JPH0982597A - Exposure method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the accuracy of the alignment of a mask and a substrate and the accuracy of the superposition of a mask pattern image and shot regions on the substrate, by measuring the positions on the substrate of the images of a plurality of marks for positional detection on the mask, and subjecting the result of the measurement to statistical operation to obtain the offset value of the mask pattern. SOLUTION: The position of each of at least three marks for positional detection formed on a mask 2, is measured on the substrate 8 side after they are passed through a projection optical system PL. Based on the result of the measurement, data on offset to the position of the mask 2 is subjected to statistical operation. Based on the offset data, the state of the image formation in the projection optical system PL is adjusted and the mask 2 is aligned with the substrate 8. Thereafter, the substrate 8 is exposed to the pattern of the mask 2, and then the offset data is stored. Subsequently, the state of the image formation in the projection optical system PL is adjusted and the alignment of the mask 2 and the substrate 8 is performed based on the stored offset data when the position of the mask 2 and the base line are measured afresh.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an exposure method for transferring an exposure pattern formed on a reticle or the like onto a photosensitive substrate such as a wafer, and more particularly to a precise exposure pattern of the exposure pattern and the exposed region of the photosensitive substrate. The present invention relates to alignment and superposition techniques.

[0002]

2. Description of the Related Art In manufacturing a semiconductor device, a liquid crystal display device or the like by a photolithography process, a pattern image of a photomask or a reticle (hereinafter referred to as "reticle") is coated with a photoresist through a projection optical system. A projection exposure apparatus that projects on each shot area on a wafer (or a glass plate or the like) is used. As a projection exposure apparatus, a wafer is placed on a two-dimensionally movable stage, and exposure and stepping operations of the wafer are repeated to sequentially expose a reticle pattern image to each shot area on the wafer. A step-and-repeat type exposure apparatus, a so-called stepper, is often used.

In recent years, semiconductor elements are formed by stacking a large number of layers of circuit patterns on a wafer. Therefore, when projecting and exposing the circuit patterns of the second and subsequent layers onto the wafer, each shot area and the shot area It is necessary to accurately perform alignment, that is, alignment with the pattern image of the next reticle to be exposed. A conventional wafer alignment method for a stepper or the like includes the following enhanced
Global Alignment (hereinafter referred to as "EGA")
A method is known (see, for example, JP-A-61-44429).

That is, a chip pattern including alignment marks (alignment marks) is formed on each of a large number of shot areas which are regularly arranged on a wafer based on preset arrangement coordinates. . When another pattern is overlaid on the formed pattern, when the wafer is stepped based on the set array coordinates, various factors cause overlay errors. That is, an overlay error occurs due to the residual rotation error of the wafer, the orthogonality error of the stage coordinate system (or shot array), the linear expansion and contraction of the wafer, and the offset (parallel movement) of the wafer center position.

Therefore, the average deviation between the array coordinate values actually measured for a plurality of shot areas selected on the wafer and the calculated array coordinate values for the corresponding shot areas is minimized. As described above, the predetermined transformation matrix is calculated using the least squares method. Next, based on this conversion matrix and the designed array coordinate value, the calculated array coordinate value of the position to be actually aligned is calculated, and each shot of the wafer is calculated based on the calculated coordinate value. The area was positioned.

However, even if the wafer is aligned based on the array coordinate values calculated by the EGA method as described above, if rotation or offset (parallel movement) occurs on the pattern side of the reticle, the reticle may be aligned. Pattern image does not exactly overlap with the chip pattern in each shot area on the wafer. Therefore, conventionally, the positioning accuracy of the pattern of the reticle is detected using an alignment microscope or the like, and the exposure is performed in a state where the positioning accuracy is within a predetermined allowable range.

[0007]

However, when the pattern image of the reticle and the chip pattern of each shot area on the wafer are positioned by the conventional method as described above, the complicated work is repeated every time the reticle is replaced. However, there is a problem that the throughput of the exposure process is reduced.

Conventionally, the projection magnification of the projection optical system was considered to be the designed value. However, in reality, the pattern of the reticle is affected by an error in the projection magnification of the projection optical system, a drawing error in the pattern of the reticle, or the like. In some cases, the magnification of the projected image through the projection optical system of was out of the designed value. Although such an error amount is extremely small and has not been regarded as a problem in the past, it cannot be ignored as the semiconductor element or the like to be manufactured is further miniaturized.

In view of the above problems, the present invention provides an exposure method for a projection exposure apparatus capable of improving the alignment accuracy between a mask and a photosensitive substrate and the overlay accuracy between a mask pattern image and a shot area on the substrate. With the goal. It is another object of the present invention to provide an exposure method for a projection exposure apparatus that can improve the mask alignment accuracy without degrading the throughput for each work lot composed of a plurality of wafers.

[0010]

In order to achieve the above object, in the exposure method of the present invention, first, the position of a mask on which a pattern is formed is measured, and the detection center position of a position detection device for the photosensitive substrate is measured. The baseline, which is the relative distance between the center of the mask and the projected point on the photosensitive substrate, is detected. Next, prior to the exposure of the preceding photosensitive substrate on which the pattern is exposed, the positions on the photosensitive substrate side of the image of at least three position detection marks formed on the mask through the projection optical system are measured. Then, based on this, the offset data for the previously measured mask position is calculated by statistical calculation. Next, based on this offset data, after adjusting the image formation state of the projection optical system and aligning the mask and the photosensitive substrate, the mask pattern is exposed on the photosensitive substrate and the offset data is stored. To do. After that, when the position of the mask and the baseline are newly measured after the preceding exposure of the photosensitive substrate, the image formation state of the projection optical system is adjusted based on the stored offset data. Align the
The pattern of the mask is exposed on the photosensitive substrate.

The above-mentioned offset data is, for example, a mask center position offset, a mask rotation offset, and a mask magnification error. Further, it is desirable that the measurement of the mask position is performed at the same time as the measurement of the baseline by using the reference mark formed on the stage.

[0012]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of an exposure method according to the present invention will be described below with reference to the drawings. In this embodiment, the present invention is applied to the exposure process of a reduction projection exposure apparatus (stepper) which reduces and projects a reticle pattern onto each shot area on a wafer as a photosensitive substrate by a step-and-repeat method.

FIG. 3 shows the exposure apparatus of this embodiment. In FIG. 3, the exposure light IL emitted from the illumination optical system 1 is shown.
Illuminates the reticle 2 with a substantially uniform illuminance. The reticle 2 is held on a reticle stage 3, and the reticle stage 3 is supported so as to be movable and minutely rotatable within a two-dimensional plane on a base 4. A main control system 6 that controls the operation of the entire apparatus controls the operation of the reticle stage 3 via a drive device 5 on the base 4.

Under the exposure light IL, the pattern image of the reticle 2 is projected onto each shot area on the wafer 8 via the projection optical system PL. The wafer 8 is placed on the wafer stage 10 via the wafer holder 9. The wafer stage 10 is an XY stage that two-dimensionally positions the wafer 8 in a plane perpendicular to the optical axis AX of the projection optical system PL, and a Z stage that positions the wafer 8 in a direction parallel to the optical axis AX (Z direction). , And a stage for rotating the wafer 8 minutely.

Wafer holder 9 (or wafer stage 1)
The photoelectric sensor 7 is fixed near the wafer 8 on the uppermost stage (0). The photoelectric sensor 7 photoelectrically converts the light flux transmitted through the slit plate below the slit plate on which the slits having the same pattern as the bright portion of the projected image of the alignment mark (described later) formed on the reticle 2 are formed. A photoelectric detector is arranged and the surface of the slit plate is set at the same height as the exposed surface of the wafer 8. The output signal of the photoelectric detector in the photoelectric sensor 7 is supplied to the main control system 6. In this embodiment, the position of the projection image of the alignment mark formed on the reticle 2 is measured by scanning the photoelectric sensor 7 in a plane perpendicular to the optical axis AX of the projection optical system PL.

A movable mirror 11 is fixed on the upper surface of the wafer stage 10, and a laser interferometer 12 is arranged so as to face the movable mirror 11. Although shown in a simplified manner in FIG. 2, the movable mirror 11 has an X-axis and a Y-axis as orthogonal coordinate systems in a plane perpendicular to the optical axis AX of the projection optical system PL.
It is composed of a plane mirror having a reflection surface perpendicular to the axis and a plane mirror having a reflection surface perpendicular to the Y axis. Further, the laser interferometer 12 includes two laser interferometers for X-axis that irradiate the moving mirror 11 with a laser beam along the X-axis and a Y-axis for irradiating the moving mirror 11 with a laser beam along the Y-axis. The laser interferometer is used for measuring the X coordinate and the Y coordinate of the wafer stage 10 with one laser interferometer for the X axis and one laser interferometer for the Y axis. The coordinate system (X, Y) composed of the X coordinate and the Y coordinate measured in this way is hereinafter referred to as a stage coordinate system or a stationary coordinate system.

The rotation angle of the wafer stage 10 is measured by the difference between the measured values of the two laser interferometers for the X axis.
Information on the X coordinate, the Y coordinate, and the rotation angle measured by the laser interferometer 12 is supplied to the coordinate measuring circuit 12a and the main control system 6, and the main control system 6 monitors the supplied coordinate and drives the drive device 13. Through this, the positioning operation of the wafer stage 10 is controlled. Although not shown in FIG. 2,
The reticle side is also provided with the same interferometer system as the wafer side.

An image forming characteristic control device 14 is mounted on the projection optical system PL of this embodiment. Imaging characteristic control device 14
Is adjusted, for example, by adjusting the distance between predetermined lens groups in the lens groups forming the projection optical system PL, or by adjusting the pressure of gas in the lens chamber between the predetermined lens groups.
The projection magnification and distortion of the projection optical system PL are adjusted. The operation of the imaging characteristic control device 14 is also controlled by the main control system 6. Further, wafer lot management data is input by an operator, or a bar code provided on a wafer is read by a bar code reader (not shown) to control the wafer management data (how many wafers to be exposed are in the lot. Data and the like) are stored in the main control system 6.

In this embodiment, an off-axis type alignment system 15 is arranged on the side surface of the projection optical system PL, and in this alignment system 15, the illumination light from the light source 16 is collimator lens 17, beam splitter 18, and mirror 19. And, the vicinity of the alignment mark 29 on the wafer 8 is irradiated through the objective lens 20. The reflected light from the alignment mark 29 is reflected by the objective lens 20 and the mirror 1.
9, the beam is irradiated onto the index plate 22 via the beam splitter 18 and the condenser lens 21, and an image of the alignment mark 29 is formed on the index plate 22. The baseline amount, which is the distance between the optical axis 20a of the objective lens 20 and the optical axis AX of the projection optical system PL, is measured in advance.

The light transmitted through the index plate 22 travels through the first relay lens 23 toward the beam splitter 24, and the light transmitted through the beam splitter 24 is transferred by the second relay lens 25X for X axis to the X-axis composed of a two-dimensional CCD. Is focused on the image pickup surface of the image pickup device 26X for use in the beam splitter 2
The light reflected at 4 is focused on the image pickup surface of the Y-axis image pickup device 26Y including a two-dimensional CCD by the Y-axis second relay lens 25Y. An image of the alignment mark 19 and an image of the index mark on the index plate 22 are superimposed and formed on the image pickup surfaces of the image pickup devices 26X and 26Y, respectively. The image pickup signals of the image pickup devices 26X and 26Y are both coordinate position measurement circuit 1
2a.

FIG. 4 shows the pattern on the index plate 22 of FIG. 3, in which the image 29 of the cross-shaped alignment mark 29 is formed in the center.
P is imaged, and the XP direction and Y perpendicular to the linear pattern images 29XP and 29YP orthogonal to this image 29P, respectively.
The P direction is conjugated with the X direction and the Y direction of the stage coordinate system of the wafer stage 10 of FIG. 2, respectively. Then, two index marks 31A and 31B are formed so as to sandwich the alignment mark image 29P in the XP direction, and two index marks 32A and 32B are formed so as to sandwich the alignment mark image 29P in the YP direction.

In this case, the index mark 31A,
31B and the detection area 33 surrounding the linear pattern image 29XP
The image in X is picked up by the X-axis image pickup device 26X in FIG.
An image in the detection area 33Y surrounding the index marks 32A and 32B and the linear pattern image 29YP in the P direction is imaged by the Y-axis imaging device 26Y in FIG. Furthermore, the scanning directions when reading photoelectric conversion signals from the pixels of the image pickup devices 26X and 26Y are set to the XP direction and the YP direction, respectively, and by processing the image pickup signals of the image pickup devices 26X and 26Y,
Alignment mark image 29P and index marks 31A, 31
It is possible to obtain the amount of positional deviation between B and 32A, 32B in the XP and YP directions. Therefore, in FIG.
The coordinate measuring circuit 12a determines the stage coordinate system (X) of the alignment mark 29 based on the positional relationship between the image of the alignment mark 29 on the wafer 8 and the index mark on the index plate 22 and the measurement result of the laser interferometer 12 at that time. , Y), and the coordinate values thus measured are supplied to the main control system 6.

Although not shown, a first fiducial mark detectable by the off-axis alignment system 15 and a second fiducial mark detectable via the projection optical system PL are predetermined on the wafer stage. They are juxtaposed in the designed positional relationship. FIG. 5A shows the reticle 2 of this embodiment. A circuit pattern for exposure is formed in the pattern area 41 at the center of the reticle 2. Pattern area 41
8 and the same alignment marks 43A to 43H are formed inside the region 42R conjugate with the effective exposure field of the projection optical system PL. The positions of these alignment marks 43A to 43H are predetermined on the orthogonal coordinate system (ξR, ηR) with the center of the reticle 2 as the origin (0, 0). In addition to the alignment marks 43A to 43H, reticle marks RM1 and RM2 for detecting the position of the reticle 2 are formed on the reticle 2 at diagonal positions.

FIG. 5B shows the shape of the alignment mark 43A. The alignment mark 43A is a light shielding film 44.
Line-and-space pattern 4 consisting of aperture patterns arranged at a predetermined pitch in the ξR direction against the background
5 and a line and space pattern 46 formed of opening patterns arranged at a predetermined pitch in the ηR direction. The alignment marks 43 are formed by the line and space patterns 45 and 46, respectively.
The positions of A in the ξR direction and the ηR direction are designated. The other alignment marks 43B to 43H have the same shape. In addition, the slit plate of the photoelectric sensor 7 of FIG.
(B) Line and space patterns 45 and 4
A pattern conjugate with 6 is formed.

Next, referring to the flow charts shown in FIGS. 1 and 2, the position of the wafer 8 and the reticle 2 in the operation of exposing the pattern of the reticle 2 to each shot area of the wafer 8 in this embodiment. The matching operation will be mainly described. First, the reticle 2 is loaded on the reticle stage 3 (step 101), and alignment of the reticle 2 and baseline measurement are performed simultaneously (step 10).
2). When performing alignment of the reticle 2, an alignment microscope (not shown) arranged above the end of the reticle 2 is used to set the positions of the reticle marks RM1 and RM2 to the TTL (Through The Lens) system or the TTR (Throug).
h The Reticle) method, the center positions Xc and Yc of the reticle 2 and the residual rotation amount θR of the reticle 2 are detected.
To measure. Hereinafter, such an operation is referred to as a "reticle alignment check". Such alignment check is performed by detecting the positional deviation between the reticle marks RM1 and RM2 and the second reference mark (not shown) on the wafer stage.

On the other hand, in the baseline measurement performed simultaneously with the reticle alignment check, the relative distance between the reticle center position and the detection center position of the alignment system 15 is measured. That is, the first on the wafer stage
Of the reference mark (not shown) from the detection center is detected, and based on the designed distance between the first reference mark and the second reference mark, and the amount of each position deviation detected as described above. Then, a baseline which is a relative distance between the projection points of the reticle marks RM1 and RM2 (center position of the reticle 2) on the wafer stage and the detection center point of the off-axis alignment system 15 is calculated. As described above, in this embodiment, the so-called “simultaneous baseline measurement” is performed in which the reticle alignment check and the baseline measurement are performed simultaneously.
However, for details of the simultaneous baseline measurement, refer to Japanese Patent Laid-Open No. 4-324923.

As a sensor for performing the baseline measurement, the measurement is performed by detecting the wafer mark using a CCD or the like with the wafer stage stationary.
IA (Field Image Alignment), LIA (Laser Interflight) that utilizes the interference of two light beams irradiated on the wafer mark.
erometric alignment) and LSA (Laser Step Alignment) for scanning and measuring the stage. Further, when the baseline measurement is performed, the static type sensor and the scan type sensor cannot perform the measurement at the same time, and therefore the measurement is performed separately. At this time, the center position of the reticle measured by the reticle alignment has different values. That is, since the reticle center position uses the value measured at the same time as the static sensor, the scan type sensor has a baseline offset.

The reticle alignment check described above adopts the image processing alignment (VRA) method, and enables simultaneous baseline measurement with the stage static alignment sensor such as FIA or LIA described above. Further, in the case of a stage scanning type alignment sensor such as LSA, the reticle alignment check adopts an ISS (Image Slit Sensor) method in which a slit provided on the wafer stage is used. The reticle alignment check error due to the difference in the sensor is
It is stored as an offset, and the baseline correction process is performed based on this. Next, in step 102-1 the main control system 6 determines whether or not the wafer to be loaded is the first (first) wafer in the lot. If it is at the beginning of the lot, the process proceeds to step 103, and if it is not at the beginning, the process proceeds to step 106.

Next, in step 103, the wafer stage 1 having three or more alignment marks among the eight alignment marks 43A to 43H on the reticle 2
The coordinates of the projected image on 0 in the stage coordinate system (X, Y) are measured. FIG. 6 shows projection images of the alignment marks 43A to 43H of the reticle 2 shown in FIG. 5A onto the wafer stage 10. In FIG. 6, the projection optical system PL is shown.
Eight alignment mark images 43AW to 43HW are projected in the effective exposure field 42 of FIG. Further, when the imaging state of the projection optical system PL is ideal, the projected image of the coordinate system (ξR, ηR) on the reticle 2 onto the wafer stage 10 is represented by the coordinate system (ξ, η). . In this embodiment, the designed array coordinates (CRξ) of the alignment marks 43A to 43H on the reticle 2 coordinate system (ξR, ηR).
Xn, CRηYn) (n = 1 to 8) is set on the wafer stage 1
Alignment mark image 4 on the coordinate system (ξ, η) on 0
Convert to array coordinates (CRXn, CRYn) of 3AW to 43HW. Then, the alignment mark images 43AW-43
The array coordinates in the HW stage coordinate system (X, Y) are (FR
Xn, FRYn).

In this case, the reticle 2 has the following error factors with respect to the stage coordinate system (X, Y).
The array coordinates (CRXn, CRYn) are array coordinates (FRXn,
FRYn) does not match. (1) Rotation of reticle: This is the stage coordinate system (X,
It is represented by the residual rotation error ΘR of the coordinate system (ξ, η) converted to the wafer side of the reticle with respect to Y). (2) Orthogonality of reticle coordinate system (ξ, η): This is ξ
The pattern in the axial direction and the pattern in the η-axis direction is an error that is not accurately orthogonal due to a drawing error and is represented by an orthogonality error WR. (3) Magnification error in the ξ direction and the η direction in the reticle coordinate system (ξ, η): This is when the pattern length of the reticle 2 has an error or when the projection magnification of the projection optical system PL is a design value (for example, This is an error caused by deviation from 1/5). This error is represented by magnification errors RRx and RRy in the ξ direction and the η direction, respectively. However,
The magnification error RRx in the ξ direction is the ratio between the measured value and the design value of the distance in the ξ direction between the two alignment mark images on the wafer stage, and the magnification error RRy in the η direction is the distance in the η direction between the two alignment mark images. It shall be expressed by the ratio of the actual measurement value and the design value. (4) Offset of the reticle coordinate system (ξ, η) with respect to the stage coordinate system (X, Y): This is caused by the projection image of the reticle 2 deviating from the wafer stage 10 by a small amount as a whole, which causes an offset. Error ORX, ORY
It is represented by

When the above error factors (1) to (4) are added, the coordinates (FRXn) on the stage coordinate system (X, Y) of the alignment mark image whose design coordinate value is (CRXn, CRYn). , FRYn) is expressed as follows.

[0032]

[Equation 1]

Here, when the orthogonality error WR and the residual rotation error ΘR are assumed to be small amounts, the first-order approximation is performed (Equation 1).
Is as follows.

[0034]

[Equation 2]

Next, the matrix and vector of (Equation 2) are replaced with the matrix FRn, AR and the vector CRn, OR as follows.

[0036]

(Equation 3)

As a result, (Equation 2) is expressed as follows.

[0038]

(Equation 4)

Four errors (hereinafter, also referred to as "error parameter") RRx, R in the matrix AR in (Equation 4)
The purpose is to obtain two offset errors (error parameters) ORX, ORY in Ry, WR, ΘR and the vector OR, and for this purpose, the EGA method calculation method for calculating the wafer shot arrangement can be applied. That is, since there are six error parameters to be obtained, the coordinates in the stage coordinate system (X, Y) of three or more alignment mark images from the alignment mark images 43AW to 43HW in FIG. 6 are obtained and measured. The sum of squares (residual error component) between the calculated coordinates and the calculated coordinates determined by (Equation 4) may be minimized.

The alignment mark images 43 shown in FIG.
AW to 43HW are marks in which the X mark and the Y mark are combined, and like the EGA of the wafer, there are three or more sets of the X and Y marks in the reticle. Further, the condition is that three or more alignment mark images selected from the alignment mark images 43AW to 43HW in FIG. 6 are not located on the same straight line. Further, it is desirable that the alignment mark image to be measured be distributed so as to be substantially point-symmetric with respect to the optical axis without being biased within the effective exposure field 42 of the projection optical system PL.

Therefore, in step 103, as shown by hatching in FIG. 6, for example, four alignment mark images 43BW and 43D which are substantially point symmetric with respect to the optical axis.
The array coordinates in the stage coordinate system (X, Y) of W, 43FW, 43HW are measured. At this time, the photoelectric sensor 7 shown in FIG. 3 may be scanned in the XY plane to find the position where the photoelectric conversion signal becomes maximum. Then, step 104
In the above, the six error parameters (RRx, RRy, WR, ΘR, O which satisfy (Equation 4) by the above least square method are
Calculate the value of RX, ORY). Note that the reticle measurement performed using three or more alignment marks as described above will be referred to as "reticle multipoint measurement" for convenience.
In addition, although ISS is used for this reticle multipoint measurement in the present embodiment, a slit through which light passes and a detector below it are arranged on the stage to detect the projected image of the reticle mark by these configurations, A method of measuring the position of the reticle 2 (stage slit method) may be adopted.

Among the error parameters calculated by the reticle multipoint measurement as described above, RRx, RRy, ΘR, OR
X, ORY or RRx, RRy, ΘR, ORX, ORY
Of these, necessary parameters are stored as offset data in the memory in the main control system 6 (step 10).
5). In this embodiment, the reticle multi-point measurement described above is performed only at the beginning of the lot, and the obtained offset data is stored. After that, when a new simultaneous baseline measurement is performed in the middle of a lot (for example, every time 10 wafers are exposed in the lot, or before each wafer is exposed next, the baseline measurement is performed. In the case), the stored offset data is read (step 106) and used. Further, when performing the baseline measurement for each wafer, the offset data stored every few wafers is read (step 106),
You may make it utilize this.

Although the above-mentioned method applies the normal EGA method, the measurement data of the alignment mark image shown in FIG. 6 is weighted according to the distance from the center of the reticle 2, for example. The six error parameters may be obtained by a so-called weighted EGA method. The weighted EGA method will be described later. Also,
In some cases, the orthogonality error WR is ignored (it is regarded as 0), and the magnification errors RRx and RRy are commonly used as the average value R.
It may be replaced with R (= (RRx + RRy) / 2). In this case, the matrix AR of (Equation 4) can be approximated as follows.

[0044]

(Equation 5)

Under this approximation, since the number of error parameters to be obtained is 4, the number of alignment mark images to be measured in order to apply the least square method is 2 or more in the case of a two-dimensional mark. Get better in When the number of marks is two, it is desirable to arrange these marks in the vicinity of the conventional RA mark. Next, in this embodiment, the rotation, linear expansion and contraction of the wafer are measured by a method that is an extension of the conventional EGA method. The method will be described below.

FIG. 7A shows the wafer 8 of this embodiment,
In FIG. 7A, a plurality of shot areas 27-n (n = n) are formed along the orthogonal coordinate system (α, β) on the wafer 8.
, 1, ... Are arranged in a matrix, and chip patterns are formed in each shot area 27-n by exposure and development in the previous step. In FIG. 7A, five shot areas 27 among the plurality of shot areas 27 are shown.
Only -1 to 27-5 are shown as a representative.

A reference position is defined for each shot area 27-n. For example, when the reference position is the reference point 28-n at the center of each shot area 27-n, the design coordinate values of this reference point 28-n in the coordinate system (α, β) on the wafer 8 are ( CXn, CYn). In addition, four alignment marks 29- for alignment are provided in each shot area 27-n.
n, 30-n, 34-n, and 35-n are provided together. In this case, each shot area 27-n is formed in parallel with the coordinate system (α, β) on the wafer of FIG.
Coordinate system (x, y) on the shot area as shown in (b)
Is set, the alignment marks 29-n, 30-
Design coordinates on the coordinate system (x, y) of n, 34-n, and 35-n are (S1Xn, S1Yn), (S2, respectively).
Xn, S2Yn), (S3Xn, S3Yn) and (S4X
n, S4Yn).

Returning to FIG. 7A, the wafer 8 is placed on the wafer stage 10, and the projection image of the reticle is sequentially placed on each of the plurality of shot areas in which the chip patterns are already formed by the step-and-repeat method. Exposure is performed by overlapping. At this time, the correspondence relationship between the stage coordinate system (X, Y) that defines the movement position of the wafer stage 10 and the coordinate system (α, β) of the wafer 8 itself is not necessarily the same as the relationship in the previous process. Therefore, from the design coordinate value (CXn, CYn) of the reference point 28-n of each shot area 27-n related to the coordinate system (α, β), the stage coordinate system (X,
Even if the coordinates on Y) are obtained and the wafer 8 is moved based on these coordinates, the shot areas 27-n may not be precisely aligned. Therefore, in the present embodiment, first, it is assumed that the alignment error is caused by the following four factors as in the conventional example. (1) Wafer rotation: This is the stage coordinate system (X, Y)
Is represented by the residual rotation error Θ of the wafer coordinate system (α, β) with respect to. (2) Orthogonality of stage coordinate system (X, Y): This occurs because the feed of the wafer stage 10 in the X-axis direction and the Y-axis direction is not exactly orthogonal, and is represented by an orthogonality error W. (3) α direction and β in the wafer coordinate system (α, β)
Directional linear expansion / contraction (wafer scaling): This means that the wafer 8 expands / contracts as a whole due to a processing process or the like. This expansion / contraction amount is represented by wafer scaling Rx and Ry in the α direction and the β direction, respectively. Where α
The wafer scaling Rx in the direction is the ratio between the measured value and the design value of the distance between the two points in the α direction on the wafer 8, and the wafer scaling Ry in the β direction is the ratio between the measured value and the design value between the two points in the β direction. Shall be represented by. (4) Offset of the coordinate system (α, β) on the wafer with respect to the stage coordinate system (X, Y): This is caused by the wafer 8 being entirely displaced from the wafer stage 10 by a minute amount, and the offset amount OX , OY.

When the above-mentioned error factors (1) to (4) are added, the stage coordinate system to be positioned for actual exposure for the shot area having the design coordinate value of the reference point (CXn, CYn). Coordinates on (X, Y) (C'X
n, C'Yn) is expressed as follows.

[0050]

(Equation 6)

Here, the orthogonality error W and the residual rotation error Θ
When the first-order approximation is performed assuming that is a small amount, (Equation 6) is as follows.

[0052]

(Equation 7)

Up to this point, accurate alignment of the reference position on each shot area 27-n (the reference point at the center of each shot area in this embodiment) has been described. However, even if the reference points of the shot areas are accurately aligned, the entire chip pattern in each shot area and the projected image of the reticle 2 do not always exactly overlap each other.

Next, the overlay error in each shot area will be described. As described above, in FIG. 7B, the coordinate system (x,
y) design coordinate values are (S1Xn, S1Yn) to (S1Xn, S1Yn)
4Xn, S4Yn), alignment marks 29-n, 30-n, 34-n, 35-n are formed respectively. In this embodiment, it is assumed that the overlay error in each shot area is caused by the following factors. (5) Rotation of chip pattern (chip rotation): This is because, for example, when the projection image of the reticle 2 is exposed on the wafer 8, the reticle 2 moves in the stage coordinate system (X,
This occurs when the wafer is rotating with respect to Y) or yawing is mixed in the movement of the wafer stage 10, and is represented by a rotation error θ with respect to the coordinate system (x, y) of the shot area. (6) Chip orthogonality error: This occurs due to, for example, distortion of the pattern itself on the reticle 2 or distortion (distortion aberration) of the projection optical system PL when the projection image of the reticle 2 is exposed on the wafer 8. This is an error in the orthogonality of the chip pattern and is represented by an angular error w. (7) Linear expansion / contraction of chips (chip scaling): This is an error in the projection magnification when the projection image of the reticle 2 is exposed on the wafer 8, or the wafer 8 is wholly or partially covered by the processing process of the wafer 8. It is caused by stretching. Here, the chip scaling rx in the x direction, which is the ratio between the measured value and the design value of the distance between two points in the x direction of the coordinate system (x, y) of the shot area, and the distance between the two points in the y direction, It is assumed that the linear expansion / contraction in two directions is represented by the chip scaling ry in the y direction which is the ratio of the actually measured value (or the variation of the actually measured value) to the design value.

For example, FIG. 8A shows a wafer 8A in which a rotation error and a magnification error occur in the chip pattern of each shot area 27-n formed in the previous step.
In (a), an example of a shot area in the case where there is no rotation error and magnification error is a shot area 36-6 surrounded by a broken line.
It is represented by 36-10. On the other hand, the shot areas 27-6 to 27-10 actually formed on the wafer 8A have different rotation angles and magnifications. These errors are
As shown in (b), a shot rotation error in which the shot area 27-n is inclined with respect to the original shot area 36-n, and a magnification of the shot area 27-n as shown in FIG. Can be separated into a chip scaling error different from the original magnification of the shot area 36-n.

However, in the example of FIG. 8, there is no orthogonality error w of the chip pattern, and the chip scaling rx in the x direction.
And the chip scaling ry in the y direction are equal to each other. When the above error factors (5) to (7) are added, the designed coordinate value on the shot area 27-n is (S
NXn, SNYn) (N = 1 to 4) alignment marks 29-n, 30-n, 34-n, 35-n, the coordinate system (x, y) of the shot area to be actually aligned.
The coordinate values (S'NXn, S'NYn) above are expressed as follows.

[0057]

(Equation 8)

Here, if the linear approximation is performed assuming that the orthogonality error w and the rotation error θ are minute amounts, (Equation 8) is expressed by the following equation.

[0059]

[Equation 9]

In FIG. 7B, the stage coordinate system (X, Y) of the reference point 28-n of the arbitrary shot area 27-n.
Since the array coordinate value above is (CXn, CYn), any alignment mark (2
The design coordinate values (DNXn, DNYn) on the stage coordinate system (X, Y) of 9-n or 30-n) are expressed as follows. However, as described above, the alignment marks 29-n to 35-n are distinguished by the value of N (1 to 4).

[0061]

(Equation 10)

The above-mentioned three errors (5) to (7) occur when the chip pattern is printed on the layer on which the alignment mark of each shot area on the wafer 8 is printed. In reality, the alignment marks 29-n, 30-n, 34-n, and 35-n are affected by the above-mentioned errors (2) and (3) caused by the processing process of the wafer 8.
Let (FNXn, FNYn) (N = 1 to 4) be the coordinates of the position that should actually be on the stage coordinate system (X, Y),
The coordinate values (FNXn, FNYn) are (Equation 8) and (Equation 1)
0) is expressed as follows.

[0063]

[Equation 11]

Next, in the present embodiment, in order to facilitate the application of the method of least squares, the α-direction wafer scaling Rx and the β-direction wafer scaling Ry in (Equation 11) are respectively changed to new parameters Γx and It is expressed as in the following (Equation 12) using Γy. Similarly, the chip scaling rx in the x direction and the chip scaling ry in the y direction in the (Equation 11) are expressed as in the following (Equation 11) using new parameters γx and γy, respectively.

[0065]

(Equation 12)

When the (Equation 11) is rewritten using four parameters Γx, Γy, γx, and γy showing these new changes in linear expansion and contraction, (Equation 11) is approximately as follows. Become.

[0067]

(Equation 13)

In this (Equation 13), if the two-dimensional vector is regarded as a matrix of 2 rows × 1 column, this (Equation 13) can be rewritten as a coordinate conversion equation using the following conversion matrix.

[0069]

[Equation 14]

However, each conversion matrix of (Equation 14) is defined as follows.

[0071]

(Equation 15)

The ten error parameters (θ, θ) included in the transformation matrices A, B, and O in the coordinate transformation equation of (Equation 14).
W, Γx (= Rx−1), Γy, OX, OY, θ, w, γ
x (= rx−1), γy) can be obtained by, for example, the method of least squares. In the present embodiment, the calculated coordinates of each shot area of the wafer 8 on the stage coordinate system (X, Y) and each error of the chip are obtained based on the coordinate conversion formula (Equation 14). Then, based on this, after correcting the chip rotation error, the chip magnification error, etc., each shot area of the wafer 8 is aligned with the reticle. Note that the least squares method is not necessarily applied to (Equation 14), and 10 error parameters may be obtained at the stage of (Equation 11), for example.

Further, in the present embodiment, as described below, the coordinate position of the further selected alignment mark is measured in the shot area (sample shot) preselected from the shot area on the wafer 8 and The measurement result is applied to (Equation 14) to obtain 10 parameters by the least squares method, and the array coordinates of each shot area are calculated based on these parameters.

Specifically, in step 107 of FIG. 2, the wafer 8 to be exposed this time is loaded on the wafer holder 9. A chip pattern is already formed in each shot area of the wafer 8 in the previous process. Further, as shown in FIG. 7B, four cross-shaped alignment marks 29-n, 30-n, 34-n and 35-n are formed in each shot area 27-n on the wafer 8. Has been done.

Next, in step 108 of FIG. 2, the origin of the wafer 8 is set (pre-alignment). Then, in step 109, stage coordinates of five or more alignment marks (29-n, 30-n, 34-n, or 35-n) on the wafer 8 are aligned using the off-axis alignment system 15 of FIG. The coordinate value (FMNXn, FMNYn) on the system (X, Y) is measured. Since one alignment mark has two components in the X direction and the Y direction, the values of 10 or more parameters can be determined by actually measuring the coordinate values of 5 or more alignment marks.

The alignment mark actually measured in this embodiment needs to be selected from three or more shot areas 27-n, but one shot area 27-n to 4 is not necessarily used.
It is not necessary to select the alignment marks 29-n to 35-n, and one alignment mark (29-n, 30-n, 34) is selected from each shot area 27-n.
-N or 35-n) may be selected.

In this case, the design array coordinate values (CX of the reference points 28-n of the plurality of shot areas 27-n selected on the wafer 8 on the coordinate system (α, β) on the wafer 8 (CX
n, CYn) and the designed coordinate values (relative coordinate values) (SNXn, SNYn) of the measured alignment marks in the coordinate system (x, y) on each shot area 27-n are known in advance. . Next, in step 110, (Equation 11)
On the right side of, the designed array coordinate values (CXn, CY) of the reference point of the shot area to which the measured alignment mark belongs.
n) and relative design coordinate values (SNXn, SNYn) regarding the reference point of the alignment mark, the calculated coordinate value () of the alignment mark on the stage coordinate system (X, Y) ( FNXn, F
NYn) is calculated.

Next, ten error parameters (Θ, W, Γx, Γy, O) satisfying (Equation 14) are calculated by the least squares method.
X, OY, θ, w, γx, γy). In particular,
The difference (ENX) between the actually measured coordinate value (FMNXn, FMNYn) and the calculated coordinate value (FNXn, FNYn).
n, ENY n) is considered an alignment error. Therefore, E
NXn = FMNXn-FNXn, ENYn = FMN
Yn-FNYn is established. Then, five or more sets of alignment errors (ENXn, ENYn), that is, the sum of squares of 10 or more alignment errors are sequentially partial differentiated with these 10 error parameters, and equations are created so that the respective values become 0. Then, by solving these 10 simultaneous equations, 10 error parameters can be obtained. This is the EGA calculation of this embodiment.

Then, in step 111, (Equation 4)
With reference to the rotation error ΘR of the reticle in the conversion matrix AR, the residual rotation error (wafer rotation error) Θ of the wafer in the conversion matrix A of (Equation 9) and the rotation error θ of the chip rotation in the conversion matrix B. The sum (Θ + θ) is adjusted to the rotation error ΘR. For that purpose, the wafer 8 is rotated by the wafer stage 10 in FIG. 3, and the traveling direction of the wafer stage 10 is set obliquely in accordance with the rotation of the reticle 2. Instead of rotating the wafer 8,
The reticle 2 side may be rotated.

However, when the wafer 8 is rotated, the offset error (OX, OY) of the wafer 8 may change. Therefore, after the coordinate values of the alignment marks are measured again, the conventional normal EGA is used. It is necessary to recalculate the error parameter by performing calculation. Therefore, for example, when the wafer 8 is rotated by a predetermined angle, as in the case where the error in the conventional chip pattern is not considered, the stage coordinate system (X, X, Remeasure the coordinate values in Y). Then, from the result, six error parameters (Θ, W, Rx, Ry,
The value of (OX, OY) is determined, and each shot area is aligned based on the array coordinates calculated from this result and exposure is performed. This also means to confirm whether the new residual rotation error Θ after the rotation of the wafer 8 is a value corresponding to the angle rotated in step 111.

The orthogonality error WR of the reticle 2 and the orthogonality error w of the chip cannot be corrected in a strict sense, but the error can be suppressed to a small value by appropriately rotating the reticle 2. Therefore, the orthogonality error W of the reticle 2
It is also possible to optimize the rotation amount of the reticle 2 so that the sum of the absolute values of R, the rotation error Θ of the wafer 8, the rotation error θ of the chip, and the orthogonality error w is minimized.

Next, in step 112, (Equation 4)
Magnification error RRx of the projected image of the reticle 2 in the matrix AR of
The projection magnification of the projection optical system PL is adjusted via the imaging characteristic control device 14 of FIG. 2 so as to correct RRy and the chip scaling errors rx and ry in the conversion matrix B of (Equation 9). In this case, in this embodiment, since the magnification of the projected image of the reticle 2 can be directly measured, the magnification errors RRx and RRy of the reticle 2 are exactly 1 when the first layer on the wafer 8 is exposed. The projection magnification of the projection optical system PL is adjusted so that Magnification errors RRx and RRy are exactly 1
The projection optical system PL for the pattern of the reticle 2
The projected image by is the size as designed.

As shown in FIG. 6, the alignment mark images 43AW to 43HW of the reticle 2 of this embodiment may be outside the actual rectangular exposure area within the effective exposure field 42 of the projection optical system PL. Therefore, the projection optical system P
When measuring the projection magnification of L, it is greatly affected by lens distortion. Therefore, by using a reference reticle in which an alignment mark image is projected in the exposure area in advance, offset correction of the projection magnification of the projection optical system PL is performed so that the magnification of the projection image of the reference reticle is equal to the design value. You may go. In this case, it is not necessary to correct the projection magnification when exposing the first layer on the wafer 8.

Next, when the pattern image of the reticle 2 is exposed on the second and subsequent layers of the wafer 8, the projection magnification of the projection optical system PL is corrected with reference to the chip scaling errors rx and ry. As a result, the size of the projected image of the reticle 2 becomes equal to the size of the chip pattern already formed on the wafer 2. Then, step 113 in FIG.
In the following, using the conversion matrices A and O including the elements composed of the error parameters obtained in step 110, the design array coordinate value (of the reference point 28-n of each shot area 27-n on the wafer 8 is designed as follows: By substituting CXn, CYn), a calculated array coordinate value (GXn, GYn) of the reference point 28-n on the stage coordinate system (X, Y) is obtained. However, as described above, when the wafer 8 side is rotated to correct the rotation error in step 107,
Based on the measured coordinates of the alignment mark again, the calculated array coordinate value (GXn, GYn) of each reference point 28-n on the stage coordinate system (X, Y) is obtained.

[0085]

(Equation 16)

Then, in step 114, based on the array coordinates (GXn, GYn) obtained by the calculation and the baseline amount previously obtained by the simultaneous baseline measurement, each shot area 27-n on the wafer 8 is determined. The reference points 28-n are sequentially set to the exposure field 42 of the projection optical system PL.
The shot area 2 is aligned with a predetermined position inside the shot area 2
The pattern image of the reticle 2 is projected and exposed to 7-n. Then, after the exposure of all shot areas on the wafer 8 is completed, processing such as development of the wafer 8 is performed.

After that, at the time when the exposure processing of a predetermined number of wafers in the middle of the lot or one wafer is completed, new simultaneous baseline measurement is performed. At this time, the reticle multi-point measurement (steps 103 and 104) described above is not performed, and the offset data (RRx, RRy, ΘR, ORX, ORY) that is calculated and stored in advance at the beginning of the lot is read. After that, the exposure operation is performed by the procedure described above, but the point that steps 111 and 112 are executed using the offset data read in step 106 is different from the case of processing the first wafer of the lot.

In this embodiment, as shown in (Equation 4) and (Equation 14), not only the transformation matrices A and O, but also reticle rotation error, reticle magnification error, chip rotation,
Since the parameters of the chip scaling are also taken into consideration, the effects of the rotation of the reticle 2, the magnification error of the projection optical system PL, the expansion and contraction and rotation of the chip pattern itself transferred to each shot area are suppressed, and It is possible to superimpose the chip pattern of each shot area and the projected image of the reticle pattern with higher accuracy.

In the above-described embodiment, as shown in FIG. 7B, a cross-shaped alignment mark 29-n which can be simultaneously aligned in the X and Y directions on the stage coordinate system.
2 to 35-n on the x-axis on the coordinate system of the shot area,
And two on the y-axis. However, for example, 1
If the four alignment marks are not arranged on a straight line, such an arrangement is not always necessary. For example, alignment marks are arranged at the four corners of each shot area. Further, each alignment mark may be formed on the street line area between each shot area. Also,
One-dimensional alignment mark for X direction and 1 for Y direction
Even if the dimensional alignment marks are provided separately,
If attention is paid to the arrangement, exactly the same as the above-mentioned embodiment (Equation 1)
The transformation matrices A, B and O of 4) can be obtained.

However, in the case of using a one-dimensional alignment mark (diffraction grating mark or the like) capable of detecting only the position in the X direction or the Y direction, instead of the cross mark that can specify the two-dimensional position as in the present embodiment. In order to determine the values of 10 parameters, it is necessary to actually measure the coordinate values of 10 or more one-dimensional alignment marks. Further, in the above-described embodiment, the rotation error θ of the chip rotation, the orthogonality error w of the chip, and the chip scaling rx, ry.
In order to obtain the two-dimensional alignment marks 29-n to 35-in each shot area 27-n of the wafer 8.
n is provided. However, each shot area 27-
Even if the offsets of the n reference points (in the x direction and the y direction) are taken into consideration, there are six parameters to be obtained, so that three two-dimensional alignment marks (29-n, for example, 29-n) are provided in each shot area 27-n. , 30-n and 34-n) may be provided. Thus, when using the two-dimensional alignment mark, two alignment marks are always selected. However, if it is a one-dimensional alignment mark, it is necessary to form six alignment marks in each shot area 27-n.

In the above-described embodiment, multipoint measurement is performed within each shot area on the wafer 8.
The method may be applied as it is to detect the position of one specific mark in each shot area. In this case, the matrix B in (Equation 14) may be ignored, and the wafer rotation error Θ may be matched with the reticle rotation error ΘR, for example.

In the above embodiment, the orthogonality error WR of the reticle 2 is not positively corrected. However, for example, a part of the lenses forming the projection optical system PL is formed on the toroidal surface, and a mechanism capable of rotating the lenses around the optical axis is provided. With this mechanism, the orthogonality error WR of the projected image is minimized. The lens may be rotated so that Thereby, the orthogonality error WR due to the drawing error of the reticle 2 or the like can be reduced.

Further, the alignment method in which the shot arrangement error of the wafer 8 is linear as in the above-described embodiment belongs to the EGA method. Further, in the EGA method of the above-described embodiment, the error in the chip rotation and the chip magnification (including distortion) for each shot area is constant within the same wafer, and the chip rotation and the chip magnification error are obtained. Therefore, it is desirable to provide an alignment method capable of performing highly accurate alignment by appropriately correcting the chip rotation and the chip magnification error even when the local arrangement error on the wafer and the fluctuation (non-linearity) of the distortion component are large. Be done. In the following, a second embodiment of the present invention will be described in which the above-mentioned EGA method is improved and alignment can be performed with higher accuracy. This alignment method is similar to the above-mentioned embodiment in Japanese Patent Application No. 4-2971.
The alignment method proposed in No. 21 is applied.

The second embodiment of the present invention will be described with reference to FIG. The projection exposure apparatus shown in FIG. 3 is also used in this embodiment, but in this embodiment, the EGA alignment used in the first embodiment is further improved, and the first weighted enhanced global alignment method ( Less than,
"W1-EGA method") is performed.
This W1-EGA method alignment is effective against “regular non-linear distortion”, and “even if the substrate has a regular non-linear distortion, the alignment error in the local region on the substrate is small. It is focused on "almost equal". Then, in this W1-EGA alignment, weighting is performed according to the distance to the sample shot, as described later.

FIG. 9 shows a wafer 8 to be exposed in this embodiment.
In FIG. 9, when determining the calculated coordinate position of the i-th shot area ESi on the wafer 8, this shot area ESi and m (in FIG. 8, m = 9) sample shots SA1 to SA9 are shown. Distance between LK1 and LK9
Accordingly, the weight Win is given to each of the measured coordinate positions (alignment data) of the alignment marks in the nine sample shots. Specifically, the two alignment marks MA of the sample shot SA1
A weight Wi1 corresponding to the distance LK1 is given to the measured coordinate positions of 1 and MB1. More strictly, it is desirable to give weights according to the distance from the reference point of the shot area ESi to each alignment mark in each sample shot. Further, it is not always necessary to measure the coordinates of the two alignment marks in each sample shot.

In the W1-EGA method, the residual error component Ei of the following (Equation 12) is defined instead of the residual error component of the simple sum of squares in the EGA method. In this (Equation 17), coordinate values (FMNXn, FMNY
n) is the actually measured coordinate value of the Nth alignment mark in the nth sample shot, the coordinate value (F
NXn, FNYn) are the calculated coordinate values.

[0097]

[Equation 17]

Then, ten error parameters (Θ, W, Γx, Γy, OX, OY, which satisfy (Equation 14) are set so that the residual error component Ei defined in this way is minimized.
θ, w, γx, γy) is obtained. Note that here, sample shots SA1 to SA1 used for each shot area ESi
SA9 is the same, but naturally the distance to each sample shot SAn is different for each shot area ESi. Therefore, the weight Win given to the coordinate position (alignment data) of the sample shot SAn changes for each shot area ESi. Then, error parameters (Θ, W, Rx, Ry, OX, OY, θ, w, rx,
ry) and first correct the wafer rotation error Θ in the conversion matrix A of (Equation 14) and the chip rotation error θ in the conversion matrix B, and at the same time, correct the chip magnification in the conversion matrix B of (Equation 14). Error (chip scaling γ
x, γy) is corrected.

After that, the error parameters (Θ, W, Γx,
Γy, OX, OY) transformation matrix A and O containing elements
By substituting the designed array coordinate value of the shot area ESi on the wafer 8 into (Expression 16), the stage coordinate system (X,
Y) Calculate the calculated array coordinate value. As described above, when the wafer 8 side is rotated, the array coordinate value is obtained by the normal EGA calculation based on the coordinates of the alignment mark newly measured.

Thus, in the W1-EGA method, the wafer 8 is
For each shot area ESi above, each sample shot S
The weight Win for the coordinate data of An changes. As an example, the weight Win is set to the i-th shot area ESi.
And as a function of the distance LKn between the n-th sample shot SAn. However, the parameter S is a parameter for changing the degree of weighting.

[0101]

(Equation 18)

As is clear from this (Equation 18), the shorter the distance LKn to the i-th shot area ESi, the shorter the sample shot SAn, the greater the weight Win given to the alignment data. Also, (Equation 1)
In 8), when the value of the parameter S is sufficiently large, the result of the statistical calculation process is almost equal to the result obtained by the EGA method of the above-mentioned embodiment. On the other hand, if all shot areas ESi to be exposed on the wafer 8 are sample shots SAn and the value of the parameter S is made sufficiently close to zero, the position of the wafer mark is measured for each shot area, which is a so-called die. It is almost equal to the result obtained by the by-die method. That is, in the W1-EGA method, the parameter S
The EGA method and die
It is possible to obtain an effect intermediate to that of the by-die method. For example, for a wafer having a large non-linear component, by setting the value of the parameter S small, an effect (alignment accuracy) almost equivalent to that of the die-by-die method can be obtained, and an alignment error due to the non-linear component can be reduced. It can be removed well. Further, when the measurement reproducibility of the alignment sensor is poor, the effect of the EGA method can be obtained by setting the value of the parameter S to a large value, and the alignment error can be reduced by the averaging effect. .

Furthermore, the weighting function of (Equation 18) may be separately prepared for the X coordinate and the Y coordinate of the alignment mark, and the weight Win may be set independently for the X coordinate and the Y coordinate. . In this case, the degree of nonlinear distortion of the wafer (magnitude), regularity or step pitch, that is, the distance between the centers of two adjacent shot areas (although it depends on the width of the street line on the wafer 8, it depends on the shot size). Even if the (corresponding value) is different in the X direction and the Y direction, by setting the value of the parameter S independently,
The shot arrangement error on the wafer 8 can be corrected with high accuracy. At this time, the value of the parameter S may be different between the X coordinate and the Y coordinate as described above, and the parameter S may be the same or different even if the values of the parameter S of the X coordinate and the Y coordinate are different. The value of S may be appropriately changed according to the magnitude of the “regular non-linear distortion”, the regularity, the step pitch, the measurement reproducibility of the alignment sensor, and the like.

From the above, the effect can be changed from the EGA method to the die-by-die method by appropriately changing the value of the parameter S. Therefore, for each layer, and for each component (X direction and Y direction), alignment is performed according to, for example, the characteristics of the non-linear component (for example, size, regularity, etc.), the step pitch, and the quality of the measurement reproducibility of the alignment sensor. It is possible to flexibly change and perform alignment under optimum conditions for each layer and each component.

A second weighted enhanced global alignment method (hereinafter referred to as "W2-EGA method") will now be described with reference to FIG. Here, for simplicity of explanation, it is assumed that the wafer 8 is regularly and particularly point-symmetrically non-linearly deformed, and the center of the point symmetry coincides with the center of the wafer 8 (wafer center).

FIG. 10 shows a wafer 8 to be exposed in the present embodiment. In FIG. 10, the deformation center point of the wafer 8 (the point symmetry center of the nonlinear distortion), that is, the wafer center Wc.
, And the distance (radius) between the i-th shot area ESi on the wafer 8 and LEi, the wafer centers Wc and m.
(M = 9 in FIG. 12) sample shots SA1 to S
The distance (radius) between each of A9 is LW1 to LW9
And And even in this W2-EGA system, W1-E
Similar to the GA method, the distance LEi and the distances LW1 to LW9
Accordingly, the weight Win ′ is given to each of the measured coordinate positions (alignment data) of the alignment marks in the nine sample shots SA1 to SA9. This W2-
In the EGA method, after detecting two alignment marks (MAi, MBi) for each sample shot,
Similar to 8), the residual error component Ei ′ is defined by the following (Equation 19), and the (Equation 19) is minimized (Equation 14).
Error parameters (Θ, W, Γx, Γy, OX, OY,
θ, w, γx, γy) is determined.

[0107]

[Equation 19]

Also in the W2-EGA method, the weight Win 'given to the alignment data is the same as in the W1-EGA method.
Changes for each shot area ESi, a statistical calculation is performed for each shot area ESi to obtain error parameters (Θ, W,
Γx, Γy, OX, OY, θ, w, γx, γy) to determine the chip rotation, the chip orthogonality, the chip magnification error, and the calculated array coordinate value.

Then, each shot area ES on the wafer 8
In order to change the weight Win ′ for each sample shot for each i, the weight Win ′ in (Equation 19) is set as a function of the distance (radius) LEi between the i-th shot area ESi on the wafer 8 and the wafer center Wc. It is expressed as follows. However, the parameter S is a parameter for changing the degree of weighting.

[0110]

(Equation 20)

As is clear from this (Equation 20), the sample shot SAn has a distance LWn to the wafer center Wc close to the distance LEi between the wafer center Wc and the i-th shot area ESi on the wafer W. The weight Wi given to the alignment data
n'becomes larger. In other words, the largest weight Win ′ is given to the alignment data of the sample shots located on the circle having the radius LEi centered on the wafer enter Wc, and the weight Wi to the alignment data is increased as the distance from the circle increases in the radial direction.
n'is small.

The value of the parameter S in (Equation 20) is the same as that required in the W1-EGA method, such as alignment accuracy, nonlinear distortion characteristics (for example, size, regularity, etc.),
It may be appropriately determined depending on the step pitch, the quality of measurement reproducibility of the alignment sensor, and the like. That is, when the non-linear component is relatively large, the value of the parameter S is set to be smaller so that the distance LWn from the wafer center Wc is reduced.
It is possible to reduce the influence of the sample shots that differ greatly. On the other hand, when the non-linear component is relatively small, by setting the value of the parameter S larger, it is possible to prevent the alignment accuracy of the alignment sensor (or layer) having poor measurement reproducibility from being degraded.

Further, in the W2-EGA method, a plurality of shot areas located substantially equidistant from the center of point symmetry on the wafer W, that is, a plurality of shot areas located on the same circle centered on the center of point symmetry. In each case, of course, the weight Win ′ given to the alignment data of the sample shot
Are the same. Therefore, when multiple shot areas are located on the same circle centered on the center of point symmetry,
The above weighting and statistical calculations are performed only on any one of the shot areas to obtain error parameters (Θ, W, Γx,
Γy, OX, OY, θ, w, γx, γy)
For the remaining shot areas, the chip rotation, the chip orthogonality, the chip magnification error, and the coordinate position can be determined by using the previously calculated error parameters as they are. This has the advantage of reducing the amount of calculation.

Incidentally, it is desirable that the sample shot arrangement suitable for the W2-EGA method is designated so as to be symmetrical with respect to the point symmetry center of the nonlinear distortion, that is, the wafer center Wc. It may be specified as a character shape or a cross shape. Besides that,
The arrangement may be similar to that of the W1-EGA method. If the point symmetry center of the non-linear distortion is other than the wafer center Wc, the X-shape or the cross shape may be arranged with the point symmetry center as a reference. When determining the value of the error parameter, the weighting function shown in (Equation 20) may be set independently in each of the X direction and the Y direction.
Needless to say, the weighting function may be set independently for each alignment mark.

The W1-EGA method and the W2-EGA method
In the method, when correcting the rotation error and the magnification error based on the four error parameters (θ, w, rx, ry) related to the chip pattern, the parameters obtained for each shot region are used for each shot region. Correction may be performed. Alternatively, one set of parameters obtained for each shot area may be averaged to obtain one set of parameters, and based on this parameter, the entire wafer may be corrected only once. Further, the wafer 8 may be divided into a plurality of blocks and correction may be performed for each block. Further, in the W2-EGA method, in a shot area located concentrically with respect to the center of point symmetry, it is not necessary to obtain a parameter for each shot area, and the parameter obtained for any one shot area is used. May be.

Further, in the above-described embodiment, when the correction of the reticle rotation or the correction of the projection magnification of the projection image of the reticle 2 (correction in the exposure of the second and subsequent layers) is performed, the rotation of the wafer 8 and the chip pattern side and It also considers expansion and contraction. However, for example, when it is known that the magnification error and the rotation error of the wafer 8 and the chip pattern are sufficiently small due to the exposure up to that time in the same lot, the measurement (EGA 4) of the reticle 2 (EGA) is performed. The rotation correction of the reticle or the projection magnification of the projection optical system may be performed using only the measurement result.

Further, the present invention is not limited to the step-and-repeat type exposure apparatus (for example, the reduction projection type stepper or the equal-magnification projection type stepper), but also a so-called step-and-scan exposure type exposure apparatus or a proxy. The present invention can be widely applied to a mitty type stepper (X-ray exposure device, etc.). In addition to the exposure equipment, it is an equipment (defect inspection equipment, prober, etc.) that inspects semiconductor wafers and reticles with multiple chip patterns, and uses a step-and-repeat method for each chip as a standard for inspection fields and probe needles. The present invention can be similarly applied to a device for aligning with respect to a position.

Further, when the above-described alignment method is applied to a scanning type exposure apparatus such as a step-and-scan exposure type exposure apparatus, a predetermined offset ( The scanning exposure is performed after the wafer is positioned at a position to which a value uniquely determined according to the pattern size, the reticle, the run-up section of the wafer, etc. is added. As described above, the present invention is not limited to the above-described embodiment, and can take various configurations without departing from the gist of the present invention. In the above-described embodiment, the reticle offset data is stored when the first wafer in the lot is processed. However, when the baseline measurement is performed in the middle of the lot, the reticle offset data is stored in the same manner as described above ( Alternatively, the offset data stored (or updated) in the middle of this lot may be used when performing baseline measurement when processing wafers in the subsequent lot.

[0119]

According to the present invention, the positions of the images of at least three position detecting marks on the mask on the substrate are measured, the measurement results are statistically processed, and the rotation angle of the pattern of the mask and the Since the offset value such as the orthogonality of the pattern is obtained, the alignment accuracy and the overlay accuracy of the pattern on the mask and the pattern on the substrate can be improved. Furthermore, when considering exposure control on a lot-by-lot basis, multipoint measurement of the reticle is performed only at the beginning of the lot, and when performing simultaneous baseline measurement during the subsequent lots, the multipoint measurement of the reticle is performed again. Since the alignment is performed using the previously obtained data without performing the alignment, there is an effect that the alignment precision and the overlay precision of the mask and the photosensitive substrate can be improved without lowering the throughput.

[Brief description of drawings]

FIG. 1 is a flowchart showing the first half of the operation of an embodiment of an exposure method according to the present invention.

FIG. 2 is a flowchart showing the latter half of the operation of the embodiment of the exposure method according to the present invention.

FIG. 3 is a configuration diagram showing an example of a projection exposure apparatus in which the exposure method of the present embodiment is implemented.

FIG. 4 is an enlarged view showing an image of an alignment mark on the index plate of FIG.

5A is a plan view showing a pattern arrangement of the reticle 2 used in the embodiment, and FIG. 5B is an enlarged plan view showing a configuration of an alignment mark on the reticle 2. FIG.

FIG. 6 is a plan view showing a projected image of the reticle 2 of FIG. 5 (a).

7A is a plan view showing an example of an arrangement of shot areas on a wafer of the embodiment, and FIG. 7B is a view showing FIG.
It is an enlarged plan view showing a shot area in (a).

8A is a plan view showing an example of a wafer including a rotation error of a chip pattern and an error of a chip magnification, and FIG. 8B is an explanatory view of a chip rotation error;
(C) is an explanatory view of a chip magnification error.

FIG. 9 shows W1-E in another embodiment of the present invention.
FIG. 6 is a plan view showing a shot array on a wafer on which GA alignment is performed.

FIG. 10 shows W2 in another embodiment of the present invention.
FIG. 6 is a plan view showing a shot array on a wafer on which EGA alignment is performed.

[Explanation of symbols]

 1 Illumination Optical System 2 Reticle 6 Main Control System 7 Photoelectric Sensor 8 Wafer 10 Wafer Stage 12 Laser Interferometer 14 Imaging Characteristic Control Device 15 Off-Axis Alignment System 27-1 to 27-5, 27-n Shot Area 43A ~ 43E Reticle alignment mark PL Projection optical system RM1, RM2 Reticle mark

Claims (3)

[Claims] 1. An exposure method for projecting a pattern of a mask onto a photosensitive substrate set on a stage via a projection optical system, comprising: a first step of measuring the position of the mask; position detection of the photosensitive substrate. A second step of detecting a baseline which is a relative distance between a detection center position of an apparatus and a projection point on the photosensitive substrate at a center of the mask; And a third step of measuring the positions of at least three position detection marks formed on the mask on the photosensitive substrate side of the image through the projection optical system, respectively; A fourth step of calculating offset data with respect to the position of the mask measured in the first step by statistical calculation based on the detected position of the detected position detection mark; and the fourth step. On the basis of the offset data calculated in step,
Fifth step of adjusting the image formation state of the projection optical system and aligning the mask with the photosensitive substrate, exposing the pattern of the mask on the photosensitive substrate, and storing the offset data. When the position of the mask and the baseline are newly measured after the exposure of the preceding photosensitive substrate, the connection of the projection optical system is performed based on the offset data stored in the fifth step. A sixth step of exposing the pattern of the mask on the photosensitive substrate after adjusting the image state and aligning the mask with the photosensitive substrate. . 2. The offset data of the mask position in the fourth step includes a center position offset with respect to a center position of the mask, a rotation offset amount of the mask, and a mask magnification error. The exposure method according to claim 1, wherein 3. The measurement of the position of the mask in the first step is performed simultaneously with the second step by using a reference mark formed on the stage. Or the exposure method according to 2.