JP3277929B2 - Projection exposure apparatus and method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a projection exposure apparatus for exposing a photosensitive layer applied to a substrate such as a semiconductor wafer or a glass plate for a liquid crystal, and more particularly to an off-axis type alignment system (fixed outside a projection optical system). And a method of observing a substrate only via a projection optical system.

[0002]

2. Description of the Related Art Conventionally, a projection exposure apparatus having an off-axis alignment system (hereinafter referred to as a stepper for convenience) has been disclosed in Japanese Patent Application Laid-Open Nos. 53-56975 and 56-56975.
As disclosed in JP-A-134737 and the like, a reference mark plate is fixed on a wafer stage that holds a photosensitive substrate (hereinafter, referred to as a wafer) and moves two-dimensionally by a step-and-repeat method. The reference mark plate is used to manage the distance between the off-axis alignment system and the projection optical system, that is, the so-called baseline amount. FIG.
3 is a diagram schematically illustrating the principle of baseline measurement disclosed in each of the above publications. In FIG. 4, a main condenser lens ICL uniformly illuminates a reticle (mask) R during exposure. Reticle R is held by reticle stage RST, and reticle stage RST is moved so that center CC of reticle R coincides with optical axis AX of projection lens PL. On the other hand, wafer stage WST
A reference mark FM equivalent to an alignment mark formed on the wafer surface is provided on the upper side.
Is positioned at a predetermined position in the projection field of view of projection lens PL, a mark RM of reticle R and a reference mark are provided by a TTR (through-the-reticle) type alignment system DDA provided above reticle R. FM are detected at the same time. The distance La between the mark RM and the center CC of the reticle R is a predetermined value in design, and the distance between the projection point of the mark RM and the projection point of the center CC on the image plane side (wafer side) of the projection lens PL is: La
/ M. Here, M is the magnification of the projection lens PL when viewing the reticle side from the wafer side, and M = 5 in the case of a 1/5 reduction projection lens.

[0003] Outside the projection lens PL (outside the projection field of view)
Has an off-axis wafer alignment system O
The WA is fixed. The optical axis of the wafer alignment system OWA is parallel to the optical axis AX of the projection lens PL on the projection image plane side. In the wafer alignment system OWA, a mark on the wafer or an index mark TM serving as a reference when aligning the reference mark FM is provided on the glass plate, and a projection image plane (the wafer surface or the reference mark FM) is provided. Plane).

As shown in FIG. 4, the base line amount BL is obtained by aligning the position X1 of the stage WST when the reticle mark RM and the reference mark FM are aligned with the index mark TM and the reference mark FM. The position X2 of the stage WST is measured by a laser interferometer or the like, and the difference (X1-X2) is calculated. This baseline amount BL is used as a reference amount when the mark on the wafer is later aligned by the wafer alignment system OWA and is sent immediately below the projection lens PL. That is, assuming that the distance between the center of one shot (area to be exposed) on the wafer and the mark on the wafer is XP, and the position of the wafer stage WST when the wafer mark matches the index mark TM is X3, the shot center and the reticle center In order to match with CC, the wafer stage W
ST may be moved to the position of the following equation.

X 3 -BL-XP (or X 3 -BL + XP) Note that this formula only represents one-dimensional direction in principle, and it is actually necessary to consider it in two dimensions. In addition, the TTR alignment system DDA The calculation method also differs depending on the arrangement of the mark RM, the arrangement of the wafer alignment system OWA, and the like.

In any case, after the mark position on the wafer is detected using the off-axis type wafer alignment system OWA, the wafer stage WST is fed by a fixed amount, and the pattern of the reticle R is immediately transferred onto the wafer. Exposure can be performed so as to be accurately superimposed on the shot area.

[0007]

In the prior art as described above, when the reference plate on which the reference mark FM is formed is mounted on the wafer stage WST, the reference plate is moved with respect to the reflecting surface of the movable mirror provided on the wafer stage. There may be a rotation error θf. At this time, when the substrate is exposed in a step-and-repeat manner using the reticle R precisely aligned with the reference mark FM, each shot area formed on the wafer is as if the chip rotation is based on the array coordinate system. Is transcribed as if it occurred.

The positional relationship (baseline amount BL) between the detection center point of the off-axis type alignment system OWA (the center of the index mark TM) and the projection point of the mark RM of the reticle R by the projection lens PL is measured. The relative distance is obtained by moving the wafer stage WST and using a laser interferometer. Therefore, the wafer stage WST
Is at position X1 in FIG. 4 and at position X2,
The running accuracy of wafer stage WST, particularly an error due to yawing, occurs. Usually, the coordinate position of wafer stage WST is measured by a pair of laser interferometers (for the X direction and for the Y direction) having length measuring axes orthogonal to each other in the stage movement plane. The interferometer is arranged so as to satisfy the Abbe condition with respect to the projection lens PL. Therefore, when the reference mark FM is detected by the off-axis alignment system OWA, the measured position (X2) of the mark FM includes an Abbe error. No matter how much the detection resolution of each alignment sensor (DDA, OWA) is increased due to this Abbe error or yawing of the stage WST, the measurement accuracy of the baseline amount BL naturally has a limit.

Accordingly, an object of the present invention is to reduce the chip rotation of a plurality of shot areas exposed on a substrate to a negligible level. Further, it is another object of the present invention to provide a projection exposure apparatus having a configuration and a sequence capable of minimizing deterioration of measurement accuracy due to yawing of a stage which occurs during baseline measurement.

[0010]

According to the present invention, a substrate (W) is provided.
(WST) on which a reference mark (FM) is formed and a reference plate (FP) provided on the stage
A projection exposure apparatus for exposing the substrate by projecting a pattern of the mask onto the substrate, the alignment device comprising: The mask is aligned based on a mounting error including at least a rotation error (θf) of the reference plate with respect to the stage.

Further, the present invention relates to a projection exposure method for exposing a substrate by projecting a mask pattern onto a substrate placed on a stage, wherein a reference plate on which a reference mark is formed is mounted on the stage. Has been detected, the mounting error of the reference plate with respect to the stage is detected,
When performing the alignment of the mask with a reference mark, small
At least the mask is aligned based on a mounting error including a rotation error .

The present invention also relates to a projection exposure apparatus provided with an off-axis alignment system. The off-axis alignment system referred to here is provided separately from the projection optical system and on a wafer. Alternatively, it means both a system for detecting a mark on a reference plate and a system for detecting a mark on a wafer or a reference plate only via a projection optical system without using a reticle.

Further, in the present invention, a pair of first interference sensors arranged to satisfy the Abbe condition with respect to the detection center point of the off-axis alignment system and measuring the two-dimensional coordinate position of the substrate stage (WST). (IFX1, IFY)
As in the conventional case, a pair of second interferometers (IFX2, IFY) satisfying the Abbe condition with respect to the optical axis position of the projection optical system is provided.

When the reference mark plate (FP) fixed on the substrate stage is positioned immediately below the projection optical system (PL) for alignment of the reticle with the apparatus, coordinate measurement values ( Lf) and the internal counter (UDC1) of one of the interferometers are instructed by a command (setting means) from the main control system (14) so that the coordinate measurement value (Le) by the second interferometer becomes equal to each other.
The count value (DX2) of 2) was preset to an internal counter (UDC11) of the other interferometer.

[0015]

According to the present invention, when the mounting error of the reference plate with respect to the stage is accurately measured in advance and the reticle stage is positioned so as to correct the error during reticle alignment, as shown by r2 in FIG. The reticle R is aligned with the reflection surface of the movable mirror IMx without rotation. As a result, the chip rotation of a plurality of shot areas exposed on the substrate can be reduced to a negligible level.

Further, a second interferometer that satisfies the Abbe condition with respect to the projection optical system and a first interferometer that satisfies the Abbe condition with respect to the off-axis alignment system are provided. The measurement values of the first interferometer and the second interferometer are preset identically at the reference stage position where the reference mark (FP) is located below the projection optical system at the time of the check. Therefore, after the preset, the off-axis alignment system (OW
The position information of the mark on the wafer detected by the first interferometer using A) can be directly introduced into the position control of the substrate stage based on the measurement value of the second interferometer. That is, there is no need to measure the amount of error due to yawing and add it to the baseline amount as a correction.
After presetting, switching to either the first interferometer or the second interferometer to perform the stage position control does not cause any trouble.

[0017]

1 shows a configuration of a projection exposure apparatus according to an embodiment of the present invention. FIG. 2 shows an arrangement of a wafer stage and a laser interferometer of the apparatus of FIG. 1 and a block diagram of a control system. FIG.
, A reticle R having a predetermined pattern area PA
Is held on a reticle stage (not shown), and is positioned such that the optical axis AX of the projection lens PL passes through the center point of the pattern area PA. The reticle stage is finely moved by a motor in the x direction, the y direction, and the θ (rotation about the optical axis AX) direction in the drawing, and is aligned with the wafer via the projection lens PL (die-by-die). The reticle R is moved for alignment) or alignment of the reticle R itself with the apparatus (reticle alignment). Also, at four locations around the pattern area PA of the reticle R, marks RMx 1, RMx 2, RMy 1, RM for reticle alignment (or die-by-die alignment)
y 2 is provided. The marks RMx 1 and RMx 2 are used for positioning in the x-axis direction of the rectangular coordinate system xy on the reticle side and for detecting a displacement, and the marks RMy 1 and RMy 2 are used for positioning and positioning in the y-axis direction of the rectangular coordinate system xy. Used for deviation detection. All of these four marks are arranged in the field of view of the projection lens PL.

As shown in FIG. 2, the projected image of the pattern area PA of the reticle R is a wafer W as a photosensitive substrate.
Imaged on top. The wafer W is stored in a rectangular coordinate system XY by a step-and-repeat method or a step-and-scan method.
Holder W on wafer stage WST moving in parallel
Held through H. Holder WH is stage WST
Above, it can be finely moved in the θ direction, and corrects the rotation of the wafer W after pre-alignment. Stage WS
A moving mirror IMy having a reflecting surface parallel to the X axis of the rectangular coordinate system XY and a moving mirror IMx having a reflecting surface parallel to the Y axis are fixed to two sides around T. The reflecting surface of the movable mirror IMy is irradiated with a laser beam from the laser interferometer IFY in parallel with the Y axis. Interferometer IFY causes the reflected beam from moving mirror IFY and the reflected beam from a reference mirror fixed at a predetermined position to interfere with each other, and moves moving mirror IMy, that is, the moving distance of wafer stage WST from the reference position in the Y direction. Measure.

Similarly, two laser interferometers IFX1 and IFX2 are provided opposite to the moving mirror IMx in the X direction. Of these, each extension line of the length measurement axis which is the center of the laser beam of the interferometer IFX2 and the length measurement axis (beam center) of the interferometer IFY is set to be orthogonal to the position of the optical axis AX of the projection lens PL. Is done.

In FIG. 2, a circle I centered on the optical axis AX is shown.
f is the field of view (image field) of the projection lens PL
The reference plate FP having a size including the image field If is fixed to the stage WST so as to have substantially the same height as the surface on the wafer W.

As shown in FIG. 1, this reference plate FP has
When the center point Fcc coincides with the optical axis AX of the projection lens PL, the four marks RMx 1, RMx 2,
Four reference marks FMX1, FMX2, FMY1, FMY aligned with the respective projected images of RMy1, RMy2
MY2 is formed. The four reference marks FM are passed through the projection lens PL and the reticle R, and together with four corresponding reticle marks RM, alignment systems (mark detecting means) DDX1, DDX2, DDX1 and DDX2 of a through-the-reticle method (hereinafter referred to as TTR method). It is detected by each of DDY1 and DDY2. Regarding the alignment system of the TTR system, several systems have been put to practical use and are well known, and therefore detailed description thereof is omitted here. However, a typical system is a reticle R side by a projection lens PL. The image of the reference mark FM and the image of the reticle mark RM, which are back-projected to the reticle, are captured by a television camera or the like, and the position is detected based on the image signal. R, irradiates the reference plate FP via the projection lens PL onto the reference plate FP, and diffracted or scattered light generated from each mark when the one-dimensional scanning is performed across the reticle mark RM and the reference mark FM. There is a method in which the amount of positional deviation between the reference mark FM and the reticle mark RM is detected by selectively performing photoelectric detection in a plane conjugate with the EP. The alignment system of the TTR system
In addition, it may be used for detecting a positional shift between an alignment mark (referred to as a wafer mark) attached to a shot area on the wafer W and the reticle mark RM.

In this embodiment, as shown in FIG. 1, an off-axis type alignment system (mark detection means) OWA is fixed outside the projection lens PL. The off-axis alignment system OWA includes a microscope objective lens and a television camera (CCD or the like) that captures an enlarged image of the objective lens in order to detect an alignment mark on the wafer W as an image. Since the alignment system OWA is provided separately from the projection lens PL, the alignment system OWA has an optical system (objective lens) in which achromatism is independently performed, and the observation region on the wafer W is uniformly formed via the objective lens. The light to illuminate is 2
It is made of a halogen lamp or the like having a wavelength width of about 00 to 350 nm. As a matter of course, the illuminating light removes the photosensitive band of the resist on the wafer W.

In such an off-axis alignment system OWA, for example, an index mark for detecting a positional shift of a mark on the wafer W is provided, and an enlarged image of the index mark and the wafer mark is displayed on a television. The image is taken by the camera. Therefore, the off-axis alignment system OWA
When the mark position on the wafer W is detected by the camera, the coordinate position (Xw, Y) of the wafer stage WST when the index mark and the wafer mark are imaged by the television camera.
w) is stored, and the amount of displacement (ΔX, ΔY) between the index mark and the wafer mark in the X and Y directions is obtained by image signal processing, and then the coordinate values (Xw−ΔX, Yw−ΔY)
It is necessary to calculate as Since it is necessary to read the coordinate position of wafer stage WST in this manner, as shown in FIG. 2, off-axis alignment system OWA
Is set at a point where the extension line of the measurement axis of the interferometer IFY for the Y direction and the extension line of the measurement axis of the interferometer IFX1 for the X direction are orthogonal to each other. Here, the detection center point Pd of the off-axis alignment system OWA is not necessarily limited to the optical axis position of the objective lens, but rather is a virtual point determined by an internal index mark. That is,
When the index mark is projected onto the wafer W via the objective lens, it is considered that the detection center point Pd exists at a certain position with respect to the projection point of the index mark, and the off-axis alignment system OWA detects It is considered that the positional deviation amount (ΔX, ΔY) of the wafer mark with respect to the center point Pd is determined.

As described above, if the interferometers IFY, IFX1 and the off-axis alignment system OWA are arranged, the Abbe error (sine error) becomes zero when the mark position is detected by the alignment system OWA. Similarly, the interferometers IFY and IFX2 are also arranged so as to satisfy the Abbe condition with respect to the projection lens PL.

The up / down pulse UDP1 output from the interferometer IFX1 in FIG. 2 is reversibly counted by a presettable up / down counter (UDC) 11, and the counted value DX1 is used as the coordinate value of the stage WST in the X direction. Output to the main control system 14. Similarly, interferometer IF
The up / down pulse UDP2 output from X2 is counted by the presettable UDC 12, and the counted value DX2
Is the main control system 14 as the coordinate value of the stage WST in the X direction.
Output to Similarly, for the interferometer IFY1,
The up-down pulse is counted by the UDC 10 and the count value DY is sent to the main control system 14. The up / down pulse from each of these interferometers is set to be one pulse every time the stage WST moves, for example, 0.01 μm.

The main control system 14 applies preset (or reset) signals S 1, S 2,
Outputs S3. The UDCs 11 and 12 are specially preset in response to the preset signals S1 and S2. That is, the preset signal (pulse) S
When 1 is output, the count value DX2 of the UDC 12 becomes
Preset signal (pulse) preset in DC11
When S2 is output, the count value DX1 of the UDC 11 is preset in the UDC 12. In addition, UDC10, 1
1 and 12, a zero reset signal is sent from the main control system 14 when the wafer stage WST is positioned at the origin position.

Now, the TTR alignment processing system 16
TR type alignment system DDX1, DDX2, DDY1
, DDY2, the mark position deviation information S detected by
Various alignment errors, reticle stage, or wafer stage WS based on X1, SX2, SY1, SY2.
The position correction amount of T is calculated, and the result is output to the main control system 14. The off-axis alignment processing system 18 determines the coordinates of the center point of the shot area on the wafer W and the coordinates of the shot area on the basis of the positional deviation information SGx and SGy in the X and Y directions of the plurality of wafer marks detected by the alignment system OWA. The regularity of the arrangement, the expansion and contraction of the wafer W, the distortion, and the like are obtained by calculation, and the results are output to the main control system 14. The stage driver 20 controls the motors 21 and 22 for driving the wafer stage WST in the X direction and the Y direction, respectively. The results from the processing systems 16 and 18 and U
The coordinate values (DX1, DX2,
DY), the wafer stage WST is precisely positioned.

Next, the operation of this embodiment will be described.
Since the entire sequence is generally known, a detailed description thereof will be omitted, and only a characteristic sequence will be described. First, before the wafer W is pre-aligned and mounted on the holder WH, the reticle R is mounted on the reticle stage. The reticle R is also held on the reticle stage by mechanical pre-alignment. At this time, UDC10 of interferometer IFY, IFX1, IFX2,
Assume that 11 and 12 are reset to zero at the origin position of wafer stage WST.

Next, the main control system 14 controls the motors 21 and 22 via the stage driver 20 so as to position the wafer stage WST as shown in FIG. 1 or FIG. Since the position of FIG. 1 (or FIG. 2) with respect to the origin position of wafer stage WST is predetermined, center point Fcc of reference plate FP is several μm with respect to optical axis AX of projection lens PL.
Positioning is performed with an accuracy of about m. When this positioning is achieved, wafer stage WST is moved to interferometers IFY and IFX.
The servo is locked under the control of 2.

Thereafter, the alignment system DD of the TTR system is used.
By using X1, DDX2, DDY1, and DDY2, a positional shift amount SX1, between the reticle mark RM and the reference mark FM,
SX2, SY1, SY2 are obtained. The processing system 16 determines the center point of the reticle R and the reference plate F based on these positional deviation amounts.
The amount of displacement of the P from the center point Fcc in the X and Y directions and the amount of relative rotation error in the θ direction are calculated. The main control system 14 drives a motor (not shown) for finely moving the reticle stage and the like so that the displacement and the rotation error are both driven to zero. Driving of reticle stage is TTR
Close control may be performed based on the amount of positional deviation or rotation error obtained from the processing system 16 by sequentially detecting the mark positional deviation in the alignment system, or open if there is a sensor that measures the position of the reticle stage with high accuracy. Control may be used.

By the above operation, reticle R is aligned with reference to reference mark FM on reference plate FP. The wafer stage WS for precisely aligning the reticle R with the reference plate FP in this manner
At position T, each UD of two interferometers IFX1 and IFX2
C11 and C11 are mutually preset. In particular,
At that time, the measured value DX2 of the UDC12 for the interferometer IFX2 used for the servo lock of the wafer stage WST
The main control system 14 outputs a preset signal S1 so that the signal is preset to the counter UDC11 for the other interferometer IFX1.

However, immediately before that, the measured value DX1 of the counter UDC11 is read into the main control system 14, and the measured value DX1 is read.
2 is calculated, and the difference or the read measurement value DX1 is stored. These stored values are required when restoring the mutual measurement relationship between the interferometers IFX1 and IFX2.

The above-described operation is a characteristic sequence of the present invention. After that, the wafer stage WST is moved so that the reference marks FM on the reference plate FP are sequentially detected by the off-axis alignment system OWA. After positioning, each position of wafer stage WST at that time is set to interferometer IFX1.
, IFY and store. Further, the off-axis alignment processing system 18 obtains the amount of displacement of each reference mark FM with respect to the detection center Pd, and based on the information of the amount of displacement and the stored position of each stage, the coordinates on the orthogonal coordinate system XY are determined. Detection center point Pd and reticle R
The relative position relationship (base line) between the center point of the and the projection point is calculated and stored.

Thereafter, the wafer W is placed on the holder WH, and the wafer W is set by the off-axis alignment system OWA.
When detecting a wafer mark attached to some of the above shot areas, the interferometers IFX1 and IFY are used to measure the coordinate position of the wafer stage WST. When performing exposure, the measured values of the interferometers IFX2 and IFY are used. The position of the wafer stage WST is controlled under the above conditions.

That is, wafer stage W capable of aligning reticle R with reference plate FP as a reference.
After the internal counters (UDC11, UDC12) of the two interferometers IFX1 and IFX2 are preset to the same value at the position ST, the value measured by the interferometer IFX1 when the mark is detected by the off-axis alignment system OWA. Can be directly introduced as a value necessary for positioning the wafer stage WST based on the measurement value of the interferometer IFX2 at the time of exposure. That is, a position error due to yawing of wafer stage WST does not occur between off-axis alignment and exposure.

FIG. 3 shows the sequence described above.
This will be described in more detail with reference to FIG. FIG. 3 exaggerates the positional relationship between the reference mark FM, the movable mirror IMy, and the interferometers IFY, IFX1, IFX2 during reticle alignment. First, it is assumed that the interferometer IF
The length measuring axis of Y and the length measuring axes of the interferometers IFX1 and IFX2 are assumed to be precisely orthogonal to each other, and the detection center PDx in the X direction of the off-axis alignment system OWA, the detection center PD in the Y direction, and PDy And the center point Fcc of the reference plate FP are located on the length measurement axis of the interferometer IFY. Now, when the reference plate FP is mounted on the wafer stage WST, it has a rotation error θf with respect to the reflection surface of the movable mirror IMx. Therefore, when the rotation error θf is precisely measured in advance and the reticle stage is positioned so that the error θf is corrected during reticle alignment, FIG.
As shown by r2 in the figure, the reticle R is aligned with the reflecting surface of the movable mirror IMx without rotation.
Note that r1 in FIG. 3 indicates the position of the reticle when the reticle mark RM is aligned with the reference mark FM with an error of zero, and a straight line K1 passing through the center point Fcc is a line extending through the center of the reticle R in the Y direction. When the reticle R is aligned so that the error θf is corrected, the straight line K1 becomes parallel to the reflecting surface of the movable mirror IMx.

On the other hand, in this state, the wafer stage WS
The positioning servo control of T is performed based on the interferometers IFX2 and IFX. At this time, the measured value (DX2) of the interferometer IFX2 is Le, and the measured value (D
X1) is Lf, the main control system 14 has its ΔLy (L
f-Le) is calculated and stored. The measured values Le and Lf are the values of the movable mirror I when the wafer stage WST was at the origin position.
The position is based on the position of the reflection surface of Mx, and the reference is represented by a straight line Lir in FIG. Therefore, after storing the difference ΔLy, the measured value Lf of the interferometer IFX1 is
When the UDC 11 is preset so as to be equal to the measured value Le of X2, the reference of the two interferometers IFX1 and IFX2 changes like a straight line Lir '. This straight line Lir 'is parallel to the reflecting surface of the movable mirror IMx and also parallel to the straight line K1. Accordingly, thereafter, the mark on the wafer W or the reference mark FM on the reference plate FP is detected by the off-axis alignment system OWA, and the position in the X direction is detected by the interferometer IFX.
The coordinate value detected using 1 is used as is for the interferometer IFX2.
The position of the wafer stage WST at the time of exposure is controlled by the interferometer IFX2 without performing a correction operation or the like in consideration of yawing of the stage.
The control is switched to IFY control.

When the wafer stage WST is stepped, for example, in the Y direction by the step and repeat method, the stage WST moves along the direction of the reflecting surface of the movable mirror IMx. For this reason, if the reticle R is corrected by the error θf like r2 and aligned, the rotation of each shot area with respect to the array coordinate system determined by the array of the plurality of shot areas exposed on the wafer W can be performed. The so-called chip rotation becomes so small that it can be ignored.

In this embodiment, only one off-axis alignment system OWA is provided. However, even when two or more off-axis alignment systems are provided, an interferometer satisfying the Abbe condition may be similarly added. Furthermore, the BB is also applicable to a TTL off-axis alignment system in which the image field of the projection lens is large and has a detection center at a peripheral portion in the image field (a method of detecting a wafer mark substantially only through a projection lens). The interferometer may be arranged to satisfy the conditions. In the case where the number of interferometers in the X direction is two and the number of interferometers in the Y direction are two, mutual presetting may be performed for each of the X direction and the Y direction. In the description with reference to FIG. 3, the measurement value Le (DX2) of the UDC 12 of the interferometer IFX2 is used.
) Is preset in the UDC 11 of the interferometer IFX1, but may be preset in the opposite direction. That is, the measurement value Lf of the UDC 11 may be preset in the UDC 12. However, in this case, before the presetting, the positioning control of the wafer stage WST is performed by the interferometer IFX1,
It is necessary to switch to control by IFY.

As the TTR type alignment system DDX1, DDX2, DDY1, DDY2 or the off-axis alignment system OWA shown in FIG. 1, for example, as disclosed in Japanese Unexamined Patent Application Publication No. Hei. A coherent beam of a book is irradiated on a reticle or a wafer so as to intersect at a predetermined angle, an interference fringe is formed on a grating pattern on the reticle or the wafer, and diffracted light generated from the grating pattern (± 1st order) Light or 0th order, ±
A two-beam interference alignment method may be employed, in which the secondary light is detected photoelectrically to determine the coordinate position of the lattice pattern on the wafer or the relative displacement between the lattice pattern of the reticle and the lattice pattern on the wafer. . This method is extremely effective especially when aligning the reticle R using the reference plate FP. That is, each reference mark FM on the reference plate FP is a grid pattern, each reticle mark RM on the reticle R is also a grid pattern, and each reticle mark RM and each reference mark FM are aligned in a two-beam interference alignment type TTR alignment system. Find the amount of displacement.
The measurement resolution of the displacement is determined by the interferometers IFX1 and IFX1
It can be higher than the resolution of X2 and IFY (± 0.01 μm).

Then, the reference plate FP
The reticle stage may be servo-controlled in the X, Y, and θ directions by adding the offset amount of each mark due to the mounting error θf. After reticle R is aligned, 2
The control of the stage driver 20 based on only the positional deviation detected by the TTR alignment system is started so that the reference plate FP is position-servo-controlled with respect to the reticle R in the TTR alignment system of the light beam interference alignment system,
With the position of wafer stage WST servo-locked with reference to aligned reticle R, the measured values of two interferometers IFX1 and IFX2 may be preset equally. In this case, two interferometers IFX
1 and IFX2 do not contribute to the position control of the wafer stage WST, so both may be preset.

For checking the reticle alignment or the residual error after the reticle alignment, besides the TTR alignment system, the wafer stage WS
The light emitting mark provided on the reference plate FP of T
Back-projecting to detect the relative position between the image of the light-emitting mark and the reticle mark RM, or by providing a photoelectric element under the slit-shaped mark of the reference plate FP and forming the reticle mark RM formed on the reference plate. The projection image may be received by the photoelectric element to detect the position of the projection image.

By the way, the mounting error θf of the reference plate FP
There is a method of using a trial bake as a method of accurately measuring in advance. In this case, a test reticle for stepper evaluation and the like are shown in FIG. 1 and TTR alignment systems DDX1 and DDX1
Using DX2, DDY1 and DDY2, precise alignment with reference plate FP is performed. That is, in this reticle alignment, the reference plate FP with the mounting error θf
The test reticle is aligned with respect to
The test reticle is set as r1 in FIG.

Next, in this state, a test wafer (bare silicon, etc.) with a resist applied on the wafer holder WH
Are pre-aligned on the basis of the flat OF and placed. Then, wafer stage WST is placed at a pitch corresponding to the size of the projected image of pattern area PA of the test reticle.
Is repeated in the X and Y directions. At this time, the movement of wafer stage WST in the X and Y directions follows the reflecting surfaces of movable mirrors IMx and IMy to the last.

FIG. 5 shows an exaggerated arrangement of the shot area TS of the test reticle formed on the test wafer W at the time of exposure. In FIG.
The orthogonal coordinate systems Xr and Yr are coordinate systems of the test reticle, and each of the plurality of shot areas TS on the wafer W has no rotational displacement with respect to the coordinate systems Xr and Yr.
However, since the stepping of the wafer stage WST in the step-and-repeat method is performed according to the moving mirrors IMx and IMy, the array coordinate system (orthogonal) αβ defined by the center point of each shot area TS is equal to the moving mirrors IMx and IMy.
Is parallel to each of the reflection surfaces. Therefore, as is apparent from FIG. 5, each shot area TS on the wafer has an array coordinate system αβ
As a reference, it is observed as if chip rotation had occurred.

Therefore, after the test-exposed wafer is developed, the average rotation error amount of each shot area TS with respect to the array coordinate system αβ is determined using the test mark (vernier or the like) in each shot area TS. If Δθc is obtained using the measurement function of a stepper or other measuring instruments,
The error amount Δθc is substantially equal to the mounting error θf.

Incidentally, there is a method of performing special stepping at the time of test exposure for chip rotation measurement. That is, stepping is performed so that the vernier patterns arranged at a plurality of locations on the outermost periphery of the pattern area PA of the test reticle overlap with each other when the adjacent shot areas are exposed. FIG. 6 shows a specific example.
(A), (B) and (C) show.

FIG. 6A shows an example of a vernier pattern arrangement in the pattern area PA of the test reticle TR. Each of the four sides around the area PA has the diffraction grating shape shown in FIG. A plurality of vernier patterns GP are formed at predetermined positions. Of these vernier patterns GP,
Any of the coordinates Xr and Yr formed along the sides of the pattern area PA parallel to the Yr axis are determined so that the pitch direction of the diffraction grating coincides with the Yr direction, and the pattern area PA parallel to the Xr axis. What was formed along the side of
In each case, the pitch direction of the diffraction grating is determined so as to coincide with the Xr direction. In addition, the plurality of vernier patterns G
Each of P is formed at a predetermined position with respect to the center point CR of the test reticle TR, and the positional relationship is precisely measured by another measuring device, and is obtained as the vernier arrangement accuracy (error) of the test reticle. It is assumed that

The vernier pattern GP in the form of a diffraction grating shown in FIG.
μm) to form a line and space, and each line and space is, for example, 2 μm on the wafer. Pattern area PA having such a vernier pattern GP
6 is printed on the wafer by stepping.
As shown in (C), the adjacent shot areas are step-moved so as to partially overlap each other in the X and Y step directions. FIG. 6C focuses on one of the plurality of shot areas shown in FIG. 5 as a specific shot area TSa, and exaggerates the arrangement of four adjacent shot areas TS1, TS2, TS3, and TS4. It is shown. As can be seen from FIG. 5, when the chip rotation is present, the adjacent shot areas are observed to have a relative positional shift in a direction substantially orthogonal to the stepping direction. For example, in FIG.
The shot areas TS1 and TS3 formed by the stepping in the α (X) direction are the shot areas TSa of interest.
Are shifted in the β (Y) direction with respect to the shot areas TS2 and TS4 formed by stepping in the β (Y) direction.
Are shifted in the α (X) direction with respect to the shot area TSa. Therefore, the direction shifted by the chip rotation and the pitch direction of the lattice-like vernier pattern GP are made coincident with each other in the overlap portion (hatched portion) of each shot region, and each vernier pattern in the adjacent shot region is juxtaposed with a slight shift in the stepping direction. Stepping is performed as follows.

When the thus baked wafer is developed, a resist image of a pair of lattice-shaped vernier patterns GP is formed on each of the four sides around the shot area TSa. The pair of vernier resist patterns are slightly displaced in the pitch direction due to the chip rotation. By measuring this positional deviation amount with high accuracy, the chip rotation amount Δθc, that is, the mounting error θf is obtained.

Here, as a specific example, in the vernier pattern GP in the test reticle of FIG. 6A, the center point C is located on a line passing through the reticle center point CR and parallel to the Xr axis.
Two vernier patterns G symmetrically positioned with respect to R
Pay attention to Pa and GPb. At this time, the shot area TS
In the overlap portion between the shot area TS1 and the shot area TS1 (the overlap area on the left side), the area TSa with respect to the vernier resist pattern GPa
The vernier resist pattern GPb 1 has a positive displacement in the Y direction, and the shot area TSa and the shot area TS
In the overlap portion with 3 (the overlap portion on the right side), the vernier resist pattern GPa in the region TS3 causes a negative displacement in the Y direction with respect to the vernier resist pattern GPb attached to the region TSa.

Therefore, the value obtained by converting the distance between the vernier patterns GPa and GPb on the test reticle in the Xr direction on the wafer side is denoted by LD, the amount of positional deviation in the Y direction obtained by the left overlap portion is denoted by ΔYg 1, When the amount of displacement in the Y direction obtained at the overlapped portion is represented by ΔYg 2, the chip rotation Δθc (= θf) is calculated by the following equation.

Δθc ≒ (ΔYg 1 −ΔYg 2) / LD By such a method, the relative positional shift amounts of the other plural pairs of vernier resist patterns GP in the focused shot area TSa are obtained, and a plurality of chip rotation amounts are obtained. After calculating Δθc and averaging them, a more accurate mounting error θf is obtained. Further, a shot area TSa focused on a plurality of different positions on the wafer is set, and the averaged chip rotation amount Δθc obtained at each of the plurality of positions is further averaged, so that a random stepping error of the wafer stage WST causes The effect will be reduced.

When two of the resist images of the vernier pattern GP in FIG. 6B are juxtaposed in a direction orthogonal to the pitch direction, the amount of positional deviation of the two vernier resist patterns in the pitch direction can be determined, for example, by the method disclosed in The measurement can be performed with extremely high accuracy by using a displacement measuring apparatus using interference fringes as disclosed in Japanese Patent Application Laid-Open No. 62-56818.

According to the method disclosed in this publication, two vernier resist patterns are simultaneously irradiated with two laser beams that are symmetrically inclined in the pitch direction, and each of the two vernier resist patterns has a lattice pitch Pg of 1 of the vernier resist pattern. In addition to producing an interference fringe having a pitch of / 2, a pair of frequency differences Δf are given between the two laser beams. At this time, the interference light of the ± 1st-order diffracted light vertically generated from each of the two vernier resist patterns becomes beat light having a frequency Δf, and when this interference light is individually photoelectrically detected, two AC signals (sine waves) having a frequency Δf are obtained. Is obtained. However, the two vernier resist patterns are displaced in the pitch direction (grating pitch Pg
(Within の of the above), a phase difference φ occurs between the two AC signals in proportion to the positional deviation amount. Therefore, using a phase difference meter or the like, the phase difference φ (± 180 °) between the two AC signals is obtained.
) And convert it to the amount of positional deviation.
Incidentally, if the Fourier transform is used for the phase difference measurement, a resolution of about ± 0.5 ° can be stably obtained as the resolution of the phase difference measurement. Assuming that the lattice pitch Pg is 4 μm on the wafer, the position shift of Pg / 2 corresponds to a phase difference of 360 ° (± Pg / 4 corresponds to ± 180 °), and therefore ± 0.5 °
When the phase measurement resolution is, the displacement detection resolution is ± Pg
/(4.180/0.5), which is approximately ± 2.8 nm. This value is 1 for the resolution of a normal laser interferometer for length measurement.
They are orders of magnitude higher.

Therefore, when the misalignment measurement function using such interference fringes is incorporated in the alignment system of the stepper, the developed wafer is placed on the stepper to perform automatic measurement, thereby immediately performing chip rotation. Since Δθc and the mounting error θf of the reference plate FP are obtained, there is an advantage that the operation until the values are stored as device constants in the computer of the stepper can be automatically performed. If the stepper has such a high-resolution and high-precision self-measurement system, the reference plate FP
Can also be determined by directly measuring the positional relationship between the reference marks. In this case, it is necessary to put a diffraction grating mark similar to that shown in FIG. 5 in a part of the reference mark.

[0057]

As described above, according to the present invention, the wafer W
With respect to an array coordinate system determined by the arrangement of a plurality of shot areas exposed above, rotation for each shot area, so-called chip rotation, can be made small enough to be ignored.

Further, a first interferometer satisfying the Abbe condition with respect to the off-axis alignment system and a second interferometer satisfying the Abbe condition with respect to the projection optical system are provided with respect to a reference plate on the stage. The stage position for aligning the mask is preset so that the same measurement value is obtained. After that, the mark position on the photosensitive substrate was detected using the off-axis alignment system and the first interferometer. The result can be used as it is for the stage position control at the time of exposure using the second interferometer, and there is an advantage that it is not necessary to sequentially perform a correction operation in consideration of the yawing of the stage and the like.

[Brief description of the drawings]

FIG. 1 is a perspective view showing a configuration of a projection exposure apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a configuration on a wafer stage and a configuration of a control system of the apparatus of FIG. 1;

FIG. 3 is a diagram illustrating the operation of the present embodiment.

FIG. 4 is a diagram illustrating baseline measurement in a conventional projection exposure apparatus.

FIG. 5 is a view showing a shot arrangement on a wafer when a mounting error of a reference plate is obtained by test baking.

FIG. 6 is a diagram illustrating a pattern arrangement of a test reticle and a method of stepping for chip rotation measurement.

[Explanation of Signs of Main Parts]

PL Projection lens R Reticle RMx1, RMx2, RMy1, RMy2 Reticle mark W Wafer WST Wafer stage FP Reference plate FMX1, FMX2, FMY1, FMY2 Reference mark OWA Off-axis alignment system Pd1, DX2 DDY1, DDY2 TTR alignment system IFY, IFX1, IFX2 Laser interferometer 10, 11, 12 Up / down counter (UDC) 14 Main control system S1, S2, S3 Preset signal

Claims (13)

(57) [Claims] A stage on which a substrate is placed; a reference plate provided on the stage, on which a reference mark is formed; and positioning means for aligning a mask using the reference mark. A projection exposure apparatus that exposes the substrate by projecting a pattern of a mask onto the substrate, wherein the positioning unit adjusts the position of the mask based on a mounting error including at least a rotation error of the reference plate with respect to the stage. A projection exposure apparatus for performing alignment. 2. The apparatus according to claim 1, further comprising a first and a second interferometer having length measuring axes parallel to each other for detecting position information of the stage, wherein the rotation error is irradiated by a length measuring beam of the first and the second interferometer. 2. The projection exposure apparatus according to claim 1, wherein a rotation error of the reference plate with respect to a reflection surface of the stage is obtained. 3. The apparatus according to claim 1, further comprising a mark detection unit configured to detect a relative positional relationship between the specific mark provided on the mask and the reference mark, wherein the alignment unit includes a detection result of the position detection unit;
3. The projection exposure apparatus according to claim 1, wherein the alignment of the mask is performed based on the mounting error. 4. The apparatus according to claim 1, further comprising a storage unit configured to store the mounting error, wherein the alignment unit performs alignment of the mask based on the mounting error stored in the storage unit. The projection exposure apparatus according to any one of claims 1 to 3. 5. The mounting error includes transferring at least one mark on a mask positioned with respect to a reference mark on the reference plate to a substrate disposed on the stage, and transferring the transferred mark image. The projection exposure apparatus according to claim 1, wherein the projection exposure apparatus is determined by detection. 6. The projection exposure apparatus according to claim 1, further comprising an error detection unit for detecting the mounting error. 7. The apparatus according to claim 6, wherein said error detecting means includes phase difference detecting means for detecting a positional shift of the diffraction grating pattern on the stage as phase difference information using interference fringes of the detected beam. 3. The projection exposure apparatus according to claim 1. 8. A measurement value of the first interferometer and a measurement value of the second interferometer in a state where the stage is positioned so that the fiducial mark is arranged in a projection field of view of the projection optical system. The projection exposure apparatus according to any one of claims 2 to 7, further comprising setting means for associating 9. The mark detection unit detects a reference mark on the stage via the projection optical system, and the setting unit determines a position of the stage when the mark detection unit detects the reference mark. 9. The projection exposure apparatus according to claim 8, wherein each measurement value of the first and second interferometers is associated. 10. The stage according to claim 1, wherein the setting unit detects the relative position of the stage when the mark detecting unit detects a relative positional relationship between the reference mark and a specific mark formed on the mask. The projection exposure apparatus according to claim 9, wherein each measurement value of the second interferometer is associated with each other. 11. A length measuring axis of the first interferometer is associated with a detection center point of an alignment system for detecting a mark on the substrate without passing through the projection optical system, and a length measuring axis of the second interferometer. An axis is associated with an optical axis position of the projection optical system, and using position information detected by the first interferometer when detecting the mark on the substrate by the alignment system or the reference mark, At the time of exposure, the second
The projection exposure apparatus according to claim 2, further comprising a control unit configured to control movement of the stage using position information detected by an interferometer. 12. The positioning device according to claim 11, wherein
The stage to be mounted is moved slightly to align the mask.
12. The method according to claim 1, wherein
Item 6. The projection exposure apparatus according to Item 1. 13. A projection exposure method for exposing a substrate by projecting a pattern of a mask onto a substrate mounted on a stage, wherein a reference plate on which a reference mark is formed is mounted on the stage. And the reference plate is slightly
When detecting the mounting error including the rotation error at least, when performing the alignment of the mask using the reference mark,
A projection exposure method, wherein the mask is aligned based on the mounting error.