JPH1064808A - Mask alignment method and projection exposure method - Google Patents

Abstract

(57) [Problem] To provide a method of aligning a mask in which a displacement of a projected image of a mask pattern does not occur due to adjustment of a projection magnification of a projection optical system. SOLUTION: Before a pattern image formed on a mask R is transferred onto a photosensitive substrate via a projection optical system PL, a projection magnification of the projection optical system PL is set in first and second steps. At least two alignment marks 1 formed in a predetermined positional relationship with the pattern of the mask R before and after the change
The projection positions 9a and 19b are respectively measured. Then, in the third step, the position of the mask R is adjusted based on the measurement results of the first and second steps so that the center of the projected image of the pattern does not change even when the projection magnification of the projection optical system PL is changed. Is done. Therefore, the displacement of the projection image of the mask pattern does not occur due to the adjustment of the projection magnification of the projection optical system PL.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of aligning a mask and a method of projecting and exposing the same, and more particularly, to a method of forming a mask used in a photolithography process in the manufacture of a semiconductor device or a liquid crystal display device. The present invention relates to a mask alignment method and a projection exposure method suitable for a projection exposure apparatus which transfers a pattern image onto a photosensitive substrate via a projection optical system.

[0002]

2. Description of the Related Art Conventionally, when a semiconductor element or a liquid crystal display element is manufactured by a photolithography process, a photomask or a reticle (hereinafter, collectively referred to as a “reticle”) is used.
A projection exposure apparatus is used for projecting an image of the above pattern onto each shot area on a substrate such as a wafer or a glass plate coated with a photosensitive agent via a projection optical system (hereinafter, appropriately referred to as a “photosensitive substrate”). . In recent years, as a projection exposure apparatus of this type, a photosensitive substrate is placed on a two-dimensionally movable stage, and the stage is moved (stepped) by the stage, so that a reticle pattern image is formed on the photosensitive substrate. A so-called step-and-repeat type exposure apparatus that repeats the operation of sequentially exposing each shot area, in particular, a reduction projection type exposure apparatus (stepper) is used relatively frequently.

[0003] For example, a semiconductor element is formed by superposing and exposing a multi-layer circuit pattern on a substrate (wafer). When projecting and exposing the circuit pattern of the second and subsequent layers on the wafer, the alignment between the already formed circuit pattern on the wafer and the reticle pattern image, that is, the alignment between the wafer and the reticle (alignment) Must be performed accurately.

[0004] In this positioning (overlapping),
The alignment mark on the wafer is illuminated with the light beam of the exposure wavelength, an image is formed on the reticle pattern surface, and the relative position between the image position and the alignment mark on the reticle is directly measured, so that the relative position is obtained. There is also a method of measuring and positioning the relationship, but this method is hardly used because the vicinity of the mark on the wafer is exposed, and the general method currently used is a method other than the exposure wavelength. A method of detecting the position of a mark on a photosensitive substrate with a light beam, using a projection optical system as a part of the mark detection optical system (TTL (Through The Lens) system), or using a dedicated position detection optical system (Hereinafter referred to as “off-axis method”). In these methods, the reticle is not directly aligned with the photosensitive substrate, but is indirectly adjusted via a reference mark provided in a projection exposure apparatus (generally, on a substrate stage on which the photosensitive substrate is placed). Do.

[0005] As an example, the off-axis method will be specifically described. First, prior to the overlay exposure, the reference mark is aligned with the position of the alignment mark on the reticle projected on the substrate stage, and the position of the substrate stage at that time is measured. Subsequently, the reference mark is moved below the mark detection optical system, and is aligned with the detection reference of the mark detection optical system. At this time, the position of the substrate stage is measured. The difference between these two stage positions is
The relative distance between the reticle reference position and the reference position of the mark detection optical system is referred to as a baseline amount, and the above sequence is referred to as baseline measurement.

At the time of overlay exposure on the photosensitive substrate, the position detection mark on the photosensitive substrate is aligned with the mark detection optical system, and the substrate stage is displaced from the current position of the substrate stage by a baseline amount. And the exposure is performed, the existing circuit pattern on the photosensitive substrate and the image of the reticle pattern can be superimposed.

As described above, in the conventional projection exposure apparatus, the position of the projected image of the alignment mark on the reticle on the substrate stage (the projected position of the reticle pattern on the photosensitive substrate) and the position of the mark detection optical system The absolute position of the reticle relative to the projection exposure apparatus is only required if the position of the reticle (with respect to the photosensitive substrate but not with respect to the apparatus) can be accurately determined, since the positional relationship with the detection reference need only be accurately determined. Did not need to be so strictly controlled. For example, the position of the reticle with respect to the projection optical system does not need to be set very accurately as long as it does not extend beyond the exposure area where the accuracy of the projection optical system is guaranteed. Also, there are usually multiple projection optical systems.
(For example, 20 or more) lens elements (lens elements), and the true optical axis of the projection optical system is
Since it is determined by the synthesis of the delicate center axis misalignment of each lens element, it cannot be determined by measuring not only the outer diameter standard of the projection optical system but also the lens elements at both ends. On the other hand, if there is too much room, it is necessary to create a large exposure area of the projection optical system, which is disadvantageous in cost.

For the above reasons, heretofore, conventionally, for example, a level at which mechanical adjustment is performed with reference to the outer diameter of the barrel of the projection optical system, specifically, roughly based on the outer diameter of the projection optical system (for example, 200 (About 400 μm) to adjust the position of the reticle.

As described above, in the prior art, the center of the reticle pattern does not exactly coincide with the optical axis of the projection optical system. However, as described above, the position of the reticle with respect to the projection optical system is superimposed. Is not important for
There was no particular problem.

[0010]

By the way, with the recent miniaturization of the pattern line width of a semiconductor integrated circuit or the like,
The required value of the overlay accuracy is becoming more and more strict. For this reason, a method of adjusting the magnification of the projection optical system according to the distortion of the photosensitive substrate has been implemented. That is, the position of the position detection mark formed on the photosensitive substrate is detected to determine the amount of expansion and contraction of the photosensitive substrate, and the exposure operation is performed in a state in which the magnification of the projection optical system is optimally adjusted accordingly. This is mainly due to consideration of thermal deformation (expansion and contraction) of the substrate when the substrate goes through a high-temperature process or the like.

In the above-described magnification adjustment, the magnification of the projection optical system is slightly changed from the original magnification, and as described above, in the conventional apparatus, the reticle is not used with respect to the projection optical system. And the reticle center does not exactly match the optical axis center of the projection optical system, so if the magnification of the projection optical system is changed, the center of the reticle pattern image will be slightly Will shift. That is, when the center of the reticle does not coincide with the optical axis of the projection optical system, the center of the reticle changes with respect to the projection exposure apparatus (with respect to the alignment sensor). For this reason, even when the magnification components match in the operation of adjusting the magnification, a position shift of superimposition occurs as a whole. This displacement will be described with reference to FIG. In FIG. 3, the center RCT of the reticle pattern image and the projection optical system (effective area 2)
The optical axis AX of 01) is at a position shifted obliquely. It is assumed that the reticle image 203 is formed by adjusting the projection magnification of the initial reticle image 202 in accordance with the expansion and contraction of the photosensitive substrate. At this time, the magnification changes in a direction away from the optical axis AX by an amount proportional to the distance from the optical axis AX as a fixed point. Therefore, each vertex moves on dotted lines 204, 205, and 206, respectively. At this time, the reticle image center RCT is slightly moved (shifted) along the dotted line 204 because it is slightly away from the optical axis AX. This is an error.

Normally, when the reticle position is mechanically adjusted based on the outer diameter of the projection optical system as described above, the deviation between the center of the reticle pattern image and the center of the optical axis is 200 to 400 μm.
m is expected. In this state, for example, a magnification of 20 pp
m, the center of the reticle pattern image is 4 to 8n
m. Recent projection exposure apparatuses are required to have an overlay accuracy of 100 nm or less. In addition to magnification correction, the imaging performance of the projection optical system (mainly distortion), changes in the baseline, stage control accuracy, and alignment sensors Considering that there are various factors of overlay error such as measurement error of (mark detection optical system),
It is not an acceptable level to use about 1/10 of the value only for magnification correction.

As a magnification changing mechanism, for example, a mechanism for driving a part of lens elements of a projection optical system in the optical axis direction is well known. However, the lens element cannot be accurately driven in the optical axis direction. If the image is slightly shifted or tilted in a direction perpendicular to the optical axis, the image of the reticle pattern moves. In such a case, the displacement of the image of the pattern of the reticle and the above-described shift of the center of the pattern image of the reticle and the center of the optical axis are superimposed, and there is a problem that the error when changing the magnification is further increased.

The present invention has been made under such circumstances, and an object of the present invention is to prevent a positional shift of a projected image of a mask pattern by adjusting a projection magnification of a projection optical system. An object of the present invention is to provide a method of aligning a mask.

It is another object of the present invention to prevent the occurrence of a registration error caused by a shift in the image of the mask pattern even when the registration exposure is performed while adjusting the projection magnification of the projection optical system. An object of the present invention is to provide a projection exposure method.

[0016]

According to the first aspect of the present invention, an image of a pattern formed on a mask (R) is transferred onto a photosensitive substrate (W) via a projection optical system (PL). Prior to this, a method of aligning the mask, wherein a projection position of at least two alignment marks (19a, 19b) formed on the mask (R) in a predetermined positional relationship with the pattern is determined. A first step of measuring;
After changing the projection magnification of the projection optical system (PL), the at least two alignment marks (19a, 19b)
A second step of measuring the projection position of the pattern; the center of the projected image of the pattern also changes when the projection magnification of the projection optical system (PL) is changed based on the measurement results of the first and second steps. A third step of adjusting the position of the mask so as not to perform the process.

According to this, prior to the transfer of the image of the pattern formed on the mask onto the photosensitive substrate via the projection optical system, the projection of the projection optical system is performed in the first and second steps. The projection positions of at least two alignment marks formed in a predetermined positional relationship with the mask pattern before and after the magnification change are measured. Then, in the third step, the position of the mask is adjusted based on the measurement results of the first and second steps so that the center of the projected image of the pattern does not change even when the projection magnification of the projection optical system is changed. . In other words, the mask is precisely positioned with respect to the projection optical system at a position where the imaging position of the center position of the projected image of the pattern does not change even when the projection magnification of the projection optical system changes. The displacement of the projected image of the mask pattern does not occur by adjusting the magnification. In other words,
By changing the projection magnification of the projection optical system, the image of the mask pattern can be accurately aligned with the pattern already formed on the substrate.

By the way, when exposing a plurality of masks sequentially by exchanging masks, each time the masks are exchanged, another mask after the exchange may be positioned by the above method. As in the invention described in item 2, after the replacement of the mask, another mask after the replacement may be positioned at the same position using information relating to the position of the mask in the third step. In this way, if the information related to the position of the mask in the above-described third step is left as it is, when the mask is replaced, the mask after replacement can be promptly positioned at the same position. .

Here, as the information relating to the position of the mask in the third step, the mask is located at the same position as in the third step.
That is, any information can be used as long as the information can be positioned at a position where the image formation position of the center position of the projected image of the pattern does not change even when the projection magnification of the projection optical system changes. Position information of the reference marks (15a, 15b) corresponding to the mask positions in the three steps may be used. In this case, as will be described later, when performing off-axis alignment, the procedure of baseline measurement is simplified.

Alternatively, the information relating to the position of the mask in the third step is the third information.
The information may be information obtained from the mask position adjusting means at the time of mask alignment in the process. In this case,
By adjusting the position adjusting means based on this information by, for example, the same amount, the position of the mask at the same position as in the third step, that is, even when the projection magnification of the projection optical system changes, the image forming position of the center position of the projected image of the pattern is changed. Positioning can be automatically and reliably performed at a position that does not change.

According to a fifth aspect of the present invention, there is provided a projection exposure method for transferring an image of a pattern formed on a mask (R) onto a photosensitive substrate (W) via a projection optical system (PL). Prior to the alignment of the mask (R) with the substrate (W), the mask (R) and the projection optical system (PL) are aligned by the mask alignment method according to any one of claims 1 to 4. )). According to this, the mask is precisely positioned with respect to the projection optical system at a position where the imaging position of the center position of the projected image of the pattern does not change even when the projection magnification of the projection optical system changes. By changing the projection magnification of the system,
With the mask pattern image accurately aligned with the pattern already formed on the substrate, the mask pattern image can be transferred onto the photosensitive substrate via the projection optical system, thereby achieving high precision Superimposed exposure can be realized.

[0022]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to FIG.
A description will be given based on FIG.

FIG. 1 schematically shows the configuration of a projection exposure apparatus 10 according to one embodiment. This projection exposure apparatus is a step-and-repeat type reduction projection exposure apparatus (so-called stepper).

The projection exposure apparatus 10 has an illumination system 12 for illuminating a reticle R as a mask with exposure illumination light IL.
A reticle stage 14 for holding a reticle R, a projection optical system PL for projecting an image of a pattern (original) formed on the reticle R onto a wafer W as a substrate to be exposed, and a two-dimensional reference plane for holding the wafer W Moving wafer stage W
ST, an off-state detecting an alignment mark (wafer mark) as a position detecting mark formed on the wafer W;
A main control unit including an axis-alignment microscope 18, a laser interferometer 20 for measuring the position of the wafer stage WST, a drive system 22 for driving the wafer stage WST, and a minicomputer (or microcomputer) for controlling the entire apparatus as a whole. 24 and the like.

The illumination system 12 includes a light source (a mercury lamp or an excimer laser), a shutter, a blind, an input lens, a fly-eye lens, a relay lens, a main condenser lens (all not shown), and the like. . The illumination system 12 includes illumination light I for exposure from a light source.
A reticle R, which is an original plate on which a semiconductor circuit pattern or the like is drawn by L, is illuminated with a uniform illuminance distribution and under predetermined illumination conditions. Here, the exposure illumination light IL is monochromatic light (or quasi-monochromatic light) supplied via a fly-eye lens, and its wavelength (exposure wavelength) is, for example, a mercury emission line (i-line).
Of 365 nm, and 248 of KrF excimer laser.
nm or the like.

A reticle R is fixed on the reticle stage 14 via reticle holders 24a and 24b by vacuum suction or the like. This reticle stage 14
A drive system (not shown) enables minute driving in the X direction (horizontal direction on the paper surface in FIG. 1), the Y direction (direction orthogonal to the paper surface in FIG. 1), and the θ direction (rotation direction in the XY plane). On the reticle R, a reticle alignment mark (for example,
Cross marks 19a, 19b are formed, and a pair of reticle microscopes 29A, 29B for detecting the reticle alignment marks 19a, 19b are provided above the reticle stage 14.

These alignment microscopes 29A, 29
B is an image processing type alignment sensor, and one alignment microscope 29A has a bending mirror 26a,
An optical system 27a including an objective lens and the like (not shown) and an image sensor 28a such as a CCD, and a signal processing circuit and the like (not shown);
The optical system 27a includes an index plate on which index marks are formed. As the light source of the alignment microscope 29A, for example, a part of the exposure light obtained by branching out the light from the exposure light source is used. The other alignment microscope 2
Similarly to the alignment microscope 29A, the optical system 9B includes a bending mirror 26b, an objective lens (not shown), and the like.
7b, an image pickup device 28b such as a CCD, a signal processing circuit (not shown) and the like, and an optical system 27b includes an index plate on which index marks are formed.

In this embodiment, these reticle microscopes 29A and 29B are used for reticle alignment and relative position detection between a reticle mark at the time of baseline measurement and a later-described reference mark. These reticle microscopes 2
The outputs of 9A and 29B are supplied to main controller 24, which performs reticle alignment and the like described later.

The reticle microscopes 29A and 29B are
Since it is an obstacle during the exposure operation, it can be evacuated (see FIG. 2).

The projection optical system PL is composed of a plurality of lens elements 40, 41,... Having a common optical axis so as to have a telecentric optical arrangement on both sides. Among the lens elements, the reticle stage 14
Is held by a ring-shaped support member 42, which is
Are telescopic drive elements, for example, piezo elements 44a, 4
4b, 44c (the drive element 44c on the far side of the drawing is not shown)
Thus, it is supported at three points and is connected to the lens barrel 46. The driving elements 44a, 44b, and 44c allow the three peripheral points of the lens element 40 to be independently moved in the optical axis AX direction of the projection optical system PL. That is, the lens element 40 is moved along the optical axis AX according to the displacement of the driving elements 44a, 44b, 44c.
Along the optical axis A
It can also be arbitrarily inclined with respect to a plane perpendicular to X.
The voltages applied to these drive elements 44a, 44b, 44c are controlled by a magnification correction controller 48 based on a command from the main controller 24, whereby the displacement of the drive elements 44a, 44b, 44c is controlled. It has become so. The projection optical system PL in FIG.
Is the common optical axis of the lens element 41 and other lens elements (not shown) fixed to the lens barrel 46, that is, the light determined by the synthesis of the delicate misalignment of the central axes of these lens elements. Point to the axis.

Here, the optical axis AX of the lens element 40
Since it is clear that the magnification of the projection optical system PL can be changed by moving in the direction, the driving element 44 in the present embodiment.
Magnification correction means is constituted by the magnification correction controller 48 which controls the displacement amounts of a, 44b, and 44c.

The projection optical system PL according to the present embodiment
Is defined as a Z-axis direction whose optical axis AX is orthogonal to the moving surface of the reticle stage 14, and a predetermined reduction magnification β (β is, for example, 1
/ 5) is used. Therefore, when the reticle R is illuminated by the illumination light IL with uniform illuminance in a state where the pattern of the reticle R is aligned with the shot area on the wafer W as described later, pattern formation is performed. The pattern on the surface is reduced by the projection optical system PL at a reduction magnification β, projected onto the wafer W coated with the photoresist, and reduced in a pattern on each shot area (for example, an area of each LSI chip) on the wafer W. Is formed.

The wafer W is placed on the wafer holder 30 and the wafer stage WST, and is vacuum-sucked on the wafer holder 30. Wafer stage WST is actually a Y stage that moves on a base (not shown) in the Y direction (a direction orthogonal to the paper surface of FIG. 1), and an X stage that moves on the Y stage in the X direction (the horizontal direction of the paper surface of FIG. 1). Stage and wafer holder 3
0 and an optical axis AX direction (Z
(In the axial direction), it is possible to finely move the wafer stage WST in FIG.
It is illustrated as

A movable mirror 34 is provided on the upper surface of one end of the wafer stage WST, and the laser interferometer 20 is arranged to face the movable mirror 34. Laser interferometer 20 projects the laser beam onto movable mirror 34 and receives the reflected light to measure the position of wafer stage WST with a resolution of, for example, 0.01 μm. Incidentally, the movable mirror is actually provided with an X movable mirror having a reflective surface orthogonal to the X axis and a Y movable mirror having a reflective surface orthogonal to the Y axis. An X laser interferometer for measuring the direction position and a Y laser interferometer for measuring the Y direction position are provided. In FIG. 1, these are typically shown as a movable mirror 34 and a laser interferometer 20. I have. Therefore, XY coordinates of wafer stage WST are measured by laser interferometer 20.

The measured value of the laser interferometer 20 is stored in the main controller 2
The main controller 24 positions the wafer stage WST, that is, the wafer W via the drive system 22 while monitoring the measurement value of the laser interferometer 20.

A reference plate FP is fixed on wafer stage WST so that its surface is at the same height as the surface of wafer W. On the surface of the reference plate FP, various reference marks (including reference marks 15a and 15b described later) including a reference mark used for baseline measurement and the like described later are formed.

The alignment microscope 18 is an off-axis type alignment sensor fixed to the side of the projection optical system PL. Here, the reason for using the off-axis type alignment sensor is that the projection optical system PL is usually designed and manufactured in consideration of the transmittance of the light of the exposure wavelength and the like, so that light of another wavelength is used as the alignment light. When this is used, while a problem such as chromatic aberration occurs, it is desirable to use another wavelength that does not expose the resist as alignment light, because it has an adverse effect due to absorption of the resist and does not affect the process. Therefore, when an alignment microscope using a different wavelength of the TTL system is used, there are some restrictions on the design of the projection optical system PL. However, as shown in FIG. 1, when another optical system is provided in addition to the projection optical system PL. Does not have such a restriction.

In this embodiment, an image processing system is used as the alignment microscope 18.
The alignment microscope 18 includes a light source that emits broadband illumination light such as a halogen lamp, an objective lens, an index plate, and an image sensor such as a CCD (all not shown). The illumination light emitted from the light source constituting the alignment microscope 18
After passing through the internal objective lens, the wafer W (or the reference plate F
P). The reflected light from the wafer mark area (not shown) on the surface of the wafer W returns to the inside of the alignment microscope 18, sequentially passes through the objective lens and the index plate, and the image of the wafer mark on an imaging surface such as a CCD, and on the index plate. An image of the index is formed.

The photoelectric conversion signals of these images are processed by the signal processing circuit 16, and the main controller 24 calculates the relative position between the wafer mark and the index.

By the way, as is apparent from FIG. 1, in this embodiment, the position of the wafer stage WST when projecting the pattern of the reticle R and the position of the wafer stage WST when measuring the mark by the alignment microscope 18 are different. It is necessary to accurately measure and correct the positional relationship (baseline = BL). For this reason, the baseline measurement is performed prior to the detection of the wafer mark. Here, this procedure will be described.

First, the reticle R is aligned before the baseline measurement (reticle alignment).
I do. In the case of the present embodiment, this reticle alignment means that the reticle stage 14 is driven and the center R of a reticle pattern image (hereinafter, appropriately referred to as a “reticle image”).
This means that CT (see FIG. 3) is accurately aligned with the optical axis AX of the projection optical system PL.

This reticle alignment is performed by main controller 2
4 is performed as follows. When performing continuous exposure processing on a large number of wafers, this reticle alignment may be performed only at the beginning of the multiple exposures.

The alignment marks 15a and 15b (see FIG. 1) on the reference plate FP are positioned at predetermined alignment positions (the method for obtaining the alignment will be described later) which are measured in advance when adjusting the apparatus.
The main controller 24 monitors the measurement value of the laser interferometer 20 based on the data in the memory (not shown) while the wafer stage W
Move ST.

Next, main controller 24 monitors alignment signals 15a, 15b and reticle marks 19a, 19a while monitoring output signals of reticle microscopes 29A, 29B.
The position of the reticle stage 14 is adjusted via a drive system (not shown) so that 9b coincides. In this adjustment, the reticle microscopes 29A and 29B detect the relative positions of the reticle alignment marks 19a and 19b and the alignment marks (reference marks) 15a and 15b via the projection optical system PL using a detection wavelength equal to the exposure wavelength. . Since the detection wavelength equal to the exposure wavelength is used for this detection, the relationship between the projection position of the reticle alignment mark and the projection position of another pattern (circuit pattern) on the reticle R formed at the time of exposure is due to chromatic aberration and the like. The reticle pattern is maintained as designed without being affected.

By the above adjustment, reticle image center RCT
Coincides with the optical axis AX. FIG. 1 shows a state in which the reticle alignment has been completed.

Here, a method of obtaining the above-mentioned predetermined alignment position at the time of adjusting the apparatus will be described in detail.

First, main controller 24 moves wafer stage WST such that reference plate FP is located immediately below projection optical system PL. In this state, the reticle microscope 2
While simultaneously observing the reticle marks 19a and 19b and the marks 15a and 15b on the reference plate FP by 9A and 29B, an appropriate command is given to the magnification correction controller 48 to greatly change the projection magnification of the projection optical system PL. The marks 15a and 15b on the reference plate FP because the projection magnification changes
Appears to move with respect to reticle marks 19a and 19b. At this time, the drive system 2 is controlled based on the output of a not-shown optical axis AX direction position sensor (focus detection system) so that the focus position is adjusted.
The reference plate FP is made to follow the focal plane through the Z stage 2 and a Z stage (not shown).

When the optical axis AX is deviated from the reticle center (the center of the pattern area of the reticle R), the amount of change differs between the marks 15a and 15b. That is, the positions of the centers of the marks 15a and 15b are changed and observed. This is because the optical axis AX does not coincide with the center of the reference plate FP.
This means that the reticle center of the reticle R aligned with reference to the reticle R and the reticle R does not coincide with the optical axis AX.

In this case, the amount of change in the center of the reference plate FP can be expressed by {(distance from the optical axis AX) × (change in magnification)}, and the amount of change and the change in magnification are determined by the reticle microscope 29A. 29B and the command value to the magnification correction controller 48, the distance between the center of the reference plate FP and the optical axis AX can be obtained as a result. The distance from the axis AX can be calculated back). Since the distance from the optical axis AX is a vector, the center of the reference plate FP is made coincident with the optical axis AX by driving the wafer stage WST in the X and Y two-dimensional directions according to the X axis component and the Y axis component of this vector. be able to.

Subsequently, the reticle stage 14 is driven so that the reticle marks coincide with the marks 15a and 15b on the reference plate FP at the new position, and the reticle R is again positioned. Further, the confirmation is performed by changing the projection magnification of the projection optical system PL. If there is a deviation, the same operation is repeated again, and the drive is performed so that the center of the reference plate FP coincides with the optical axis AX.

The alignment between the optical axis AX and the reticle R can be performed by exposing a test pattern on the wafer W without necessarily using the microscopes 29A and 29B and the reference plate FP. For example, a test pattern is exposed at a normal magnification, and a test pattern is superposed and exposed at a different magnification on the test pattern so that the amount of deviation is the same at each point (point where the image height is equal in design) in the wafer. To
A method of adjusting the position of the reticle R is conceivable.

The projection magnification in each of the above cases can be adjusted more precisely by making the projection magnification as large as possible by the magnification correction means (for example, using all corrections for changes in atmospheric pressure, etc.). It can be adjusted with no problematic precision.

As described above, the laser interferometer 2 at the position where the positioning of the reference plate FP with respect to the optical axis AX is completed.
The measured coordinate value of 0 is the predetermined alignment position described above. Therefore, the coordinate value of the predetermined alignment position is stored in the memory.

Next, a baseline measurement method will be briefly described. The reference plate FP is connected to the wafer stage WS as described above.
It is fixed on T and held so as to be substantially the same surface as the wafer W. At the time of measurement, the main controller 24 causes the reference plate FP to be substantially immediately below the optical axis AX of the projection optical system PL and the optical axis AXX of the alignment microscope 18, that is, the reticle alignment marks 19a, 1 of the reticle R.
The wafer stage WST is positioned via the drive system 22 so that the image of 9b can be observed by the reticle microscopes 29A and 29B at the same time as the marks 15a and 15b on the reference plate FP. Since this condition has already been satisfied, no special operation is required).

Here, the main controller 24 operates the marks 15a, 15a on the reference plate FP whose positional relationship is accurately known in advance.
The wafer stage WST is positioned (both shift and rotation) such that the images of the reticle alignment marks 19a and 19b of the reticle R come to the distribution positions with respect to the position b.
I do. Thereby, the marks 15a, 15 on the reference plate FP are
This means that the center position of b corresponds to the reticle image center RCT.

In this case, the mark 15c on the reference plate FP is accurately formed so as to be located at the design position of the alignment microscope 18, and the mark 15c on the reference plate FP is determined based on the output of the alignment microscope 18. By observing, the amount of deviation from the design position of the index for alignment inside the alignment microscope 18 can be accurately measured.
Thus, the positional relationship (baseline = BL) between the reticle image center RCT and the alignment microscope 18 can be accurately obtained.

In the projection exposure apparatus 10 of the present embodiment,
After the above baseline measurement sequence, the overlay exposure on the wafer W is started. That is, main controller 2
4 moves the wafer stage WST as shown in FIG. 2 and detects the position of a wafer alignment mark (not shown) on the wafer W via the alignment microscope 18.
Then, main controller 24 recognizes the sum of the relative positional relationship between the wafer alignment mark and the center of the index mark in wafer microscope 18 at this time and the output value of laser interferometer 20 as the mark position. Subsequently, the main controller 2
In step 4, the wafer W (ie, the wafer stage WST) is moved from the mark position by the sum of the baseline amount and the design coordinates of the wafer alignment mark based on the measurement value of the laser interferometer 20. As a result, the projected image of the pattern on the reticle R and the existing pattern on the wafer W are accurately aligned. In this state, exposure is performed, and the pattern on the reticle R is projected and transferred onto the wafer W.

In this manner, by repeatedly performing the exposure (projection transfer) while sequentially moving each shot area on the wafer W to the projection position of the image of the reticle pattern, the exposure of the step-and-repeat method is performed. Will be

In this embodiment, the above-mentioned step and
In the repeat type exposure operation, each alignment mark in a plurality of shots (sample shots) is detected by the alignment microscope 18 prior to exposure, and the detected values are statistically processed by the least square method or the like to arrange the shot area arrangement. It is assumed that the exposure is performed by a so-called EGA (Enhanced Global Alignment) method in which all shots are exposed based on the determined arrangement. According to the EGA method, a deviation from a design value of each shot area (for example, each chip when one chip is taken for one shot) is obtained by being decomposed into a shift component, a rotation component, a magnification component, and the like. Accordingly, the driving elements 44a, 44b, and 44c are controlled via the magnification correction controller 48 in accordance with the obtained magnification component, and the magnification of the projection optical system PL is adjusted in accordance with the expansion and contraction of the wafer W generated by a high-temperature process or the like. Can be adjusted, but in the present embodiment, the reticle image center RCT and the optical axis A of the projection optical system PL are adjusted by the reticle alignment described above.
Since X is adjusted so as to coincide with each other, alignment correction does not occur due to this magnification correction, and high-accuracy overlay exposure can be realized.

At the time of EGA measurement, if, for example, the number of error parameters is increased and a more precise calculation is performed such that the rotation of each shot, the magnification of each shot, and the like are obtained in addition to the arrangement coordinates of the shots. For example, a difference in change in magnification between the outer peripheral portion and the central portion of the wafer W can be obtained with respect to the distortion of the process, and finer magnification correction can be performed.

Although the productivity is not good, it is conceivable that instead of the EGA method, the magnification of the shot area is obtained and corrected for each shot area. In this case, the TTL alignment sensor may be used together with the above-described off-axis alignment microscope 18 to more accurately detect the change in the magnification of the shot. As the TTL type alignment sensor used in this case, for example, as shown in FIG. 2, an alignment beam, for example, a slit-shaped laser beam is horizontally incident from an opening on the side surface of the reticle holder 24b, and a mirror M The alignment beam is bent vertically downward to irradiate an alignment mark formed of a lattice mark on the wafer W through a peripheral portion of the projection optical system PL, and scans the alignment mark, and diffracts the light in the alignment mark area. LSA (Laser Step Ali) for detecting the position of the alignment mark by receiving the scattered light (reflected light) via the projection optical system PL and the mirror M
gnment) type alignment sensor 50 can be used. Of course, the off-axis type alignment microscope 18 may be constituted by an LSA (Laser Step Alignment) type alignment sensor.

In this embodiment as well, the magnification correction is, of course,
It is desirable to perform this also for a change in magnification characteristic of the projection optical system PL itself (according to a change in atmospheric pressure, etc.).

As described above, according to the present embodiment, the alignment mark (15) as a reference mark for positioning the reticle R at the time of device adjustment is set in advance.
a, 15b), so that the center of the reticle image RCT and the center of the optical axis AX of the projection optical system PL coincide with each other.
The projection magnification of the projection optical system PL is largely changed to adjust the coordinates of the reticle image center RCT so as not to change. When an actual apparatus is used, the information at the time of this adjustment (the position adjustment of the alignment marks (15a, 15b)) is performed. Wafer stage WST based on the wafer stage WST coordinate position information).
Is positioned at that position, and reticle alignment is performed with reference to the alignment marks (15a, 15b) at this position. For this reason, even if the projection magnification of the projection optical system PL is changed for the purpose of correcting the expansion and contraction of the wafer W during exposure, the center position of the pattern image of the reticle R does not change. Does not occur.

In this embodiment, the marks 15a and 15b on the reference plate FP are positioned at predetermined positions for positioning the reticle image center RCT and the optical axis AX at the time of adjusting the apparatus. The position coordinates of wafer stage WST at that time are measured, and thereafter, reticle alignment is performed based on this position, so that reticle alignment can be performed in a short time and throughput can be improved. In the reticle alignment, the reticle alignment marks 19a and 19b of the reticle R are aligned with the marks 15a and 15b on the reference plate FP at the predetermined alignment positions. Considering the state where the reticle alignment has been completed, at this time, the marks 15a, 15
The wafer stage WST is positioned such that the images of the reticle alignment marks 19a and 19b of the reticle R come to the distribution positions with respect to b, and the center positions of the marks 15a and 15b on the reference plate FP and the reticle image center RCT
Matches. Therefore, if the reticle alignment method of the present embodiment is performed, the reference plate FP is output based on the output of the alignment microscope 18 following the reticle alignment.
The baseline measurement can be performed only by observing the upper mark 15c, and there is an advantage that the procedure of the baseline measurement is simplified and the required time is shortened. In this embodiment, the reticle microscope 29A,
A case has been described in which a large reference plate FP is used so that the measurement of the reference marks 15c by the alignment microscope 18 can be performed simultaneously with the measurement of the reference marks 15a and 15b by 29B. However, the present invention is not limited to this. The fiducial mark is aligned with the position of the projected image on wafer stage WST, and subsequently, the fiducial mark is moved below alignment microscope 18 and aligned with the detection reference of the alignment microscope to obtain a baseline. Even when a normal reference mark plate for performing measurement is used, the procedure of baseline measurement can be simplified.

In the above description of the embodiment, the reticle image is shifted at the time of magnification adjustment only by the deviation between the optical axis AX and the reticle R. However, actually, the reticle image may be shifted by the error of the magnification correction means. The image shifts. When the lens element 40 in FIG. 1 is not accurately driven in the optical axis AX direction and moves or tilts obliquely, the reticle pattern image shifts. Since this phenomenon and the shift of the image due to the reticle position cannot be measured separately, when the reticle alignment of the above embodiment is performed, the error between the two is corrected simultaneously. If the error of the magnification correcting means is reproduced every time as a constant value, both errors can be corrected without any problem. That is, it is possible to simultaneously correct the positional shift of the reticle image that occurs during magnification adjustment due to an error of the magnification correction unit.

In the description of the above embodiment, from the viewpoint of improving the throughput, the reference plate FP is set at a predetermined position for positioning the reticle image center RCT and the optical axis AX in advance when the apparatus is adjusted. Marks 15a, 15
b, the position coordinates of the wafer stage at this time are measured, and thereafter, the position coordinates are assumed to be invariable, and the case where the reticle alignment is performed based on this position has been described. The position coordinates may be periodically measured again. If the reticle alignment can be performed in a sufficiently short time, the reticle alignment may be performed every time the reticle R is replaced.
In this case, an index such as fiducial marks 15a and 15b is not required for reticle alignment.

In the above embodiment, the reticle alignment reference marks 15a and 15b are
Although the case where it is on T has been described, it is not always necessary, and for example, it may be arranged in a reticle microscope or directly below the reticle. However, since the index position is adjusted as in the above embodiment, the index position needs to be adjustable, and position stability is required.

Although the above embodiment has been described with reference to the case where the reticle is aligned based on the predetermined information on the reticle alignment reference marks 15a and 15b, the present invention is not limited to this. For example, when positioning by pressing against a positioning pin or the like based on the outer shape of the reticle, the reticle is positioned at a position where the center coordinates of the reticle image do not change regardless of the change in the projection magnification, and the pin is pressed against this reticle. It is sufficient to store the position of the pin, and then adjust the position of the pin to this position. In short, the reticle is aligned to the same position based on information obtained from the adjusting means for adjusting the position of the reticle. Just do it.

Further, since the baseline and the reticle position slightly change due to a change in the temperature of the apparatus or the like, for example, at the beginning of the lot and at the beginning of the lot, measurement and correction are appropriately performed to prevent the occurrence of an overlay error. It is desirable to do so.

In the above embodiment, the case where the mask alignment method according to the present invention is applied to a batch exposure type projection exposure apparatus has been described, but the scope of the present invention is limited to this. Instead, the present invention is also applicable to a step-and-scan type exposure apparatus that relatively scans the reticle R and the wafer W.

[0071]

As described above, claims 1 to 4
According to the invention described in (1), there is provided a method of aligning a mask in which a positional shift of a projected image of a mask pattern does not occur by adjusting a projection magnification of a projection optical system.

According to the fifth aspect of the present invention, even when the overlay exposure is performed while adjusting the projection magnification of the projection optical system, the overlay error caused by the shift of the image of the mask pattern does not occur. A projection exposure method is provided.

[Brief description of the drawings]

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

FIG. 2 is a view showing a state during exposure of the apparatus of FIG. 1;

FIG. 3 is a diagram for explaining a reticle positioning method and a problem to be solved by the invention in the apparatus of FIG. 1;

[Explanation of symbols]

 15a, 15b Reference mark 19a, 19b Reticle mark 24 Main controller 29A, 29B Reticle microscope 48 Magnification correction controller R Reticle W Wafer PL Projection optical system FP Reference plate

Claims (5)

[Claims] 1. A method of positioning a mask prior to transferring an image of a pattern formed on the mask onto a photosensitive substrate via a projection optical system, the method comprising: positioning the pattern on the mask. And a first step of measuring the projection positions of at least two alignment marks formed in a predetermined positional relationship; and changing the projection positions of the at least two alignment marks after changing the projection magnification of the projection optical system. A second step of measuring; and a position of the mask such that the center of the projected image of the pattern does not change even when the projection magnification of the projection optical system is changed based on the measurement results of the first and second steps. And a third step of adjusting the mask position. 2. The method according to claim 1, wherein after replacing the mask, another mask after the replacement is aligned to the same position using information relating to the position of the mask in the third step. Mask alignment method. 3. The method according to claim 2, wherein the information related to the position of the mask is position information of a reference mark corresponding to the mask position in the third step. 4. The position of the mask according to claim 2, wherein the information related to the position of the mask is information obtained from a mask position adjusting means at the time of positioning the mask in the third step. Matching method. 5. A projection exposure method for transferring an image of a pattern formed on a mask onto a photosensitive substrate via a projection optical system, wherein prior to positioning the mask and the substrate,
A projection exposure method comprising a step of aligning a mask and a projection optical system by the mask alignment method according to claim 1.