JPH0774094A - Projection aligner and therefor - Google Patents

Abstract

PURPOSE:To detect the deviation in rotation between a mask and an exposed body with high accuracy in a short time by detecting the quantity of error in the rotational direction of a mask to orthogonal coordinates with a first rotation error detection means, and detecting the quantity of error in the rotational direction of the arrangement coordinate system in the exposed area to the orthogonal coordinates with a second rotation error detection means. CONSTITUTION:A minute opening member 8 and a photoelectric detector 9 detect the projected images of the translucent marks RR and RL of a reticle 5, and a laser interference meter 13 seeks the position coordinates, and detects the rotational deviation as an angle epsilon, The arrangement coordinate system alphabetaof a wafer 10 is positioned being slant by the angle epsilon' to the XY-coordinates of a stage 7. Accordingly, the projection pattern area P0 of the reticle 5 of the second layer and each pattern area E1-E6 of the first layer always existing on the wafer 10 are set under the condition that they are slant counterclockwise by the angle approx.=', respectively, to the XY-coordinate. This way, the rotational deviation to the wafer shift coordinate system of the mask can be sought with accuracy in a short time.

Description

Detailed Description of the Invention

[0001]

The present invention relates to a semiconductor (wafer) for transferring a transfer mask pattern of a high density integrated circuit through a projection optical system.
More specifically, the present invention relates to an apparatus for exposing on the upper surface, and more particularly to an apparatus for performing relative alignment between a mask and a wafer without rotational displacement with respect to an optical image of a mask pattern.

[0002]

2. Description of the Related Art Large scale integrated circuits (LSI: Large Scale)
The miniaturization of integrated circuit patterns is advancing year by year, but a reduction projection type exposure apparatus has become widespread as a circuit pattern printing apparatus satisfying the requirements for miniaturization and having high productivity. In these devices which have been conventionally used, a reticle pattern that is several times (for example, 5 times) the pattern to be printed on a silicon wafer is reduced and projected by a projection lens and printed in one exposure. The area is smaller than a square having a diagonal length of 21 mm on the wafer. Therefore, in order to print a pattern on the entire surface of a wafer having a diameter of about 125 mm, a so-called step-and-repeat method is adopted in which the wafer is placed on a stage, moved for a certain distance, and exposure is repeated.

In the manufacture of an LSI, patterns of several layers or more are sequentially formed on a wafer. However, unless the overlay error (positional deviation) of different interlayer patterns is kept below a certain value, the conductivity between layers or The insulation state becomes unintended and the function of the LSI cannot be achieved. For example, for a circuit having a minimum line width of 1 μm, only a positional deviation of about 0.2 μm is allowed.

The positional deviation during printing is roughly classified into rotational deviation and parallel deviation. FIG. 12 is an explanatory view exaggeratingly showing the state where the rotational deviation has occurred. The solid line rectangular pattern area P ₁ is a circuit pattern area already formed on the wafer, and is minute with respect to this area P ₁ . Rectangular pattern area P indicated by a broken line in a state of being rotated by an amount
₂ are overlaid and exposed and baked. In this case, the area P
Assuming that the center O of the diagonal line of ₁ is completely coincident with the center of the diagonal line of P ₂ , the region P ₁ is present at a position other than the center O.
On the other hand, rotational deviation (rotational direction holding error) occurs in the region P ₂ .

In the conventional reduction projection type exposure apparatus, when performing the step-and-repeat operation, the stage on which the wafer is placed is moved via the wafer holder with reference to the origin of the orthogonal coordinates forming the moving plane. On the other hand, the reticle on which the circuit pattern is drawn as a mask is positioned and held so that the projected image of the reticle on the wafer has as little rotational displacement as possible with respect to the Cartesian coordinates. In such a state, a series of step-and-repeat type exposure printing is performed in which the exposure is performed by positioning in the orthogonal coordinate system.

When the detection center of a position detector (for example, a reticle alignment microscope) used for positioning the reticle is deviated, the reticle positioned with reference to the position detector is positioned with respect to the orthogonal coordinate system of the stage. Due to the error, the pattern area printed on the wafer will generally have a rotational deviation.

Conventionally, the correction of the rotational deviation is performed by actually repeating a reticle pattern and printing it on a wafer, observing the pattern on the printed wafer with an optical microscope or the like, and detecting the deviation between the patterns printed adjacent to each other. The position is measured, and the deviation of the detection center of the position detector is determined from the measured value in reverse to correct it.

[0008]

However, in the conventional method as described above, since a pattern is printed on a wafer, the pattern is developed, and the amount of rotational deviation is observed and determined, it takes a lot of time and labor. Moreover, the accuracy of the correction is not high, and it has been extremely difficult to suppress the positional deviation of the circuit pattern due to the rotational deviation to the 0.1 μm position.

The present invention solves the above-mentioned problems and detects the rotational deviation between the mask and the object to be exposed with high accuracy in a short time,
An object of the present invention is to obtain a projection exposure apparatus and an exposure method capable of performing an exposure operation for each mask pattern with high overlay accuracy and high efficiency.

[0010]

To achieve the above object, in a projection exposure apparatus according to a first aspect of the present invention, holding means for holding a mask having a predetermined pattern formed therein and a plurality of exposed regions. A projection optical system that projects the pattern onto a photosensitive substrate formed according to a predetermined array coordinate system, a substrate stage that holds the photosensitive substrate and is two-dimensionally movable within a predetermined orthogonal coordinate system, and the orthogonal coordinate system. Coordinate measuring means for measuring the coordinate position of the photosensitive substrate by detecting the amount of movement of the substrate stage inside, and control means for controlling the movement of the substrate stage based on the measurement value of the coordinate measuring means. In the apparatus, which sequentially positions the substrate stage according to predetermined target position information, and exposes the pattern of the mask by superimposing it on each of the plurality of exposed regions. A photoelectric detection unit for photoelectrically detecting projected images of a plurality of marks on the mask projected by the projection optical system via a reference mark provided on the substrate stage; Illuminating means for illuminating a pattern of the mask and illuminating a plurality of marks on the mask to form a projected image of the marks by the projection optical system on the reference mark, and the holding means. With the mask attached, based on the signal output from the photoelectric detection means,
First rotation error detection means for detecting an error amount in the rotation direction of the mask with respect to the orthogonal coordinate system, and a plurality of exposed regions on the photosensitive substrate in a state where the photosensitive substrate is mounted on the substrate stage. By detecting coordinate positions of at least two preselected marks among the marks associated with each other in the Cartesian coordinate system, array coordinates of a plurality of exposed regions on the photosensitive substrate with respect to the Cartesian coordinate system. Second rotation error detection means for detecting an error amount in the rotational direction of the system, and the photosensitive substrate is rotated based on the two detected error amounts, and the mask and a plurality of exposed regions on the photosensitive substrate. Correction means for making the amount of error in the direction of rotation relative to the array coordinate system of the mask substantially zero, and the target position information is corrected and calculated based on the amount of error in the direction of rotation of the mask. And a calculation means for calculating the coordinate position of each of the plurality of exposed areas on the plate to be positioned, wherein the control means has the coordinate position measured by the coordinate measurement means calculated by the calculation means. The positioning of the substrate stage is controlled so as to substantially coincide with.

Further, in the projection exposure method according to the second aspect of the present invention, a plurality of exposed regions are arranged in a predetermined array coordinate system αβ.
A substrate stage that holds the photosensitive substrate formed in accordance with the above-mentioned method and that moves two-dimensionally within a predetermined Cartesian coordinate system XY;
An apparatus having a projection optical system for projecting a pattern formed on a mask on a photosensitive substrate according to a predetermined orthogonal internal coordinate system xy, wherein the substrate stage is sequentially positioned according to predetermined target position information. A method of performing overlay exposure of the pattern onto each of a plurality of exposed regions on the photosensitive substrate via the projection optical system, wherein a plurality of masks on the mask are provided by illuminating the pattern for the overlay exposure. Of the plurality of marks on the photosensitive substrate while photoelectrically detecting projection images of the plurality of marks projected by the projection optical system via a reference mark provided on the substrate stage. Of the marks associated with each of the exposed areas, the coordinate positions of at least two preselected marks in the Cartesian coordinate system are detected, A first step of detecting a relative rotational direction error amount between the orthogonal internal coordinate system xy of the mask and the array coordinate system αβ of the plurality of exposed regions on the photosensitive substrate, and the detected rotational direction error amount So as to be substantially zero, the second step of relatively rotating the mask and the photosensitive substrate, and the rotation direction of the orthogonal internal coordinate system xy of the mask with respect to the orthogonal coordinate system XY after the relative rotation. By correcting the target position information based on the error amount of
A third step of calculating a coordinate position to be positioned for each of the plurality of exposed regions on the photosensitive substrate, and positioning the substrate stage based on the calculated coordinate position in the overlay exposure. To control.

[0012]

According to the present invention, the rotational deviation of the mask pattern with respect to the moving coordinate axis of the stage is suppressed to a sufficiently small value, and even if this rotational deviation exists, it does not directly affect the printing pattern. The position of the step movement of the stage can be controlled.

[0013]

EXAMPLES The present invention will be described in detail below with reference to examples. FIG. 1 is a block diagram showing the outline of a reduction projection exposure apparatus according to an embodiment of the present invention. In FIG. 1, the illumination light from the exposure illumination light source 1 is once converged by the first condenser lens 2 and then reaches the second condenser lens 3. A shutter 4a for blocking passage of illumination light when desired is provided at a position where the light is converged in the optical path. The light flux that has passed through the second condenser lens 3 is a test reticle as a mask (hereinafter simply referred to as a reticle).
Illuminate 5.

The light flux transmitted through the reticle 5 enters a projection lens 6 as an image forming optical system. This projection lens 6
Is an optical system in which the reticle 5 side, that is, the object side, is non-telecentric and the image side is telecentric. The stage 7 directly below the projection lens 6 is normally mounted with the semiconductor wafer 10 thereon, and is moved in the XY directions of the first orthogonal coordinate system which forms the plane of its movement.
The wafer 10 is dimensionally moved, and the wafer 10 is placed on a wafer holder 11 that is two-dimensionally moved integrally with the stage 7.

The wafer holder 11 is provided so as to be capable of minute rotation and vertical movement with respect to the stage 7. The wafer holder 11 moves up and down so that a projected image of a circuit pattern (not shown) of the reticle 5 by the projection lens 6 is imaged on the surface of the wafer 10, that is, focusing is possible.

On the lower surface of the reticle 5, in addition to the circuit pattern, light-transmitting marks RR and RL are drawn at predetermined positions (local portions) at the left and right ends. Focusing on the light flux l ₁ transmitted through the mark RR of the reticle 5, the light flux l ₁
Is condensed by the projection lens 6 and the light flux l from the image side
The image of the mark RR is formed on the minute aperture member 8 provided on the stage 7 after being emitted as ₂ .

The micro aperture member 8 is combined with a photoelectric detector 9 as a photoelectric conversion means for receiving light passing through the micro aperture member 8 and outputting an electric signal, and the aperture surface of the micro aperture member 8 is a stage. The height of the surface of the wafer 10 on the wafer 7 is substantially the same as that of the wafer 10. Therefore, the minute aperture member 8 and the photoelectric detector 9 are integrally moved up and down as the wafer holder 11 is moved up and down. There is.

For such focusing, a gap sensor 12 for measuring the distance between the projection lens 6 and the surface of the wafer 10 (or the opening surface of the minute opening member 8) is provided. The focus sensor can be automatically adjusted by the vertical movement mechanism of the gap sensor 12 and the wafer holder 11. When the circuit pattern of the reticle 5 is printed on the wafer 10, the height of the surface of the wafer 10 is detected and projection with high contrast is always performed. The image can be transferred.

On the other hand, the position of the stage 7 in the first orthogonal coordinate system XY is determined by measuring the distance to the reflecting mirror fixed to the stage 7 by a laser interferometer using laser light. In FIG. 1, only the laser interferometer 13 and the reflecting mirror 14 in the X-axis direction (left and right direction on the paper surface) are shown, but the Y-axis direction (front and back surface on the paper surface) orthogonal to the X-axis forming the moving plane of the stage 7 is shown. It goes without saying that another combination of the laser interferometer and the reflecting mirror is similarly provided for the (direction).

With these laser interferometers, the position coordinate values of the stage 7 with respect to the origin of the first Cartesian coordinate system XY preset in the apparatus are successively measured during the step movement. The origin of the system XY is on the optical axis of the projection lens 6 in this embodiment, and therefore the X
Both the laser interferometers in the axial and Y-axis directions are arranged so that the intersection of the two measurement axes formed by the respective laser beams is located on the optical axis of the projection lens 6.

The reticle holder 15 holds the reticle 5 and is two-dimensionally movable in a second orthogonal coordinate system xy in a plane parallel to the XY plane of the first orthogonal coordinate system, and the reticle alignment control described later is performed. The reticle 5 is positioned by drive control by the system.

Further, the shutters 4b and 4c arranged on both sides of the incident side of the second condenser lens 3 only receive the light incident on the marks RR and RL of the reticle 5 in the illumination light path from the illumination light source 1 to the reticle 5. Is to shield the light when desired, and the arrangement position thereof is not limited to the illustrated position.

The reticle 5 is provided with the light transmissive marks RR and RL as described above. Specifically, the marks RR and RL are specifically the circuit pattern areas of the reticle 5 as shown in FIG. It is a light-transmitting slit provided in the light-shielding portion around 5a, for example, in the x-axis direction. This mark R
In the example of FIG. 2, R and RL are provided at two positions apart from each other on the x-axis of the second orthogonal coordinate system xy with the center of the reticle 5 as the origin, but the same translucency is also provided on the y-axis. A mark may be provided. The projection lens 6 of the pattern of the reticle 5
Since the projected image due to is half-rolled with respect to the xy coordinates, in FIG.
The Y direction is shown in reverse.

Therefore, the reticle 5 is attached to the reticle holder 15
When it is mounted and fixed on, a plane including these x-axis, the optical axis of the projection lens 6 and the X-axis can be defined, and the y-axis and the optical axis of the projection lens 6 can be defined by If another plane including the Y axis and three axes can be defined, the rotational deviation of the reticle 5 with respect to the stage 7 becomes zero.

The stage 7 is provided with the minute aperture member 8 for detecting the projected images of the marks RR and RL of the reticle 5 as described above. This minute opening member 8 is shown in FIG.
As shown in (a), a chromium layer or the like is vapor-deposited on the entire surface of the circular glass plate, and a slit opening 8a is formed in a part of the chromium layer.
Is formed.

FIG. 3B is a sectional view taken along the line AA of FIG. 3A, in which the longitudinal direction of the slit opening 8a is the stage 7.
Is determined so as to coincide with the X-axis direction of the first orthogonal coordinate system XY, and the width dimension of the slit aperture 8a in the Y-axis direction is determined by the mark RR or the mark RL in consideration of the reduction rate of the projection optical system. It is set to be approximately equal to the width dimension of the projected image of the mark in the Y-axis direction when the image is formed on the opening surface of the member 8.

As described above, the origin of the first orthogonal coordinate system XY is on the optical axis of the projection lens 6. The measurement of the position coordinates of the stage 7 by the laser interferometer in the present embodiment is performed by a digital measurement method by a digital counter in each of the laser interferometers in the X-axis direction and the Y-axis direction, which is the same as the generally known method.

This counter increases or decreases its count value as the stage 7 moves. When the power of the apparatus is turned on, the origin position of the stage 7 is determined and the counter is reset to zero at the origin position. Although it is necessary to preset it to a certain fixed value, the localization of the stage 7 to the origin position can be performed using the slit opening 8a described above.

That is, in this case, the coincidence between the slit aperture 8a and the image of the mark RR (or RL) is detected by the output of the photoelectric detector 9, and at this time, the counter (in the laser interferometer in the Y-axis direction ( Y counter) is reset to zero. To be more precise, taking into account the rotational deviation of the reticle 5, the count value Yr and Y of the Y counter when the image of the mark RR and the image of the mark RL are located at the slit opening 8a, respectively.
From l, when the stage 7 is positioned in the Y-axis direction so that the Y counter becomes (Yr + Yl) / 2, the Y counter is reset to zero and the Y-axis origin positioning is performed.

The same applies to the X-axis direction. Although not shown, a pair of similar light-transmitting marks are provided on the y-axis of the reticle 5, and the minute opening member 8 extends in the direction orthogonal to the slit 8a. Another slit opening may be provided and the counter (X counter) in the laser interferometer in the X-axis direction may be reset to zero by the same procedure. In this way, the origin of the coordinate system XY is set, and thereafter, the stage 7 is sequentially measured with the position coordinates of the first orthogonal coordinate system XY with the origin as a reference during the two-dimensional movement. The counter may be reset to zero with a limit switch or the like when the stage 7 reaches a predetermined position.

The apparatus of the present embodiment is further provided with an alignment microscope (WAM: Wafer Alignment Microscope) which detects marks on the wafer 10 and is used for alignment as the second rotation error detecting means of the present invention. The arrangement is as shown in FIG. That is, this microscope photoelectrically detects a wafer mark having a specific shape for alignment formed on the wafer 10 as described later, and as shown in FIG. 2, the alignment microscope WL is arranged around the lens barrel of the projection lens 6, Alignment microscope WX and alignment microscope WR are off-axis (off ax)
is) fixedly placed.

The first alignment microscope WL is for detecting the position of the wafer 10 in the Y-axis direction, and is arranged so that its detection center (observation center) is located on the Y-axis in the reference state. . The second alignment microscope WX
The position of the wafer 10 in the X-axis direction is detected, and the wafer 10 is arranged so that its detection center (observation center) is located on the X-axis in the reference state. In this way, both alignment microscopes W
The detection centers of L and the alignment microscope WX are made to coincide with each other on the Y axis and the X axis, respectively, in order to eliminate the Abbe error when detecting the position of the wafer 10.

The third alignment microscope WR is paired with the first alignment microscope WL to detect the rotational deviation of the wafer 10. The detection center (observation center) and the second alignment microscope WL in the reference state. Are arranged so that the line segment connecting to the detection center of is parallel to the X axis.

These three alignment microscopes (WL,
WX, WR) is a specific shape transferred and formed inside or in the vicinity of the circuit pattern on the wafer for alignment of overlay exposure of the second and subsequent layers in pattern printing of the first layer, For example, it is a photoelectric microscope that detects a short linear wafer mark by scanning it with a vibrating slit or a laser light vibrating beam, and the detection center thereof is set to coincide with the vibrating center of the vibrating slit or the vibrating boom. The optical system of the alignment microscope WR of No. 3 is provided with a rotatable plane parallel plate glass or the like in order to deviate the detection center in the Y-axis direction.

The main configuration of the control system for controlling the apparatus shown in FIGS. 1 and 2 is shown in the block diagram of FIG. The entire device is a microcomputer (C, etc.) including a memory so that control by a program and various arithmetic processes can be performed.
PU) 30 controls overall. The CPU 30 exchanges various types of information with a peripheral detection unit, measurement unit, or drive unit via an interface (IF) 31.

The shutter drive unit 32 receives the commands from the CPU 30 and operates the shutters 4a, 4b and 4c.
The reticle alignment control system (R-ALG) 33 moves and aligns the reticle holder 15 so that the reticle 5 is located at a predetermined position with respect to the optical axis of the projection lens 6.

On the other hand, in order to measure the position coordinates of the stage 7, the X-axis position information of the stage 7 read by the X-axis laser interferometer 13 and the Y-axis laser interferometer 34 are used. The read position information of the stage 7 in the Y-axis direction is sent to the CPU 30 via the interface 31.

Further, in order to move the stage 7 two-dimensionally, an X-axis drive unit (X- that drives the stage 7 in the X-axis direction).
ACT) 35 and a Y-axis drive unit (Y-ACT) 36 that drives the stage 7 in the Y-axis direction are provided so as to operate according to a command from the CPU 30.
A θ-axis rotation drive unit (θ-ACT) 37 for finely rotating the upper wafer holder 11 and a Z-axis drive unit for vertically moving the combination of the wafer holder 11, the fine aperture member 8 and the photoelectric detector 9 integrally. (Z-ACT) 38 is provided and is operated by a command from the CPU 30.

The photoelectric output signal of the photoelectric detector 9 is also input to the CPU 30 via the interface 31, and the focus detection section (AFD) 39 receives the signal from the gap sensor 12 shown in FIG. The information (focus information) between the surface (or the opening surface of the minute aperture member 8) and the focus position of the projection lens 6 is provided to the CPU 30 via the interface 31. CPU3
An output terminal device 40 such as a CRT display for monitoring or a printer for displaying a measurement result, an operating state, or the like is also connected to 0 via an interface 31.

As shown in FIG. 2, the off-axis arrangement of the alignment microscope WL, the alignment microscope WX, and the alignment microscope WR are photoelectric microscope or laser spot scanning type wafer mark detecting means.
When the respective detection centers coincide with a predetermined wafer mark on the wafer 10, a mark detection signal is given to the CPU 30 via the interface 31. Needless to say, the focus information from the gap sensor 12 and the focus detection unit 39 is input to the CPU 30 when the wafer marks are detected by these alignment microscopes (WL, WX, WR).

In the alignment microscope WR, a parallel plate glass is used to slightly shift the vibration center of the vibrating slit or the vibrating beam and the light-transmitting optical path of the laser light in the Y-axis direction on the first orthogonal coordinate meter XY. And prism is CPU30
It is provided so that it can be rotated or moved by a desired amount under the control of the command.

When the reticle 5 is placed on the reticle holder 15 and the reticle alignment control system 33 aligns the reticle 5 in the above-described structure, the pattern of the reticle 5 is projected onto the wafer 10 to form an optical image thereof. It becomes like FIG. This FIG. 5 shows a laser interferometer 1
3 and the laser interferometer 34, the pattern area 5 of the reticle 5 with respect to the first Cartesian coordinate XY of the stage 7
It shows a projected image 5a 'of a. The marks RR and RL of the reticle 5 are also projected as images RR 'and RL' on both sides of the projected image 5a 'along the x-axis. Here, the origin of the internal coordinate system xy of the projected image of the reticle 5 is made to coincide with the origin 0 of the orthogonal coordinate system XY of the stage 7, and the distortion caused by the projection lens 6 is treated as negligible.

Normally, the reticle-side second Cartesian coordinate system xy has an overlay error due to rotational deviation with respect to the wafer-side first Cartesian coordinate meter XY, which is exaggerated in FIG. The amount of rotational deviation is represented by the angle ε. In one embodiment of the present invention, instead of adjusting the angle ε to be negligibly small, the stage 7 is moved and controlled so that the printed pattern is substantially unaffected by the angle ε. It is intended to eliminate the overlay error due to the rotational deviation.

In FIG. 5, the mark image RL 'and the mark image RR' are part of the projected image of the reticle 5, and their y-axis coordinate values are the same. Also reticle 5
Of the reduced aperture member 8 coincides with the level of the aperture surface of the reduction aperture member 8, and the member 8 moves within the plane of this projection as the stage 7 moves.
It shall move dimensionally.

The method of measuring the angle ε will be described below.
First, focusing is performed using the gap sensor 12, the focus detection unit 39, and the Z-axis drive unit 38 so that the projection image of the reticle is formed on the opening surface of the minute opening member 8. Then, the stage 7 is moved to scan the mark image RR ′ and the mark image RL ′ through the slit opening 8a, and the output of the photoelectric detector 9 obtained during this scanning and the outputs of the laser interferometer 13 and the laser interferometer 34 are obtained. Thus, the positions of the mark image RR ′ and the mark image RL ′ in the subject 1 orthogonal coordinate system XY are measured. For example, in FIG. 5, when the slit aperture 8a moves in the Y-axis direction to scan the mark image RR 'or the mark image RL', the photoelectric detector 9 peaks at the moment when the slit aperture 8a and the mark image match. Since the value is shown, the position at which this peak value is obtained may be measured by the laser interferometer 13 and the laser interferometer 34.

The mark image RR 'measured in this way
Let YR be the Y coordinate value of the mark image RL ′, YL be the Y coordinate value of the mark image RL ′, and let l be the distance between the mark image RR ′ and the mark image RL ′ obtained from the amount of movement of the stage 7 in the X axis direction.
Is usually a very small angle, so it can be expressed as in Eq. (1) below. ε = tan ^-1 (YR-YL) / l ≈ (YR-YL) / l (1)

If the distance between the mark RR and the mark RL on the reticle 5 is known in advance, the distance l'in the x-axis direction between the mark image RR 'and the mark image RL' which is the projected image thereof.
Since we also know, the angle ε can be expressed by the following equation (2). ε = sin ⁻¹ (YR-YL) / l ′ ≈ (YR-YL) / l ′ (2) In this way, the angle ε is obtained in advance, and the stage 7 in the exposure scanning of the step-and-repeat method to be performed thereafter is selected. It is used as direction correction information for stepping movement.

In the above description, as the mark image RR 'and the mark image RL' are farther from the circuit pattern region image 5a 'in FIG. 7, the measurement accuracy of the rotational displacement amount (angle ε) is improved. Further, the longer the slit opening 8a of the minute aperture member 8 and the slit shape of the mark image RR 'and the mark image RL', the more the detected light amount increases, so that the measurement accuracy improves.

However, here, the interval l in the X-axis direction or the interval l'in the x-axis direction between the mark image RR 'and the mark image RL'.
If the length of the mark image RR 'is increased, the mark image RR' and the mark image RL 'are overlapped with the adjacent circuit pattern region image that is repeatedly exposed. The shutter 4b and the shutter 4c can selectively block only the incident light from the illumination light source on the light source 5, and the shutter 4b and the shutter 4c are opened only when the rotational deviation amount (angle ε) is measured. These should be closed during exposure.

If the measurement accuracy of the rotational displacement amount does not have to be so high, the mark image RR 'and the mark image R
A line parallel to the x-axis in the circuit pattern region image 5a ′, for example, both ends of the image of the boundary line between the pattern region and the periphery (i1, i2 in FIG. 7) is used without using a special mark such as L ′. The area shown by) may be used. The position measurement in this case is performed by measuring the laser interferometer 13 and the laser interferometer 34 at the center of the rising or falling of the output signal of the photoelectric detector 9 when the slit opening 8a crosses the bright / dark boundary of the boundary lines i1 and i2. Use value.

Next, the operation using the apparatus of this embodiment will be described. First, FIG. 6 shows an exposure operation when the first layer of the circuit pattern is printed on the wafer 10.
This will be described below with reference to FIGS. 7 and 8. In this case, since the circuit pattern and the wafer mark for alignment still do not exist on the surface of the wafer 10, the wafer 10 is a flat cutout provided on a part of the outer peripheral portion thereof as shown in FIG. It is placed on the wafer holder 11 with the (or flat) 10a as a reference and is fixed by suction.

In the present embodiment, the straight line direction of the flap 10a is positioned so as to coincide with the X-axis direction of the stage 7. Then, the surface of the wafer 10 is focused on the projection image plane by the gap sensor 12, the focus detection unit 39, and the Z-axis drive unit 38. After that, the laser interferometer 13, the laser interferometer 34, the X-axis driving unit 35, and the Y-axis driving unit 36 move the stage 7 by a predetermined distance, and the shutter 4a is opened for a predetermined time to reduce the pattern area 5a of the reticle 5. The exposure and transfer of the projected image onto the photoresist on the wafer 10 is repeated.

In this case, the shutters 4b and 4c are closed to prevent the images of the marks RR and RL from being transferred onto the wafer 10. In this way, the wafer 10
The reduced image of the pattern region 5a is transferred onto the matrix and printed on almost the entire surface of the. Here, if the step-and-repeat exposure transfer is performed with the reticle 5 having the rotational displacement of the angle ε as described above,
As shown in FIG. 6, the line segment connecting the centers of the pattern regions P ₀ and P ₁ successively transferred to the wafer 10 by the stepwise movement of the stage 7 in the X-axis direction is located on the X-axis. Each pattern area is transferred while being rotated about the X axis (or Y axis) by the deviation angle ε. Therefore, in the present embodiment, as described above, the wafer is provided with another orthogonal coordinate system αβ (hereinafter referred to as a third orthogonal coordinate system) which is rotated by the angle ε with respect to the XY coordinate without the CPU 30 using the previously determined angle ε. It is set for 10.

Now, after transferring one pattern area P ₀ , the stage 7 is used for transferring the next pattern area P _1.
In the X-axis direction, the step 7 moves along the α-axis to become P ₂ instead of the next transfer pattern region P ₁ , and the pattern region P ₂ and the pattern region P ₀ As can be seen from the arrangement of (3), the rotational deviation (angle ε) is substantially offset by making the transfer pattern on the wafer 10 the matrix arrangement coordinate of the third orthogonal coordinate system αβ.

Therefore, the array coordinates in the wafer 10 are set to the third orthogonal coordinate system αβ and the pattern area exposure position is positioned along this coordinate axis αβ, and the stage 7 has the third orthogonal coordinate system and the third orthogonal coordinate system. The stepping position is given by coordinate conversion with the Cartesian coordinate system of 1.

As shown in FIG. 7, the pattern area projected by the projection lens 6 is P ₀ , the pattern area to be exposed and transferred next is P _2, and the center of the pattern area P ₂ is O ₂ . Also, the pattern area P in the third orthogonal coordinate system α bar, etc.
The coordinates of the center O of _{0 (α 0, β 0)} , the pattern area P ₂
The coordinate value of the center O ₂ of the stage is (α ₁ , β ₁ ), and 10 is above the first orthogonal coordinate system XY of the stage 7 whose origin is the optical axis of the projection lens 6, that is, the center of the pattern region P _0. The upper array coordinate system αβ is rotated counterclockwise by the angle ε, and the origin O ₁ of the coordinate system αβ is the coordinate value (X
₀ , Y ₀ ).

[0057] pattern regions P exposure ₀ is completed, in order to match the center O ₂ and the coordinate system XY of the origin O of the pattern area P ₂ (the optical axis of the projection lens 6), a stage 7 of the current position That is, it is sufficient to move from (X ₀ , Y ₀ ) by ΔX in the X-axis direction and ΔY in the Y-direction as shown in FIG. Here, ΔX and ΔY are expressed as follows. ΔX = (α ₁ −α ₀ ) cos ε− (β ₁ −β ₀ ) sin ε ΔY = (α ₁ −α ₀ ) sin ε + (β ₁ −β ₀ ) cos ε Further, if the angle ε is sufficiently small, (3) of and
It can be approximated by equation (4). ΔX = (α ₁ −α ₀ ) − (β ₁ −β ₀ ) ε (3) ΔY = (α ₁ −α ₀ ) ε− (β ₁ −β ₀ ) (4)

The above equations (3) and (4) are programmed in the CPU 30 shown in FIG. 4, and the center position of each pattern area to be exposed and transferred is stored in advance as coordinate values of the array coordinate system αβ. The angle ε previously detected
Are used to perform the operations of the expressions (3) and (4), and as a result, the stepping movement of the stage 7 is controlled so as to cancel the angle ε.

In this stepping, the position of the stage 7 is instructed so that the counter measurement values of the laser interferometer 13 and the laser interferometer 34 are changed by ΔX and ΔY with respect to the current position (X ₀ , Y ₀ ) of the stage 7. For example, in order to expose one line parallel to the straight line of the substrate 10a of the wafer 10 by the step-and-repeat method,
Among the center coordinate values of the pattern area, β ₀ and β ₁ are equal to each other, and α ₁ −α ₀ is the pitch E _P.
From the equation and the equation (4), stepping of the stage 7 may be performed so that ΔX = E _P ΔY = E _P · ε.

For example, when nine pattern areas are transferred in a line, as shown in FIG.
Centers C _{1 to} C ₉ of _{1 to} E ₉ are all aligned on the α axis,
All the pattern areas E _{1 to} E ₉ are arranged without rotational displacement with respect to the coordinate system αβ, and the α axis is inclined by the angle ε with respect to the straight line of the furacet 10a parallel to the X axis of the coordinate system XY. Only. In this way, the second row and the third row are similarly exposed and transferred while the pattern area is set on the coordinate system αβ, so that the entire surface of the wafer 10 can be arrayed without causing the rotational displacement as shown in FIG. A transfer pattern aligned according to the coordinate system αβ can be obtained.

By the transfer of the first layer, the second layer
As described above, the wafer mark for the alignment of the superposition of the layers onward is transferred onto the street line in each pattern region or in the vicinity thereof. As described above, even under the set condition in which the reticle 5 is rotationally displaced by the angle ε, the pattern regions of the first layer transferred onto the wafer 10 are substantially affected by the angle ε. Never.

Next, the pattern superposition exposure transfer of the second and subsequent layers will be described below with reference to FIGS. 9 and 10. FIG. 9 is an explanatory diagram showing the positional relationship among the projection lens 6, the alignment microscope WL, the alignment microscope WX, and the alignment microscope WR shown in FIG. 2 on the first orthogonal coordinate XY plane, and the maximum exposure area of the projection lens 6. 6a, the center of the projection area 6b of the reticle pattern area 5a,
That is, the optical axis of the projection lens 6 is assumed to coincide with the origin of the first orthogonal coordinate XY. Each alignment microscope (WL, WX, W) arranged at a predetermined position around the projection lens 6
R) is shown by the broken circles that are used for each field of view,
The detection center of each is LC, XC, RC, and the detection center LC
And L is the interval between the detection centers RC in the X-axis direction.

FIG. 10 is a plan view schematically showing the wafer 10 on which the first layer pattern has already been transferred.
Only one row of pattern areas E _{1 to} E ₆ in the wafer 10 and its vicinity are drawn. Wafer marks for long and narrow alignment have already been formed on the upper and lower and left and right street lines of the pattern areas E _{1 to} E ₆ along each axis of β of the array coordinate system. Here, among these wafer marks, on one street line parallel to the α axis, two wafer marks AL and AR which are spaced apart from each other by approximately L, and on one street line parallel to the β axis. It is assumed that one wafer mark AX on the line is used for alignment of the wafer 10 for overlay exposure of the second and subsequent layers.

Prior to the exposure of the second layer pattern, a reticle 5 having a circuit pattern different from that of the first layer is mounted on the reticle holder 15.
Therefore, it is necessary to consider the rotational deviation due to the reticle again. Therefore, also in the pattern exposure of the second layer, first, the stage 7 is moved to detect the projected images of the transparent mark RR and the transparent mark RL of the reticle 5 by the minute aperture member 8 and the photoelectric detector 9, and the laser interference is performed. The position coordinate is obtained by the total 13 and the laser interferometer 34, and the rotation deviation is (1) or
The angle ε'is detected according to the equation (2).

Next, the detection center L of the alignment microscope WL
C and the detection center RC of the alignment microscope WR are adjusted. In this adjustment, as shown in FIG. 9, both the detection centers LC of the alignment microscope WL and the alignment microscope WR are adjusted.
And a line segment connecting RC with the RC is inclined on the XY coordinate plane by the angle ε ′ with respect to the X axis. For this purpose, the detection center RC of the alignment microscope WR may be changed in the Y-axis direction by the deviation amount δy obtained by δy≈L · ε ′ from the rotation deviation angle ε ′ of the second layer reticle.

Specifically, this can be finely adjusted by rotating or shift-moving the optical components such as the herring glass and the prism inside the optical system of the alignment microscope WR. In this case, in order to deviate the detection center RC by a predetermined amount, another detection means for detecting the rotation angle of the harving glass or the movement amount of the prism may be provided and the detection signal may be used. Is the laser interferometer 1
2. It is more convenient to use the laser interferometer 34.

This laser interferometer 13 and laser interferometer 34
In the case of using, the slit opening 8a of the fine aperture member 8 for position detection fixed to the stage 7 is used to first detect the detection center LC of the alignment microscope WL and the center of the slit opening 8a in the Y-axis direction. The stage 7 is positioned so that
In step 4, the Y coordinate value Y ^W1 at that time is measured. Next, the stage 7 is moved in the X-axis direction by the distance L so that the same slit opening 8a can be detected by the alignment microscope WR. At this time, the stage is positioned so that the Y-axis coordinate measurement value by the laser interferometer 34 becomes (y ^W1 + δy), and then the alignment microscope WR is aligned so that the detection center RC and the Y coordinate value of the center of the slit opening 8a coincide with each other. The optical components in the optical system are rotated or moved.

As described above, the alignment microscope WL
After the adjustment of the alignment microscope WR and the wafer 10 (the transfer pattern of the first layer is formed)
Is placed on the holder 11 and fixed by suction by vacuum or the like.
In this case, the wafer 10 is roughly positioned on the holder 11 using the flap set 10a. After that, the stage 7 is positioned so that the two wafer marks AL and the wafer marks AR formed on the one street line on the wafer 10 at a distance L are within the detection fields of the alignment microscope WL and the alignment microscope WR, respectively. To do. In this positioning, since the position of the wafer 10 with respect to the minute aperture member 8 on the stage 7 is substantially determined, the detection center LC of the alignment microscope WL and the detection center RC of the alignment microscope WR are preliminarily defined by the slit aperture 8 of the minute aperture member 8.
This can be easily performed by storing the coordinate value of the stage 7 when it coincides with a.

Now, when the two wafer marks AL and the wafer marks AR on the wafer 10 are caught by the alignment microscope WR and the alignment microscope WL, the detection center LC of the alignment microscope WL and the wafer mark AL coincide, and the alignment microscope WR. Of the wafer holder 11 so that the detection center RC of the
Is rotated by the θ-axis rotation drive unit 37. At this time, if the rotation center of the wafer holder 11 is set near the wafer mark AL, when the holder 11 is rotated, the deviation amount of the wafer mark AL from the detection center LC of the alignment microscope WL becomes extremely small. In this case, the wafer mark A substantially from the detection center RC of the alignment microscope WR
The wafer holder 11 is set so that the deviation of R in the Y-axis direction becomes zero.
It is only necessary to make a minute rotation or a minute movement of the stage 7 in the Y-axis direction. In this way, when the rotational position of the wafer 10 with respect to the stage 7 is determined, the holder 11 moves to the stage 7
Fixed to.

As described above, the array coordinate system αβ in the wafer 10 is positioned at an angle ε'with respect to the XY coordinate axes of the stage 7 as shown in FIG. 11, and therefore the projection pattern of the reticle 5 of the second layer is positioned. The area P ₀ and the pattern areas E _{1 to} E ₆ of the first layer which are already present on the wafer 10 are set in a state of being tilted counterclockwise by an angle ε ′ with respect to the XY coordinate axes.

When the positioning of the wafer 10 is completed, the Y coordinate value of the stage 7 when the wafer mark AL and the detection center LC of the alignment microscope WL match, and the stage 7 is moved to align the wafer mark AX. Based on the X coordinate value of the stage 7 when the detection center XC of the microscope WX coincides, the CPU 30 determines the center positional relationship between the pattern area P ₀ ′ and the pattern areas E _{1 to} E ₆ on the wafer 10. Then, the stage 7 is moved so that the pattern area P ₀ ′ overlaps with E ₁ , and the shutter 4a is opened for exposure. Then, the pattern area P ₀ 'is set to the first area equal to or less than E ₂ .
The stepping movement of the stage 7 is performed in order to sequentially superimpose it on each of the layer pattern areas.
If the stage 7 is positioned based on the equations (3) and (4), the rotational deviation of the reticle 5 at the angle ε ′ is canceled in all the pattern areas on the wafer 10, and therefore the first layer and the second layer The pattern areas of the layers are overlapped with each other without rotational displacement.

Thereafter, the alignment microscope WL and the alignment microscope WR are adjusted for the third and subsequent layers, and the step-and-repeat printing is similarly performed without the occurrence of rotational deviation.

In the above embodiment, the alignment microscope WL and the alignment microscope WR were adjusted using the fine aperture member 8 and the photoelectric detector 9, but the wafer 1
When there is a wafer mark AL on 0, the same adjustment can be performed using it. In this case, first, the wafer mark AL
Is aligned with the detection center LC of the alignment microscope WL, the stage 7 is moved in the X-axis direction, and when the alignment microscope WR captures the wafer mark AL within its detection field, the stage 7 is stopped and the detection center RC and the wafer are stopped. The position of the detection center RC is adjusted so that it matches the mark AL. Thereby, the detection center L of the alignment microscope WL
The line segment connecting C and the detection center of the alignment microscope WR becomes parallel to the X axis.

After that, the stage 7 is moved in the Y-axis direction and positioned by δy≈Lε ′ based on the angle ε ′ obtained in advance, and the detection center RC of the alignment microscope WR is detected.
The Herbung glass, prism, etc. in the optical system of the alignment microscope WR may be deviated so as to coincide with the wafer mark AL again.

The above embodiment is applied to a projection type exposure apparatus, but the present invention is not limited to this, and the mask and the wafer are faced with each other through a minute gap, and an exposure light beam or light is applied from the mask side. Is also applied to a step-and-repeat exposure apparatus of a so-called proximity method (proximity method) which irradiates X-rays, transfers a mask pattern image on a wafer, moves the mask in parallel for a certain distance, and repeats this. It goes without saying that it is applicable.

[0076]

As described above, according to the present invention,
Since the rotational displacement of the mask (reticle) with respect to the moving coordinate system of the object to be exposed (wafer) can be quantitatively and accurately obtained in a short time, and the wafer moving position can be corrected prior to exposure, circuit patterns that are overprinted are reticles. It is possible to effectively prevent the positional deviation caused by the rotational deviation of the reticle, and the rotational deviation of the reticle does not remain during the subsequent stepping by only adjusting the alignment microscope once for each reticle. Therefore, the superposition error due to the rotational deviation can be prevented by the adjustment in a very short time.

Further, according to the present invention, the plurality of marks on the reticle are illuminated by the means for illuminating the pattern of the reticle for the overlay exposure, and the plurality of images of the plurality of marks projected by the projection optical system are: Since a rotational displacement of the reticle with respect to the wafer moving coordinate system is obtained by photoelectrically detecting the reference mark on the substrate stage, a plurality of mark detection systems are provided corresponding to each of the plurality of marks on the reticle. There is no need, and the device configuration can be greatly simplified. Further, since the reticle pattern and the plurality of marks are illuminated by the same illumination means, it is possible to more accurately obtain the rotational deviation of the reticle as compared with the case where separate illumination means are used.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram showing a reduction projection exposure apparatus according to an embodiment of the present invention.

FIG. 2 is a schematic diagram showing the positions of translucent marks of the reticle 5 and the arrangement of three alignment microscopes around the projection lens 6 in FIG. 1 in relation to a first orthogonal coordinate system XY and a second orthogonal coordinate system xy. It is a perspective view.

3A and 3B are schematic diagrams showing the position of the minute opening member 8 in FIG. 1, where FIG. 3A is a plan view and FIG. 3B is a sectional view taken along the line AA of FIG.

FIG. 4 is a block diagram showing a configuration of a control system of the reduction projection exposure apparatus shown in FIG.

FIG. 5 shows a first projection image of a reticle pattern on a wafer.
3A and 3B are schematic diagrams showing a case where the origins of the second orthogonal coordinate system are matched.

FIG. 6 is an explanatory diagram showing a rotational displacement of adjacent transfer patterns and a cancellation thereof when exposure is performed one by one as the stage advances.

FIG. 7 is an explanatory diagram showing a positioning coordinate conversion when the stage 7 is stepped.

FIG. 8 is a schematic diagram of a transfer pattern array on a wafer for explaining that the rotational deviation is canceled with respect to the transfer pattern of the first layer.

FIG. 9 is an explanatory diagram showing a positional relationship between a projection lens 6 and each alignment microscope on an XY plane of a first orthogonal coordinate system.

FIG. 10 is a schematic plan view of a wafer to which the pattern of the first layer is transferred.

FIG. 11 is an explanatory diagram showing the superposition of the patterns of the second layer.

FIG. 12 is an explanatory diagram showing a rotational deviation between the superimposed patterns.

[Explanation of symbols]

1: Illumination optical system 2, 3: Condenser lens 4a, 4b, 4c: Shutter 5: Reticle 6: Projection lens 7: Stage 8: Micro aperture member 8a: Slit aperture 9: Photoelectric detector 10: Wafer 11: Wafer holder 12: Gap sensor 13, 34: Laser interferometer 14: Reflector 15: Reticle holder WL, WX, WR: Alignment microscope 30: Microcomputer (CPU) 31: Interface 32: Shutter drive unit 33: Reticle alignment control system 35: X-axis Drive unit 36: Y-axis drive unit 37: θ-axis drive unit 38: Z-axis drive unit 39: Focus detection unit RR, RL: translucent mark RR ', RL': mark image AL, AR, AX: wafer mark E ₁ , E ₂ , E ₃ , E ₄ , E ₅ , E ₆ , E ₇ , E ₈ , E
₉ : Transferred pattern area

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location G03F 9/00 H 9122-2H

Claims (2)

[Claims] 1. A holding means for holding a mask having a predetermined pattern formed thereon, a projection optical system for projecting the pattern onto a photosensitive substrate having a plurality of exposed regions formed in accordance with a predetermined arrangement coordinate system, and the photosensitive body. A substrate stage that holds a substrate and is two-dimensionally movable within a predetermined orthogonal coordinate system, and coordinate measurement for measuring the coordinate position of the photosensitive substrate by detecting the amount of movement of the substrate stage within the orthogonal coordinate system. Means and control means for controlling the movement of the substrate stage based on the measurement value of the coordinate measuring means, and sequentially positioning the substrate stage according to predetermined target position information to form the mask pattern. In an apparatus for exposing the plurality of exposed areas by superimposing them on each of the plurality of exposed areas, the projection images of the plurality of marks on the mask projected by the projection optical system Photoelectrically detecting means for photoelectrically detecting through a reference mark provided on a stage, and illuminating a pattern of the mask for the overlay exposure, and illuminating a plurality of marks on the mask, The illumination means for forming a projection image by the projection optical system on the reference mark, and the orthogonality based on the signal output from the photoelectric detection means in a state where the mask is attached to the holding means. First rotation error detection means for detecting an error amount in the rotation direction of the mask with respect to the coordinate system, and the photosensitive substrate mounted on the substrate stage,
By detecting the coordinate position of at least two preselected marks in the Cartesian coordinate system among the marks associated with each of the plurality of exposed regions on the photosensitive substrate, the exposure to the Cartesian coordinate system is performed. Second rotation error detecting means for detecting an error amount in a rotation direction of an array coordinate system of a plurality of exposed regions on the substrate; and the photosensitive substrate being rotated based on the two detected error amounts, and the mask And a correction means for making the amount of error in the rotational direction relative to the array coordinate system of the plurality of exposed regions on the photosensitive substrate substantially zero, and the target position information based on the amount of error in the rotational direction of the mask. Computation means for calculating a coordinate position to be positioned for each of the plurality of exposed regions on the photosensitive substrate by performing a correction operation, and the control means is a coordinate position measured by the coordinate measuring means. Projection exposure apparatus characterized by controlling the positioning of the substrate stage so as to substantially coincide with the coordinate position calculated by said calculating means. 2. A plurality of exposed areas are arranged in a predetermined coordinate system α.
A substrate stage that holds a photosensitive substrate formed according to β and that moves two-dimensionally within a predetermined orthogonal coordinate system XY, and a projection that projects a pattern formed on a mask according to a predetermined orthogonal internal coordinate system xy onto the photosensitive substrate In a device having an optical system, the substrate stage is sequentially positioned according to predetermined target position information, and the pattern is transferred to a plurality of exposed regions on the photosensitive substrate via the projection optical system. In the method of performing overlay exposure on each of them, a plurality of marks on the mask are illuminated by means for illuminating the pattern for the overlay exposure, and projected images of the plurality of marks projected by the projection optical system. Is detected photoelectrically through a reference mark provided on the substrate stage, and is detected on each of a plurality of exposed regions on the photosensitive substrate. Of the marks incidental detects the coordinate position in the orthogonal coordinate system of at least two marks preselected orthogonal internal coordinate system xy of the mask
And an array coordinate system αβ of a plurality of exposed regions on the photosensitive substrate
And a first step of detecting an error amount in the rotation direction relative to and so that the detected error amount in the rotation direction becomes substantially zero.
Secondly rotating the mask and the photosensitive substrate relative to each other
A plurality of steps on the photosensitive substrate by correcting and calculating the target position information based on the amount of error in the rotation direction of the orthogonal internal coordinate system xy of the mask with respect to the orthogonal coordinate system XY after the relative rotation. And a third step of calculating a coordinate position of each of the exposed regions to be positioned, wherein in the overlay exposure, the positioning of the substrate stage is controlled based on the calculated coordinate position. Projection exposure method.