY, that is, the reticle rotation (hereinafter referred to as RR). First, the wafer stage is moved to a position where the index image 16a and the marks 30X, 30Y are properly aligned. At this time, the wafer stage is positioned so that at least the position in the Y direction corresponds to the previously stored Y coordinate value Y. Then c PU60 marks 30X, 3
The Unino stage is moved in the X direction so that Y coordinate value Y1 does not change so that 0Y overlaps the projected image Rs/ of mark R8 on reticle R. The image Rs' of the mark R8 overlaps the mark 30Y and is observed as shown in FIG. If the reticle is inserted accurately, the amount of reticle rotation is zero.However, there is generally a rotation error, albeit a small angle.Therefore, mark R8 and mark 30Y have a relative K
It is observed displaced in the Y direction. Therefore, Dl) A-based Canada, D
The DAC 64 detects the positional deviation in the Y direction between mark R8 and mark 30Y, and the UNINO stage is slightly moved in the Y direction so that only that position is removed. And C.P.
U60 stores the Y coordinate value Y when the positional shift is eliminated. Therefore, L
Assuming that X, the CPU 60 calculates the rotation amount OR of the reticle R with respect to the coordinate system XY using the equation (2).

eR=tan-' (XJ two knee XL) ---
(LX) However, since the rotation htθR is extremely small, an approximation calculation is performed using equation (2).

θH=sin ”(-Kouji1-Toyamajigo......(
2) LX LX This rotation angle θB represents the position of the line segment connecting the center of the mark Rxy of the reticle R and the center of the mark R8 with respect to the X axis of the coordinate system XY. This completes the RR measurement.

Step 203 Next, the CP and U 60 control a transport device (not shown) to place the wafer W on the wafer holder IO. At this time, the wafer W is pre-aligned using a linear notch (hereinafter referred to as a flat) provided around the wafer W so that the flat and the X axis of the coordinate system XY are parallel to each other.

Next, the CPU 60 determines that the wafer W to be exposed is
It is determined whether the wafer is the first wafer or not. Normally, in the manufacture of semiconductor devices, a plurality of wafers that undergo the same process are managed as a lot (t, oT).

Therefore, the wafers within one lot have little variation during manufacturing.

Here, assuming that the first wafer in the lot is to be exposed, Cf'U6o executes the next step 205.

Step 205 Next, the CPU 60 determines that the pattern to be exposed is the first layer (1s
t) to determine whether it is old. Here, the first
Since the layer reticle R is attached, the CPU 60 executes the next step 206.

Step 206 Now, in this coordinate determination A, the CPU 60 corrects the stepping coordinates of the wafer stage using the step-and-repeat method by tilting the reticle's rotational amount θR with respect to the coordinate thread XY.

This situation will be explained with reference to FIG. In FIG. 13, the projected image Pr' of the pattern area Pr of the reticle R is tilted by the amount of rotation OR with respect to the coordinate thread XY.

Therefore, assuming that there is no rotational error in the projected image 1), simply move the wafer stage in the X and Y directions by a constant pitch 'JI!' By repeating the exposure process, for example, projected images Pr'l arranged in the X direction on the wafer W.
Although the center 01 exactly coincides with X1RII, each projection image Pr', Pr'1 is still 1 including a rotation error with respect to the coordinate system XY. Therefore, an orthogonal coordinate system αβ is determined which is inclined by the amount of rotation θR of the projection image Pr' (reticle R) with respect to the coordinate system XY, and the wafer stage is stepped along this coordinate system αβ.

The projected image Pr2' exposed in this way is the projected image Pr'
Rotation errors are removed from the coordinate system αβ, including the coordinate system αβ.

Here, the coordinate system αβ represents a plurality of local areas where the same circuit pattern is transferred on the wafer W in a matrix arrangement, that is, an arrangement of chips, and will therefore be referred to as an arrangement coordinate hereinafter. Now, if the pinch in the X direction of the plurality of chips to be arranged on the UNINO W is SP as shown in Figure 13, then when repeating exposure in the X direction, the UNINO stage is next stepped from the current position by ΔX. It is sufficient to stop at a position moved by ΔY. The movement 1ΔX, ΔY is determined by equations (3) + (4) when stepping in the X direction, and when stepping in the Y direction, the pinch in the Y direction is
It is determined by equation (5) + (6) as follows.

Based on these equations (3) to (6), the CPU 60 calculates in advance the position of each chip in the XY coordinate system so that the wafer stage steps according to the arrangement coordinates αβ on the wafer W.

Step 207 Next, the CPU 60 steps the wafer stage based on the above calculation result to repeatedly print the pattern of the reticle R on the Uni-W. This v
A! For example, the first
The chips are arranged as shown in Figure 4. In FIG. 14, it is assumed that FL of the wafer W and the XYOX axis of the coordinate system are parallel, and only one specific row of chips is shown among a plurality of chips arranged in the X direction. Center 0. of each chip C1-C0. ~O
g are all located on the α axis, and each chip is not rotated in the array coordinate αβ.

Step 208 The first layer of Unino W exposed as above is
It is taken out from the Unino Boulder IO and sent to processes such as development. When exposing the next unit Wl, that is, the second unit in the lot, the CPU 60 performs wafer exchange (W exchange) and then uses the stenoa again.

? The same operation is repeated from 03 onwards.

As described above, the pre-alignment of Unino W is executed again in step 203, and the CPU 60 executes the pre-alignment in step 203.
Execute 4. In step 204, it is determined that the second part of the lot is to be exposed, and the CPU 60 executes step 207.

For the second unit, since the array coordinate αβ corresponding to the rotation amount OR has already been determined, exposure of each chip is repeatedly performed according to the coordinate system αβ as shown in FIG.

As mentioned above, when exposing the first layer pattern, there is no need for the overlapping operation, and unless the reticle R for the first layer is replaced, steps 203, 204, 207, 208
is repeated continuously. The wafer exposed to i8 light in this manner is subjected to a first layer process to form a matrix of chips Cn as shown in FIG. Each of the chips Cn has a pattern region Pr of the first layer reticle R transferred thereto, and the valley chip has marks Sxn, S'/n for laser stamp alignment, and marks Wyn, Wyn for wafer alignment. Wen is formed concomitantly.

However, in order to simplify the drawing, marks Wyn and Wθn correspond to two chips C7 and Cj2 separated to the left and right on the wafer W.
The mark attached to Wy7. Wθ7 and mark WY]2. W
Only e12 is illustrated. Also, in Fig. 15, only two marks 5xnrsY'' are provided for one chip extending in the direction of direct firing, but as is clear from the pattern of the reticle R in Fig. 12, in reality, As shown in Fig. 16, a mark S ynl is also formed at a position facing the mark Syn on the chip. Fig. 16 shows the origin of the array coordinate system αβ near the center of the wafer W on the chip C·, 9. This is an enlarged view, showing marks Sy9 and Mark S.
y9' is formed to coincide with the α axis. Furthermore, Mark W
For the yn, Wθ mouth, either the interval between the marks WY7 and Wθ12 or the interval between the marks we7° and W)'12 is off-axis W-M I 021
, 22 spot lights WYS and WθS are formed so as to be equal to the interval between them.

When exposing the first layer, stepping the wafer stage by one step with respect to the coordinate system hand,
It is also extremely effective for exposure equipment for making working masks used in gloxomity type or contact type batch exposure equipment.

Next, the operation of overlapping and exposing the second and subsequent layer patterns of the wafer WK onto which the first layer pattern has been transferred will be explained using the flowchart shown in FIG. 10 previously shown. In the first layer exposure, reticle/l/ a-cheese-
By correcting the stepping position of the wafer stage according to 6Q-, the rotation of each chip with respect to the array coordinate thread αβ on the wafer (hereinafter referred to as chip rotation CR)
is removed and transferred, but the amount of rotation does not necessarily become zero. Possible causes include a decrease in alignment accuracy and stepping accuracy due to device drift, or a positional shift that occurs when the reticle stage 2 is vertically attracted to the base after completion of reticle alignment. Further, when the 11th tl is transferred using another exposure device, the chips are not necessarily arranged so as to correct the reticle rotation, that is, remove the chin protrusion. As shown in FIG. 17, for example, the center of each chip is accurately positioned along the array coordinate system αβ, but each chip itself is tilted by the amount of rotation θC.

Therefore, in the overlapping exposure of the second and subsequent layers, it is necessary to measure and correct the two-chip rotation other than the reticle rotation.

Now, at t4iJ when exposing the second layer, attach the reticle R (N (reticle R shown in Figure 8) for the second layer to the exposure device, and perform steps 200 to 205 in Figure 10 as described above. The process is carried out in exactly the same manner as the operation.Here, the second layer pattern of the first wafer of the rod is exposed, so the next step 213 is executed.

Step 213 Here, the CPU 60 performs calibration (so-called parallelization) of the W-MICs 21 and 22. 1. The CPU 60 cooperates with the WA C 65 to check the mark FM on the fiducial mark plate 30.
The wafer stage is positioned so that y matches the spot light WYS of the W-MIC 21. And CPU60
memorizes the Y coordinate of the wafer stage at that time, and then moves the wafer stage to the
Move the distance in the direction.

Accordingly, the mark FMV on the reference mark plate 30 is W−M.
It is located within the observation field of IC22. Next, the CPU 60 moves the burping glass 22b as shown in FIG.
Adjust via 5.

By the above operation, the line segment connecting the spot lights WYS and WO3 becomes exactly parallel to the X axis of the coordinate system XY.

Now, when the parallel alignment of W-M I C21 and 22 is completed,
The CPU 60 is connected to the mark W on the wafer W shown in FIG.
y7 is located within the field of view of W-MIC21, and mark Wθ1
The wafer stage is positioned so that 2° is the internal pressure of the field of view of the w-MIC 22. Then, move the wafer stage to X so that the spot light WYS and mark Wy7 match.
After positioning in the direction, the CPU 60 outputs the spot light WYS
While the mark Wy7 remains aligned, the wafer holder 10 is rotated so that the spot light WθS and the mark Wθ12 are aligned. As a result, the rotation error of the wafer with respect to the coordinate system XY is removed, and the array coordinate system αβ is also set without rotational deviation with respect to the coordinate system XY.

Then, when the spot light wys and the mark Wy7 match, the CPU 60 stores the value of the Y counter of the laser interferometer 14 as the Y coordinate value Ygo. Next, the CPU
60 moves the wafer stage so that the spot light SPx (see FIG. 4) from the LSA system 18 aligns with the mark Sx9 of the chip C9 (see FIG. 15) near the center of the Unino W. The position of mark Sx9 is detected by LSAC62. Specifically, after positioning the mark Sx9 and the spot light SPx in parallel, the wafer stage is moved by a predetermined amount in the X direction so that the mark Sx9 and the spot light SPx scan relative to each other. Extract the intensity distribution of the diffracted light generated from 3x9. Then, for example, the peak position of the intensity distribution, the position dividing the intensity distribution into two in the X direction, or the center of gravity of the intensity distribution is determined.

By returning the wafer stage to that position, the spot light S P x and the mark Sx9 are precisely aligned.

When the mark Sx9 and the spot light SPx match in this manner, the CPU 60 stores the X counter of the laser interferometer 12 as the X coordinate value Xgo. The above operations complete the association between the array coordinate system αβ of the wafer W and the coordinate system XY. However, in this case, the α axis of the coordinate system αβ is the 15th
υ different from the figure, mark Wy7. Wθ7, WYI2, W
7”, C7,, but W-MIC21 through O12
, 22, the Y coordinate value Ygo read from the Y counter is marked Syn (Syn').
The α axis is defined in the same way as in FIG. 15 by reducing the distance between the mark Wyn (Wθn) and the β direction.

Therefore, if the wafer stage is positioned so that the counted values of the X and Y counters are equal to the stored XY coordinate values Xgo and ygo, respectively, the projection point of the optical axis AX of the projection lens 5, that is, the center RC of the reticle H. and center 0 of chip C9
9 (The origin of the coordinate system αβ can be made coincident with the origin of the coordinate system αβ.

With this, wafer global alignment is completed.

Step 210 Here, the CPU 60 defines the tengross region (CR) described above. The CR measurement uses spot light SPy from the LSA system 19 to determine the rotation amount θC by detecting the positional deviation in the Y direction between marks Syn and sy n/ formed on both sides of each chip on the wafer. First, C
P U 60 moves the wafer stage so that the spot light SPy scans the mark Sy9 of the chip C9 near the center of the wafer in the Y direction, and the spot light SPy and the mark S
The Y coordinate value Y3 when y9 matches is read from the Y counter and stored. Next, the spot light SPy is marked s
Move the wafer stage in the X direction so that it is aligned parallel to y91. The amount of movement is the distance between the marks Sy9 and Sy9', and is assumed to be Px as it can be accurately read from the X counter of the laser interferometer 12. Next CPU60
Similarly, the mark Sy9' is scanned in the Y direction with the spot light SPy, and the Y coordinate value Y4 when the spot light SPy and the mark S V 9' match is read from the Y counter and stored. Based on the above measured values, the CPU 60 calculates the rotation amount θC of the tip rotation using equation (7). However, since θC is extremely small, this is an approximation.

f) c-45in “egg) = Y3 Y4 ,,,,
,, (7) Px Above, the measurement of chip rotation is fit )X
In terms of improvement, the same process is performed for multiple chips located near the center of the wafer W, and the average rotation h1θ of each chip is calculated.
It is better to average C.

With the above steps, the CPU 60 completes the CR measurement and executes the next step 211.

Step 211 Here, the CPU 60 rotates the wafer W by the rotation amount θR of the reticle R for the second layer and the rotation amount θR of the chip rotation.
The global alignment of the wafer is performed again so that the wafer is positioned at an angle with respect to the coordinate system XY by the difference Δθ in C. There are two steps to this realignment. One is the mark wy on the wafer W.

One method uses Wθ, and the other method uses mark FMy on the fiducial mark plate 30. Mark W on wafer W
When using y and Wθ, the CPU 60 performs the previous step 2.
Similar to 03, W-MIC210 works in conjunction with WAC65.
Spot light wys matches mark WY7, W-MIC
The two-dimensional moving stage is positioned so that the spot light WθS of No. 22 coincides with the mark Wθ12.

Next, the cPU 60 moves the wafer W into the coordinate system
Mark Wy7 created when tilted by Δθ with respect to Y
The deviation amount ΔYC of the mark Wθ12 in the Y direction is calculated using equation (8).
Calculate by However, since the rotation amount θC is extremely small, this is an approximation.

ΔYC: L −5in(θC−θR)=L・(θC−θ
R>-<8) In addition, the rotation direction of the reticle rotation is positive in the coordinate system rule, and counterclockwise is the positive direction.
The direction of rotation of John is positive in the array coordinate system αβ. Then, CPU6o moves the wafer stage in the Y direction by this deviation amount ΔYC.

Whether the movement is positive or negative in the Y direction is determined depending on whether the calculation result of (θC-0R) is positive or negative. Here, when (θC-0R) is positive, the wafer stage is moved in the negative Y direction by the deviation amount ΔYC. Next, the CPU 60 controls the WAC 65 to
By rotating the burping glass 22b (see FIG. 6) in 2, the spot light WθS is displaced in the direction in which the two-dimensional moving stage is moved by the deviation amount ΔYC. And WAC6
5, the rotation of the burping glass 22b is stopped when the coincidence between the spot light WθS and the mark Wθ12 is detected.

On the other hand, when using the mark FMy for wafer realignment, the wafer stage is first positioned so that the mark FMy and the spot light WYS coincide. Then, the wafer stage is moved by a distance in parallel to the X direction, and stopped at a position where it has been moved by the deviation amount ΔYC determined by the above equation (8) in the Y direction. After that, the spot light Wθ
The burping glass 22b may be rotated so that S and mark FMY coincide.

With the above operations, the line segment connecting the spot light wy s and w θS is the coordinate system X Y (1) Rotation Mi with respect to the X axis
It is adjusted to be tilted by Δθ. Next, the CPU 60 moves the mark WY7 to the spotlight W as in step 203 above.
The two-dimensional movement stage is slightly moved in the Y direction and the wafer holder 10 is slightly rotated so that the mark Wθ beam 2 coincides with the spot light WθS. This is K
Therefore, the arrangement coordinate system αβ of the wafer W is aligned with an inclination of the rotation amount Δθ with respect to the coordinate system XY. Moreover, the relative rotation error between the projected image Pr' of the reticle R and the chips to be superimposed on the wafer W is removed.

The above situation is illustrated in FIGS. 18 and 19. Figure 18 shows the arrangement coordinate system αβ of the wafer W and the coordinate system XY.
The case where the projected image Pr' of the second layer pattern is superimposed on the chip C9 is shown.

The projected image Pr' due to the rotation of the reticle R is rotated by θR in the negative direction with respect to the coordinate system XYK, and each chip 7"Cn on the wafer W is It is rotated by θC in the direction.

Therefore, as explained earlier, by realigning the wafer W according to the rotation amount θR1θC, the array coordinate system αβ is tilted by Δθ with respect to the coordinate system XYK, as shown in FIG. are matched only by relative rotation. Therefore, when performing exposure using the step-and-repeat method, the wafer stage is not stepped according to the coordinate system XY, but rather the coordinate system x'y' (in Fig. 19, the array coordinate system αβ
is consistent with By stepping according to the pattern, overlapping exposure can be performed on each chip without any rotation error. With the above steps, the wafer realignment in step 211 is completed.

Step 212 Next, the CPU 60 calculates the amount of correction in stepping of the wafer stage in the same manner as in equations (3) to (6) above.

In other words, when sequentially exposing chips arranged in a row along the α(X') axis on the wafer W, the stepping distances ΔX and ΔY to the next adjacent chip to be exposed can be calculated using equations (9) and (10). , when sequentially passing through chips lined up in a row along the β (Y') axis, the stepping distance ΔX to the adjacent chip in the β direction is determined by the equation Qll, (J'J).
, calculate ΔY.

■ΔX=SP-cos(ΔθJ=SP...
-(9)1ΔY=S P -5in(Δθn=sP・Δ
θ ・・・・・・・・・GQ However, SP is the arrangement pitch of the chip in the α direction in the arrangement coordinate system αβ, and SP
' is the arrangement pitch in the β direction. Also, since the rotation amount Δθ is extremely small, it is used as an approximate expression.

Step 207 Next, the CPU 60 advances the wafer stage according to the stepping amount (ΔX, ΔY) calculated in step 212, and transfers the projected image Pr' of the second layer reticle R to each chip on the wafer W. The images are overlapped and exposed sequentially. As described above, each chip on the wafer W is substantially free of reticle rotation and chin protrusion, and more precise overlay is achieved.

Step 208 The wafer W thus exposed is placed in the wafer holder l of the apparatus.
When the next wafer, that is, the second wafer of the lot, is to be exposed, the cPU 60 performs step 2.
Repeat the same operation from 03.

However, as shown by the dotted line from step 204, the wafer alignment in step 213 is performed using the W-M I adjusted in step 211 after the wafer pre-alignment.
Spot light wys of C21, 22, mark Wy on wθS
Immediately, global alignment is performed to match mark 3xn with LSA system 18 spot light SPx by matching n, W&n. This is because there is little variation among the wafers within the four lots, and the amount of chip rotation does not change.

Then, in the next step 204, CPU U3O
It is determined that the wafer is the second wafer, and immediately steps 207
The print (exposure) operation starts. In this way, for the second and subsequent wafers of the same lot, step 203.
Exposure processing is performed in the order of 204, 213, 207, and 208. Further, when the lot is finished and the reticle R is to be replaced when exposing the first lot, the CPU 60 carries out the reticle R based on the judgment in step 209 and repeats the operations from step 200 over and over again.

Now, in the explanation of the above operation, in step 207, the wafer stage is simply moved forward, and the marks on the wafer are not used for alignment. However, when expansion and contraction (runout) of the unino occurs, the arrangement pinch of the chips on the unino becomes different from the designed one, resulting in a decrease in alignment accuracy. Therefore, when stepping the Unino stage to match the projected image Prr with the chip Cn, the spot lights SPx and SPy from the laser step alignment system are used to travel around the marks Sxn and 3yn around the chip, respectively, and project them. Statue P
After determining the amount of deviation of the chip with respect to the projected position of r', and slightly moving the wafer stage so that the amount of deviation becomes zero,
It is a good idea to mist the chip.

This alignment for one chip is performed in row 1 every time the wafer stage steps. (Block alignment is also possible. If a linear mark that overlaps mark R8 of reticle R is provided for each chip on the fC wafer, the matching state of this linear mark and mark R8 can be determined by the DDA system 20 after stepping by coordinate correction of Δθ. The die is detected by the DDAC64.
A small amount of positional deviation can be corrected by performing pie-die type step alignment or block alignment. In optical equipment that mainly uses die-pie-die alignment, the reticle rotation and chin approach are corrected, for example, after the reticle R and wafer W are arranged as shown in FIG. ,
During exposure, the wafer stage may simply be stepped in parallel to the X and Y axes of the coordinate system XY. In FIG. 19, the wafer W is stepped in the negative direction of the X axis by the pitch SP, and the projected image Pr' of the pattern area Pr and the chip 01
When superimposing the projections Pr' and the chip CIO, the relative rotational deviation between the projection Pr' and the chip CIO is removed, and only the positional deviations in the X and Y directions remain. - Easily removed by slightly moving the wafer stage using die-based alignment.

The first embodiment of the present invention has been described above, and by performing the R-MIC check in step 200 in FIG. 10, the reticle alignment in step 201, and the RR measurement measurement in step 202 multiple times, the reticle rotation amount It is also possible to drive it to zero. To this end, a burping glass as shown in FIG. 6 is provided in the optical system of the R-MIC 17 shown in FIG. 1, so that the index image 1.7 a of the R-MIC 17 shown in FIG. It is configured so that it is displaced by a minute amount. And step 200
, 201 and 202 to determine the rotation amount θR of the reticle rotation, rotate the burping glass of the R-MIC 17, and move the index image 17a by a minute amount in the Y direction by an amount corresponding to the rotation amount θR. shift. Specifically, this operation moves the reference mark plate 300 and the reference mark 30Y from R to M.
After moving the IC 16 from the projection position of the index image L6a to the projection position of the index image 17a, the reference mark 30Y is further moved to Y.
(The index image 17a is moved slightly in the direction according to the rotation amount θR so that the index image 17a matches the reference mark 30Y at that time).
-Achieved by shaking a ping glass. Then, reticle alignment is performed again based on the adjusted index image 17a of the R-MIC 17 and the index image 16a of the R-MIC 16. After this alignment, the reticle rotation is measured again, and if the rotation amount θR falls within the predetermined accuracy depending on the accuracy of the detection system and the accuracy of the mechanical system, the next step 203 is executed. Readjust the noping glass of R-MIC17 again. By repeating the above steps with a method that provides a certain level of accuracy, the amount of rotation θR of reticle rotation can be reduced to an extremely small value. This results in a coincidence, which lengthens the overall reticle alignment time.

Therefore, it is faster in terms of time to perform the RR measurement only once as in this embodiment, and then correct the rotational deviation by correcting the rotation of the wafer, rotating the stepping coordinates, or the like. Of course, the reticle alignment method described above is an effective method when the reticle is rarely replaced.

Next, a second embodiment of the present invention will be described. In the previous embodiment, the ching locusion was performed using the sub-bont light SPy based on the laser step alignment to mark Syn,
This was detected by measuring the displacement of the Sy port' in the Y direction. In the second embodiment, a DDA system and a DDAC 64 are used to measure the chip rotation by using the mark R8 of the reticle R as a reference window and detecting the marks on both sides of the chip by moving the wafer stage. The measurement procedure will be explained using Figure 1. First, W-M I C21, 2
2, the wafer W is globally aligned and the coordinate system XY and the array coordinate system αβ are matched. Next, relying on the f blueness of the global alignment and the stepping precision of the Unino stage, the chip C near the center of the Unino is positioned directly below the projection lens 1. Next, when mark R8 of reticle R is observed by DDA system 20, chip C
On the right side, the friendship mark DMI extending in the X direction is marked R.
Detected together with 8. Then, the edge Rsl of the mark Rs and the Y-direction positional deviation ΔYMI of the mark 1) Ml with respect to Rs2 are detected by the DDC64. Next, the wafer stage is moved in the X direction without changing the Y coordinate value, and the linear mark DM2, which is formed on the left side of the chimp C and extending in the X direction, is moved to the DDA via mark R8.
The DDAC 64 detects a positional deviation ΔYM2 of the mark DM2 in the Y direction with respect to the mark R8. Since the distance LD between the marks DMI and DM2 is predetermined in design, the amount of rotation θC of the chin approach can be approximately determined by 2a→.

Note that when the DDA system 2) is configured to irradiate mark R8K on the reticle R with illumination light of the same wavelength as the wafer light, it is necessary to block the illumination light while the wafer stage moves in the X direction. There is.

Next, a third embodiment of the present invention will be described based on FIGS. 21 and 22. In each Example 1, the measurement of the rotation -3ll:RC of the reticle rotation and the measurement of the rotation amount θC of the chingrotation are performed separately.
The rotation amount Δθ of the deviation was determined by calculation. In the third embodiment, the rotation amount Δθ is directly detected without performing the calculation. FIG. 21 is a perspective view partially showing the vicinity of the reticle stage 2 of the reduction projection type optical system shown in FIG. 1, and the same members as in FIG. 1 are given the same reference numerals. In this embodiment, the DDA system or mirror 2 (
A die-pie-die alignment system, ie, a DDA system, having the same configuration as la is provided with a mirror 20a' at a position rotationally symmetrical about the optical axis AX of the projection lens 1.

Correspondingly, a similar window-like mark Rsl is formed on the reticle R on the opposite side of the mark R8 across the pattern region Pr. The DDA system W observes the image of the mark Rs' reflected by the mirror a', and at the same time, the projection lens 1
The marks on the wafer are also observed through the wafer.

Now, if the two DDA system adders 20' shown in FIG. 21 are used, step 202 (RR measurement) in FIG. In the measurement, the amount of rotation Δθ is directly measured.The procedure for measuring the amount of rotation Δθ will be explained using FIG. After that, Unino alignment is performed to align marks Wy7 and Wθ12 on the wafer with the sub-light beams wys and wθS, respectively.This eliminates the rotation error between the alignment coordinate system αβ of the Chimp C on the Unino and the coordinate system XY of the Unino stage. FIG. 22 shows a case where the array coordinate system αβ and the coordinate system XY match. In the coordinate system Assume that they are rotated in the negative direction by the amount of rotation θR and overlap. Then, the mark DMI on the right side of chimp C is the center of edges R51 and Rs2 of mark R8 on reticle R, that is, line 1.
Changing the X coordinate value of the wafer stage so that it is aligned on r, ((positioning in the Y direction).

At this time, ci'U60 is the Y coordinate value Y of the wafer stage.

Load. Next, the wafer stage is positioned in the Y direction without changing the X coordinate value so that the mark DM2 on the left side of the tick C coincides with the center (line 1r) of the mark RS/, and the CPU 60 determines the Y coordinate value Y at that time.

Load. Assuming that the distance between marks Rs and Rs' on Unicorn 1 is equal to the distance between marks DMI and DM2, and the distance is LD, then both rotation amounts θC1θR are extremely small, so the CPU 60 uses approximation formula 1. Δ based on IQ
Calculate θ.

Ebbie center in Δθ... αOD After determining the rotation amount Δθ, the wafer is tilted by Δθ with respect to the coordinate system XY and realigned, as in the first embodiment.
Step-and-repeat exposure is performed along a coordinate system αβ tilted by Δθ.

In addition, the amount of deviation between the mark R5 (R5') and the mark DMI (DM2) is detected in the D D A C64 (if there is a geometrical function, the center (line 11r) of the mark R5 (R8') and the mark DMI ( DM2) deviation amount ΔY in the Y direction
, (ΔYt) with the line 1r as zero and considering the positive and negative polarities, it is possible to calculate the following in the same manner as the equation aQ. Like this D D A C6
If the processing circuit 4 has a function for detecting the amount of deviation, the wafer stage does not have to be moved slightly for the purpose of measuring the tengo groove, and faster measurement is possible.

Furthermore, if two marks Rs and Rs' are provided on both sides of the pattern area Pr as in this embodiment, reticle rotation can be measured using the reference mark plate 30 even after reticle alignment. In this case, first mark Rs on reticle R and mark 30Y on reference mark plate 30 are
After matching 1 place with and, mark RS / and mark 30Y
If the wafer stage is moved in the X direction without changing its Y coordinate value so that it can be observed by the DDA system 20', and the positional deviation in the Y direction between the mark RS/ and the mark 30Y is found, This means that the rotation amount θR has been calculated.

Three embodiments of the present invention have been described above, and in particular, as shown in the third embodiment, the difference Δθ between the tip rotation and the reticle rope rotation can be directly expressed without using the coordinate system XY. The measuring method can also be applied to exposure equipment that does not have a high-precision, high-resolution position detection device such as the laser interference side.For example, in an exposure equipment that measures the position of the wafer stage using an encoder with low position detection accuracy, After the global alignment of the wafer is completed, it is difficult to accurately send the unihasdage using a step-and-repeat method to follow the array coordinates on the wafer. Therefore, as in the example @3, the difference Δθ between the reticle rotation and the chin approach is directly detected.
If global alignment is performed by tilting the wafer by Δθ with respect to the coordinate system XY of the Unino stage, the relative rotational deviation between the chips on the wafer and the reticle can be eliminated.

Therefore, if alignment with the reticle is performed for each tick using the die-pie-die method, an extremely high ntJX overlay can be achieved and there is an advantage that the alignment time is not unnecessarily lengthened. In this way, in exposure equipment that does not have a position detection device using a laser interferometer, the Unino stage simply moves coarsely in the X and Y directions.
The reticle stage 2 is configured to move slightly in the X and Y directions, and a similar effect can be obtained by performing rotational correction by IΔ·0 and then performing die alignment.

In addition, in each of the first to third embodiments, the marks 3yn, Syn', and the mark D are used for chip rotation measurement.
Although MI and DM2 are provided to coincide with the lines passing through the center 0 of the chip, the lines connecting the marks on both sides of the chip do not necessarily need to pass through the center 0. Further, the pair of marks may be arranged diagonally on the chip. Since the diagonal of the tip is longer than the distance between the left and right sides of the tip, this has the advantage of improving the total chin progression by 3 kg. Furthermore, when correcting Unino rotation, the residual rotation error of Uni/% is measured as disclosed in JP-A-56-102823, and the reticle rotation is corrected at the time of Unino realignment (step 211). It is advisable to correct wafer rotation by including rotation rotation as well as rotation rotation.

As described above, the present invention is not applicable only to reduction projection type optical systems; for example, even if the present invention is an Exactly the same effect can be obtained.

(Effects of the Invention) As explained above, according to the present invention, when a single substrate to be exposed (UNINO) is exposed by repeatedly overlapping turns of a mask (reticle), reticle rotation and chin-progression are performed. Because both are corrected, regardless of differences in exposure methods or manufacturing errors between different exposure devices even when using the same exposure method, the pattern (turn) transferred to the exposed substrate will always have extremely high overlay accuracy. As a result, the yield rate of manufactured semiconductor devices is increased and the yield is significantly improved.

Furthermore, without pursuing positioning accuracy regarding rotational deviation of the reticle or wafer, alignment accuracy can be achieved simply by increasing the positioning accuracy with respect to the orthogonal coordinates (coordinate system Another advantage is that an exposure apparatus with high reproducibility accuracy can be easily obtained.

[Brief explanation of drawings]

FIG. 1 is a perspective view showing a schematic configuration of a reduction projection type exposure apparatus according to an embodiment of the present invention, FIG. 2 is an optical layout diagram of a laser stamp alignment system, and FIG. 4 is a diagram showing the arrangement of the image field if of the projection lens and the spot lights SFX and sPy. FIG. 5 is a diagram showing the arrangement of the image field if and the off-axis spot lights wys and wθS. 6 is a layout diagram of the no-ping glass that shifts the spot light WθS, FIG. 7 is a plan view of the reference mark plate 30, and FIG. 8 is a reticle for transferring the second layer pattern. 9 is a circuit block diagram of the control system in the exposure apparatus of this embodiment, FIG. 10 is a flowchart for explaining the operation of this embodiment, and FIG. 11 is a diagram of the image field IF and reticle alignment microscope. FIG. 12 is a plan view of the reticle for transferring the first layer pattern, FIG. 13 is a diagram explaining the operation of correcting reticle rotation, and FIG. 14 is a plan view of the reticle for transferring the first layer pattern. FIG. 15 is a plan view showing the arrangement of each chip on the wafer and alignment marks, FIG. 16 is an enlarged view of one chip on the wafer, and FIG. 17 is a plan view of the wafer with the pattern transferred thereto. Diagram explaining chip rotation, 1st
Figure 8 is a diagram explaining the relationship between the chin approach and reticle rotation, Figure 19 is a diagram showing how the chin approach and reticle rotation are corrected, and Figure 20 is a diagram showing a second embodiment of the present invention. FIG. 21 is a perspective view partially showing the configuration near the reticle of an exposure apparatus according to a third embodiment of the present invention, and FIG. 22 is a diagram explaining the chip rotation measurement operation according to the third embodiment of the present invention. It is a figure explaining measurement operation of rotation. 5. Explanation of symbols of main parts l... Projection lens, 2... Reticle stage, 6.
...Y spacer';, 8...X stage, 9...2 stage, 10...wafer holder, 16, 17.
...Reticle alignment microscope, 18, 19...
Laser step alignment system, 20.21j... Die-pie-die alignment system, 21, 22... Wafer alignment microscope, 60... Processor (C
PU), R...reticle, W...wafer. Applicant Nippon Kogaku Kogyo Co., Ltd. Agent Takao Escort [Figure 5, Figure 8 (a) Figure 6, Figure 7, Figure 8 (C) Hitoshi, Figure 9, Figure 10, Figure 11, Figure 12, Figures 1 and 3 Figure 14 Figure 15 Figure 1b Figure 17 Figure 18 Figure 1'? figure

Claims (2)

[Claims] (1) a mask holding means for holding a mask having a first pattern; a second mask that can be aligned with the first pattern;
a substrate moving means for holding a substrate to be exposed on which a plurality of patterns are formed along predetermined array coordinates and moving the substrate to be exposed two-dimensionally; a relative relationship between the first pattern and the second pattern; rotational error detection means for detecting a rotational error; rotation means for rotating either the mask or the substrate to be exposed so that the rotational error is corrected; and the rotationally corrected first pattern and the plurality of rotational errors. an alignment device for an exposure apparatus, further comprising a control means for controlling the substrate moving means so as to sequentially align each of the second patterns. (2) a mask holding means for holding a mask having a first pattern and moving the mask along predetermined orthogonal coordinates; a second pattern that can be aligned with the first pattern is arranged at predetermined array coordinates; a substrate moving means for holding a plurality of exposed substrates formed at regular intervals along the Jar, and moving the exposed substrates two-dimensionally along the orthogonal coordinates; the first pattern and the second pattern; rotation error detection means for detecting a relative rotation error of the patterns; rotation means for rotating either the mask or the substrate to be exposed so that the rotation error is corrected; When positioning the substrate moving means accordingly,
The first pattern and the second pattern generated due to the rotation correction
A positional deviation detection means for detecting a positional deviation of a pattern along the orthogonal coordinates; and a control means for moving either the substrate moving means or the mask holding means so as to correct the positional deviation. Characteristic alignment device for exposure equipment.