JP2707541B2 - Photosensitive substrate alignment method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention positions a wafer on a two-dimensional moving stage.
Alignment method, especially when placed on the stage
Position and orientation of the flat wafer
Pre-alignment method for aligning
You. [Prior Art] In recent years, with the miniaturization of patterns such as LSI and VLSI,
Exposure equipment suitable for manufacturing
Is often used. This reduction projection type exposure apparatus
Putter on mask original plate (hereinafter referred to as “reticle”)
Image is projected onto the wafer using a projection lens.
Therefore, in the exposure apparatus, a transfer that moves two-dimensionally is performed.
Moving stage (hereinafter simply referred to as “stage”) is fixed
Repeated reduction projection while stepping
Many patterns on the entire surface of the wafer
What is called step-and-repeat, which transfers in a box shape
G method is used. In this case, one LSI etc. is manufactured
To do this, several or more layers of patterns are sequentially superimposed on the wafer.
Since it is formed by the
Keep the alignment error (position shift including rotation) below a certain value.
If so, the conductive or insulating state between the layers is not as intended.
Therefore, the LSI cannot perform its function. For example, 1 μm
For circuits with a minimum line width of
Only a misalignment is allowed. For this reason, this type of exposure equipment has a transfer type
The formed alignment mark is detected photoelectrically and the wafer
So-called fine ala that aligns the image with the projected image
Incorporated with the optical device. However, this fine alignment optics
The projection range of the spot light that irradiates the illumination mark is extremely large
Alignment near the projection optical axis of the spot light
When there are no print marks, move the wafer
You have to search the area,
Searching takes a lot of time. Shorten this search time
To do this, when the wafer is placed on the stage,
Jeha's alignment mark is fine alignment optics
Must be at least within the field of view (detection area) of the device
No. For this reason, the prior
Position adjustment (rotation
Including direction. ) And installed on stage
Again, the wafer is cut straight again on the stage side
(Hereinafter referred to as "flat").
Positioning device for performing the alignment
It is disclosed by Japanese Patent Publication No. 4 and already known. Problems to be Solved by the Invention However, the above-described conventionally known positioning device has a
Protruding from the wafer mounting surface of the wafer holder on the page,
Multiple wafers that can contact the flat and outer peripheral surface of the wafer
Wafer mounted on the wafer holder
Is positioned with respect to the stage by the rollers
Is configured to be (including the direction of the flat)
You. Therefore, the rollers that contact the wafer and the rollers
Like the roller moving means for moving,
The stage must be equipped with a dedicated device
Becomes complicated. In addition, wafer loading and unloading
When moving the wafer, avoid the protruding roller
And no special consideration is given to the movement of the transport means.
Must be done. In addition, mechanical correction by rollers
Instead, a number of photoelectric detectors are provided on the stage,
To detect the position of the wafer and the direction of the flat
Although it is thought that there is a lot of dedicated around the wafer holder
Photodetectors must be provided, and the
Has the disadvantage of becoming more complicated. The present invention uses special positioning on the stage side as described above.
Very accurate wafer alignment without the need for additional equipment
New pre-alignment method that can perform alignment
The purpose is to provide. [MEANS FOR SOLVING THE PROBLEMS] In order to achieve the above object, the present invention provides an
Explaining in connection with the drawings representing the examples,
Projection optical system on photosensitive substrate (7) having notch (F)
Projecting a pattern image of the reticle (5) via (6)
The stage (8) that moves two-dimensionally before
Direction of the notch and the notch direction of the notch (F)
Place photosensitive substrate (7) on stage (8)
Placing (101), and the optical axis of the projection optical system (6).
An electrode having a predetermined positional relationship and formed on the photosensitive substrate (7).
Finance that can detect alignment marks (SYn, SXn)
Luminous flux from the illumination device (21-31 or 51-64A, B)
Optical information generated at different points of the notch (F) using
Detecting (102-104)
The notch (F) in the predetermined moving direction
A step (102-104) for determining the amount of rotational deviation;
When the deviation is over the allowable range, the rotational deviation is offset.
(10) aligning the photosensitive substrate (7) so that
2 to 104) and a fine alignment device (21 to 31 or
51-64A, B), the photosensitive placed on the stage (8)
The alignment mark (SYn, SX) formed on the substrate (7)
n) detecting (205, 206).
It is a technical point. Further, as a fine alignment device (21-31),
Through the shadow optical system (6), the notch (F) and the alignment
Use the type that detects the mark (SYn, SXn).
Is desirable. In the above steps, the photosensitive substrate (7) is placed on the stage.
In step (101) to be placed on (8), the stage
(8) A guide or roller like the conventional device
Since no protruding member is provided, mounting of the wafer (7)
Is extremely easy. The wafer (7) placed on the stage (8) is
For realignment, for example, a projection lens (6)
Fine alignment device (21 to 31)
Or 51-64A, B)
The rotational deviation angle of (F) is obtained. The position of this notch (F)
In the position detection, the upper surface of the wafer (7) and the stage (W)
(Including the wafer holder 9).
Notches are clearly notched because there is a step that is as large as the thickness of
(F) can be detected. Moreover, the notch
Non-contact position of (F) and fine alignment
It is detected by the device (21-31 or 51-64A, B)
With high accuracy without damaging the notch (F)
Position detection becomes possible. Therefore, this fine alignment
(21-31)
Output can also be obtained with high accuracy. By performing the above pre-alignment method, the second layer
During fine alignment for subsequent exposure,
Easy and quick detection of upper alignment marks (SYn, SZn)
It is possible to perform it quickly. And for the subsequent fine alignment
The same fine alignment device (21 to 31)
Or 51-64A, B) to detect the alignment mark
Easy alignment and fine alignment in a short time
I can. Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
I will explain. FIG. 1 shows a reduction projection type exposure apparatus suitable for carrying out the present invention.
Slope view showing the state of performing pre-alignment using
FIG. 2 illustrates the configuration of the reduction projection type exposure apparatus of FIG.
FIG. In FIG. 2, illumination light from an exposure light source 1 is a condensing lens.
Through shutter 2, shutter 3, and condenser lens 4.
(A mask having an enlarged circuit pattern) 5
I will tell. The illuminated reticle 5 pattern image is projected
Is reduced to 1/5 to 1/10 by the
Projected to The wafer 7 has a movable stage in the x and y directions.
The wafer housing provided on the page 8 and capable of rotating and moving up and down.
It is vacuum-adsorbed on the rudder 9. Wafer holder 9
Is a rotation provided on the stage 8 as shown in FIG.
It is rotationally driven by the correction drive motor 10. Also,
The movement of the stage 8 in the x direction is performed in the x direction driving mode shown in FIG.
The movement in the y direction is performed by driving the
This is performed by driving the directional drive motor 12. In addition,
As shown in FIG. 1, two orthogonal sides of the stage 8
The mirror 13 has a surface extending in the y direction, and the reflection surface has a surface extending in the x direction.
Mirrors 14 are fixedly provided. Laser light wave
Interferometer (hereinafter simply referred to as "laser interferometer")
You. 15) projects laser light on mirror 14 and reflects it
Upon receiving the light, the position (or movement) of the stage 8 in the x direction
The laser interferometer 16 detects the laser
-The light is projected, the reflected light is received, and the
Configured to detect the position (or the amount of movement)
I have. The flat F is substantially parallel to the moving direction x of the stage 8
Of the wafer 7 placed on the wafer holder 9 as shown in FIG.
The reticle 5 pattern and alignment mark
It is baked through the projection lens 6 as shown in FIG.
For example, the first layer circuit pattern (pattern exposure area
Are represented by squares C1, C2... Cn. ) Is arranged in a matrix
Formed in rows, with diffraction patterns around the outside of each pattern exposure area
Multiple alignment marks SY of the child pattern (see Fig. 4)
n, SXn and GYn, Gθn are formed. Each pattern dew
Alignment extending in x and y directions from center Pn of optical region Cn
The mark marks SYn and SXn are connected to an exposure position detecting device (21
~ 31,41) and sandwich the alignment mark SX1
The alignment marks GYn and Gθn provided by
Detected by global alignment detection system (51 to 64A, 64B)
Is done. The stage output from the laser interferometers 15 and 16
8 is sent to the main control circuit 17, and the drive mode
The motors 11 and 12 respond to control signals from the main controller 20.
Is controlled. The stay including the wafer holder 9
The wafer 7 placed on the wafer 8 is transported by a transfer arm 18 (see FIG. 1).
From the pre-alignment device (not shown)
Conveyed. Next, the wafer provided in the reduced projection type exposure apparatus
Projection lens for alignment of 7
Detection of alignment marks SYn and SXn on wafer 7
TTL type exposure position detector (hereinafter referred to as “Step
This is referred to as a "liment detection system." ) And the projection lens 6
Through a pair of objective lenses 61A and 61B provided in a row,
Off-axis method to detect GYn, Gθn
Fine Aligners for Global Alignment Detection System
The configuration of the moment detection system will be described in detail. Web alignment is required for the TTL step alignment detection system.
C. Light having a wavelength that does not expose the photosensitive layer on C7 is used.
The laser beam from the laser light source 21 shown in FIG.
As shown in 2, beam expander 22, cylindrical
After passing through the cull lens 23, the beam splitter 24
Is divided into two light beams orthogonal to each other. Divided one
The other beam is a reflecting mirror 25A, a semi-transmitting mirror 26A, and a condenser lens 27.
A, through the reflecting mirror 28A, at the position of the field stop 29A in the x direction
Connect long, elongated elliptical light spots. In addition, the field stop
The laser beam that has passed through 29A is reflected by reflector 30A.
Through the pupil 6P of the projection lens 6 (see FIG. 2)
Figure as an elliptical light spot LA extending upward in the x direction
The image is re-imaged as shown in FIG. Light spots from wafer 7
Light reflected by LA (including specularly reflected light, irregularly reflected light, and diffracted light)
Is incident on the projection lens 6 in reverse, and the reflecting mirror 30A, the field stop 2
9A, passing through the reflecting mirror 28A and the converging lens 27A to the semi-transmitting mirror 26A
And reflected by the photoelectric detector 31 described later in detail.
Shoot. Note that the reflecting mirrors 25A to 30A and the photoelectric detecting unit 31 are
Step alignment in the y direction by TTL method
Configure the detection system. Also, the other split by the beam splitter 24
As shown in FIG. 1, the laser beam of
Field of view through transmission mirror 26B, condenser lens 27B, reflection mirror 28B
An elliptical light spot that extends long in the y-direction is
Laser beam that passed through the field stop 29B
Is the pupil 6P of the projection lens 6 after being reflected by the reflecting mirror 30B.
(See FIG. 2)
It is projected on the same surface as the wafer 7 as a pot LB. This
Light from the wafer 7 due to the light spot LB (specular reflection)
Light, scattered light, diffracted light, etc.)
30B, field stop 29B, reflector 28B, condenser lens
After passing through the mirror 27B and reflected by the semi-transmissive mirror 26B, photoelectric detection
The light enters the photoelectric detection unit 41 configured in the same manner as the subtraction. What
Note that, with the reflecting mirrors 25B to 30B and the photoelectric detector 41, the TT
Step alignment detection system in the x direction using the L method
Be composed. The detection signals output from both photoelectric detectors 31 and 41 are mainly
Sent to the controller 17 and the position from the laser interferometers 15 and 16
In cooperation with the signal, the alignment marks SYn and SXn
It is used for alignment of the exposure position on the wafer 7. Next, a TTL step alignment detection system
Off-axis global alignment
The detection system will be described. Off-axis global alignment detection
The laser light source 51 used in the system
Those which oscillate light of a wavelength to which the agent is not exposed are used.
The laser beam from the laser light source 51 is
Pander 52, cylindrical lens 53, galvanometer mirror
54, through the condenser lens 55, at the position of the field stop 56 as shown in FIG.
Image an elliptical light spot that extends up and down as shown
You. This light spot is applied to the rotational vibration of the galvanometer mirror 54.
Therefore, a simple vibration occurs in a direction (y direction) perpendicular to the paper surface. The laser beam focused at the position of the field stop 56 is
It becomes a parallel light beam by 57, and by the beam splitter 58
One of the split light beams is split into two light beams,
The light enters the first objective lens 61A via the semi-transmissive mirror 60A,
The objective lens 61A has an elliptical light spot LY extending in the x direction.
Is formed on the wafer 7. Also, the beam splitter 58
The other light beam split by the second
The light enters the objective lens 61B, and the second objective lens 61B also moves in the x direction.
An extended elliptical light spot Lθ is formed on the wafer 7
You. Both of the light spots LY and Lθ are
Simple vibration in the y direction perpendicular to the paper in FIG. 2 according to the rotational vibration
I do. Light reflected from wafer 7 by light spots LY, Lθ
(Specular reflection light, scattered light, diffracted light, etc.)
After passing through 60A, 60B, it reaches the spatial filters 62A, 62B,
The required scattered light and diffracted light are selected,
Light is condensed on the light receiving surfaces of the photoelectric detectors 64A and 64B via 63A and 63B.
You. The photoelectric detectors 64A and 64B are used for a wafer described later.
7 (eg, GY7, Gθ10) the amount of diffracted light from the mark
And outputs a photoelectric signal corresponding to. The semi-transmissive mirror 60A,
1 objective lens 61A, spatial filter 62A, condenser lens 63A
And Y-alignment with photoelectric detector 64A
A semi-transmission mirror 60B, a second objective lens 6
1B, spatial filter 62B, condenser lens 63B and photoelectric detection
Global alignment detection system with detector 64B
Is done. The detection signals of the photoelectric detectors 64A and 64B are both
It is sent to the main controller 17 (see FIG. 1). Figure 5 shows the above TTL step alignment detection
Light spots LA and LB, and off-axis method
Formed by the global alignment detection system
FIG. 4 is a plan view showing a positional relationship between light spots LY and Lθ.
You. In FIG. 5, a coordinate xy having the optical axis AX as an origin is determined.
, And the x-axis and the y-axis respectively indicate the moving directions of the stage 8.
I forgot. The area of the circular broken line centered on the optical axis AX is the projection lens.
Size 6 image field if
The area indicated by the broken line is the projected image of the effective pattern area of the reticle 5.
Indicates Pr. Light spots LA and LB are image fields i
is located outside the projected image Pr in f, LA coincides with the x-axis, and LB
Is formed to coincide with the y-axis. On the other hand, the center of vibration of the light spots LY and Lθ is
Coincides with a line 11 parallel to the x-axis, separated by a distance Y0 in the direction
The distance Dx in the x direction to a value smaller than the diameter of the wafer 7
It is determined to be. In addition, both light spots LY, Lθ
Are arranged symmetrically about the y axis,
The control device 17 (see FIG. 1) controls the optical axis with respect to the projection point of the optical axis AX.
Information on the positions of the pots LY and Lθ is stored. Further
In addition, main controller 20 controls light spot for the projection point of optical axis AX.
The center position (distance X1) in the x direction of the light spot LA and the light spot LB
Information about the center position (distance Y1) in the y direction is also stored.
You. Therefore, these light spots LY, Lθ and LA, LB
Scans the alignment mark on the wafer 7 and
When the position is detected, the position of the projection optical axis Ax on the wafer (7) is determined.
You can ask. Figure 6 shows the TTL step alignment detection system.
FIG. 3 is a specific configuration diagram illustrating an example of photoelectric detection units 31 and 41. Both
Since the photoelectric detectors 31 and 41 have the same configuration,
Here, only one photoelectric detector 31 is shown. 26A transflective mirror
The reflected light from the wafer 7 is transmitted to the beam splitter 70.
Is divided into two, and one of the divided lights is a projection lens.
Spatial filter 71 placed at a position conjugate with the pupil of
Incident, and only the scattered light D0 is opened by the spatial filter 71.
The light is extracted at the opening 71a (71b),
The light is focused on the light receiving surface of the element 73. With beam splitter 70
The other split light is located at a position conjugate with the pupil 6P of the projection lens 6.
Reaches the spatial filter 74, where it is diffracted
Only the light ± D is extracted through the apertures 74a and 74b and its diffraction
The light ± D is reflected by the two reflecting surfaces 75a and 75b, respectively,
The diffracted light + D is transmitted to the photoelectric element 77a through the condenser lens 76a.
The light is focused on the light receiving surface. The other diffracted light -D is condensed.
The light is focused on the light receiving surface of the photoelectric element 77b through the lens 76b.
You. 7 and 8 show the spatial filters 71 and 74, respectively.
It is a top view which shows a shape. In FIG. 7, a spatial filter
-71 is perpendicular to the light spot LA on the wafer 7
Specularly reflected light LR (shown by a broken line) that extends long in the direction is blocked.
Scattered light outside the long direction of the specular reflected light LR
Openings 71a and 71b for transmitting D0 are provided. This opening
The openings 71a and 71b are arranged symmetrically with respect to the regular reflection light LR.
You. In the present embodiment, this spatial filter 71 allows
Light spot LA on C7 or projected on the edge of flat F
Extraction of irregularly reflected light that occurs when the light spot LB
Disturbance generated when projected on the left and right outer peripheral edges of the wafer 7
Extraction of the reflected light is performed. Therefore, from the photoelectric element 73, the end of the flat F of the wafer 7
Between the edge detection signal and the detection signals of the left and right outer peripheral edges of the wafer 7.
Both detection signals SC are sent to main controller 17. On the other hand, the spatial filter 64 shown in FIG.
The light LR is blocked, and the specular reflected light LR is sandwiched by the specular reflected light LR.
Openings 74a, 74 symmetrical in the direction perpendicular to the longitudinal direction of
b are provided, and the diffracted light ± D passes through the openings 74a and 74b.
It is configured as follows. The openings 74a and 74b are
LA and LB are projected onto the alignment marks SYn and SXn, respectively.
For example, the third order diffracted light generated when
The shape and size are determined so that Follow
From the photoelectric elements 77a and 77b to the alignment on the wafer 7.
Detection output when marks SYn and SXn are detected respectively
Signals SA and SB are sent to main controller 17. Next, each photoelectric signal SA, SB, SC from the photoelectric element 73, 77a, 77b
An example of a processing circuit for inputting the following will be described with reference to FIG. this
The processing circuit is provided by a microcomputer provided in the main controller 17.
Data and mini-computers in cooperation with
Of alignment marks SYn and SXn (see FIG. 3) on wafer 7.
Detection of position and position of wafer 7, especially position of flat F
And the rotation deviation angle with respect to the moving direction of the stage 8 are detected.
Put out. In FIG. 6, a photoelectric signal SA output from the photoelectric element 77a is shown.
Is a preamplifier 80 having a predetermined amplification factor in FIG.
After amplification by a gain control amplifier (hereinafter “AGC”)
Called. ) Enter in 81. AGC81 is preset
Adjust the level of the input signal with the adjusted gain to output the signal SA1.
You. On the other hand, the photoelectric signal SB output from the photoelectric element 77b is
Similarly, the signal is converted to signal SB1 by preamplifier 82 and AGC83.
You. The adder circuit 84 converts the two signals SA1 and SB1 into analog signals.
The added signal SD is output, and this signal SD is analog-to-digital
Tal converter (hereinafter referred to as “ADC”) 85 is a sampler.
To a digital value corresponding to the size (level)
Convert. Random, access, memory circuit (hereinafter simply
Called "RAM". 86 is the data sampled by ADC85.
The digital value is stored in the designated address. AD
C85 sampling and RAM86 addressing are both
Performed by pulse signal SP of Zata interferometer 15 (or 16)
Will be This pulse signal SP is generated when the stage 8 receives x (or
1) each time the unit moves (eg, 0.2 μm) in the y) direction.
A pulse train that becomes a pulse.
In response, one sampling of ADC85 is performed, and the address of RAM86 is added.
Only one address is updated. On the other hand, the photoelectric signal output from the photoelectric element 73 in FIG.
The signal SC was amplified by the preamplifier 87 in FIG.
Later, the level is adjusted with the gain preset by AGC87.
Is adjusted. The signal SC1 whose level has been adjusted is sent to ADC89.
Therefore, after being sampled, it is written to RAM 90 in address order.
Remembered. ADC89 sampling and RAM90 address update
New also pulse signal from laser interferometer 15 (or 16)
Performed in response to SP pulses. Arithmetic processing of microcomputers and minicomputers
Management unit (hereinafter referred to as “CPU”) 100 is RAM 86 or 9
Digital information (waveform information) from 0 and laser interference
Y (or of stage 8) read by a total of 15 (or 16)
x) Input the signal information in the direction and enter the alignment mark SY.
n, SXn and the position of the wafer 7 with respect to the stage 8
Computation processing of the position of the cut F and the rotational deviation angle by software
Run more. Next, the alignment operation of the wafer 7 will be described with reference to FIG.
This will be described in detail based on a flowchart. A pre-alignment device (not shown) provided during the transfer of the wafer 7
The orientation and position of the flat F are regulated by the
The transferred wafer 7 is pre-aligned by the transfer arm 18.
Wafer from the stage 8 of the exposure apparatus
Placed (loaded) on the holder 9 (step
101). In this case, the stage 8 easily mounts the wafer 7.
From just below the projection lens 6 to a predetermined position in the y direction.
Will be sent out. However, the placed wafer 7 is not necessarily
The direction of the flat F is equal to one of the movement directions x of the stage 8.
They are not always placed exactly in a row and in a predetermined position. So
Here, the direction of the flat F is aligned with the moving direction x of the stage 8.
Using a step alignment detection system,
The light spot LA projected through the projection lens 6
Rat F orientation (Δθ alignment)
7 are performed together with the alignment in the y direction (step 10).
2). In this step 102, the wafer 7 is placed on the stage 8.
It is placed and sent directly below the projection lens 6, and
The flat F is illuminated by the light spot LA as shown in FIG.
The wafer 7 is moved in the y direction to a position where
The right end of the flat F, for example,
Scanned by LA. At this time, the upper surface of the wafer 7 is
Between the top of the stage 8 (including the wafer holder 9)
Is a light spot because there is a step corresponding to the thickness of the wafer 7.
LA goes out of the wafer from the edge (boundary) of flat F
In the focus state, at the edge of the flat F
There is a clear change in the emitted light information (specular reflection light, scattered light, etc.)
I will. The scattered light and specularly reflected light emitted from the flat F are
The light enters the projection lens 6 and reverses the outward path of the illuminating light.
After being reflected by the mirror 26A, it reaches the photoelectric detector 31. So
The scattered light from among the beams splits as shown in FIG.
After being reflected by the filter 70, it becomes a spatial filter 71 for scattered light.
Therefore, it is extracted. The extracted scattered light is
Reaching the light receiving surface, photoelectrically converted, and according to the amount of scattered light
A signal is output from the photoelectric element 73. In this case, its output
The signal is maximum when the light spot LA matches the flat F
Becomes The y direction of the stage 8 corresponding to the maximum output signal
Direction is sampled by the pulse signal of the laser interferometer 15.
The waveform stored in the RAM 90 (see FIG. 9)
100 is detected by arithmetic processing. CPU 100
The position in the y direction together with the position in the x direction of the stage 8
Remember. Next, as shown in FIG.
After moving a predetermined distance 1x in the direction, move again in the y direction.
The left end of the flat F by the light spot LA in the y direction
Scan. By this scanning, it is generated from the flat F
Scattered light is detected again by the photoelectric element 73, at which time
The position of the stage 8 in the y direction and the position in the x direction
Is stored. Left of flat F stored in CPU 100
The difference Δy between the end and the right end in the y direction and a predetermined distance in the x direction
From 1x, the CPU 100 determines the rotational deviation of the flat F with respect to the x direction.
Calculate the position angle Δθ and determine that the value of Δθ is greater than or equal to a predetermined error range.
The motor 10 by controlling the motor 10
The holder 9 is rotated. By this rotation, Flat F becomes
It is parallel to the moving direction x of the stage 8. In addition, flat
By the sign of the difference Δy between the positions of the F in the y direction,
The rotation direction of C7 is determined. At this time, the wafer 7
If the rotational drive center of the
The position of the center point of the flat F with respect to the pot LA in the y direction is CP
It can be easily predicted by the calculation of U100.
The alignment is also completed. Generally this
Sufficient, but the rotation direction of Flat F has been corrected
Later, the stage 8 is moved again in the y direction for confirmation,
Make the center position of the light spot LA coincide with the flat F
Of the stage 8 in the y direction (step 10
3). This step 103 is performed as necessary and can be omitted.
It doesn't matter. Next, as shown in FIG.
After moving the specified amount 1y in the y direction, move further in the x direction
Then, the circumferential edge of the wafer 7 is shown in FIG.
Scan in the x-direction as shown, the x of the edges 7a and 7b around it.
Direction position is the waveform stored in the laser interferometer 16 and RAM90
And the midpoint 7 of the distance dx between the end edges 7a and 7b.
Find c. A predetermined distance 1y in the y direction from this center point 7c
If it returns, the position of the midpoint Fm of the flat F in the x direction (ie,
The position of the wafer 7 in the x direction) is the position of the flat F in the y direction.
(Step 104). As described above, the position of the flat F (the stay of the wafer 7)
Is determined, then the wafer 7 is determined.
It is determined whether or not the first layer is exposed. This place
Read a barcode provided on the reticle 5, for example.
The reticle 5 is for the first layer due to the signal
If so, proceed to step 106, where reticle 5 is in the second layer.
Or if some are used for higher layers,
Proceed to step 201. When the wafer 7 is the first layer, the wafer 7 is flat
F is immediately moved by a predetermined amount in the xy direction with reference to FIG.
The center P1 of the middle exposure area C1 was brought on the projection optical axis AX
Later, immediately exposed, using known step and repeat
Pattern by matrix exposure as shown in FIG.
Are burned (step 106). In this case,
Based on the wafer position obtained in steps 102 to 104 of
Correct step and repeat stepping position
You may make it. In addition, during this baking,
SXn, SYn, GYn, Gθn
You. When the step and repeat exposure is completed,
The wafer 7 is moved from the wafer holder 9 on the stage 11 to the outside of the exposure apparatus.
(Step 107).
When the wafer is removed from the wafer holder 9, a predetermined
It is determined whether the number of wafers N has been completely unloaded, and
If there is a wafer for which the first layer is to be exposed,
And wafer loading in step 101 is performed.
The operation up to step 107 is repeated. In step 108
The predetermined number of wafers 7 unloaded after the exposure of the first layer is completed
When N is reached, the loop from step 101 to step 108
Chin, work stops. On the other hand, in step 105, the wafer 7 is at least
If the first layer is baked, go to step 101
Pre-alignment on stage 8 from
Glowing for more precise alignment
Global Alignment with Bal Alignment Detection System
Is performed. Therefore, first, the stage 8 moves in the y direction.
It is moved by a predetermined amount (step 201). Here, the off-axis global alignment
The alignment (position alignment) by the point detection system
explain. This off-axis alignment detection system
Japanese Patent Application Laid-Open
No. -13742
I will only outline the method. This off-axis alignment detection system
As shown in FIG.
And SYn, SXn, GYn, Gθn, for example, in the first layer
Has already been formed. This off-act
The objective lenses 61A and 61B of the cis type alignment detection system
Is substantially equal to the interval Dx (see FIG. 5), for example, FIG.
Among them, GY7 and Gθ10 are selected. In this case,
The IC pattern is located in the immediate vicinity of the GY7 and Gθ10
Tens of μm or less due to pre-alignment.
If you do not align to the accuracy in the
Noise due to scattered light is mixed into the detection signal or IC pattern
Will be misidentified as an alignment mark,
There is a risk of erroneous detection. However, the step above
The rotation of Flat F has already been reduced by the
The position of the flat F in the y direction and the wafer
Since the overall position is required, the alignment mark
Can be predicted within the required accuracy. Follow
In step 201, the target alignment mark
(Eg GY7, Gθ10) projected from objective lenses 61A, 61B
The stage 8 to a position close to the light spot LY, Lθ
Can be moved immediately in any direction. Here, although slightly, global alignment
From the objective lenses 61A and 61B of the detection system on the wafer 7 respectively
Line 11 connecting the projected light spot LY and Lθ (stage 8
Of the alignment mark GY
7, the line 12 connecting Gθ10 is shifted in the y direction as shown in FIG.
And a rotational deviation (angle θ). However
The rotational deviation is extremely small due to the previous step 102.
Are being suppressed. Of course there is a deviation in the X direction, but here
In the stage at, the light spots LY, Lθ and the alignment
GY7 and Gθ10 are both elongated in the x direction,
Is somewhat tolerated. In this case, the stage 8 is slightly moved in the y direction to
Make sure that GY7 matches the vibration center of light spot LY.
While applying servo to the drive motor 12 in the y direction,
C until the angle Gθ10 coincides with the vibration center of the light spot Lθ.
If the eha holder 9 is finely rotated, it becomes a matrix
The direction in which the arranged exposure areas C1, C2,.
It can be made to coincide with the movement direction x of the die 8. Therefore,
Each of the exposure areas C1, C2, C3 ...
Alignment marks GYn and Gθn are attached.
The two arcs provided at predetermined intervals as described above
Using only the alignment marks GY7 and Gθ10,
Θ alignment is performed, and at the same time,
(The detection of the stage position in the y direction) is also performed.
202). When this Y-θ alignment is completed,
Move the page 8 and again use the step alignment optics
Using the system light spot LB, for example, near the center of the wafer 7
Of the specific alignment mark SXn
Position information of the laser interferometer 16 by scanning in the x direction and RAM 8
6 based on the waveform stored in step 6 (step 20).
3). As a result, the glow of the wafer 7 in the x direction and the y direction
This means that the valve alignment has been completed. global
・ After alignment, some alignment sequences
Alignment is considered here, but here
The following describes the method for switching. And then the Global Alli
The wafer 7 is moved to the first exposure position C1 based on the result of the
It is moved by the stage 8 (step 204). This place
If global alignment is complete
TTL alignment detection system
Alignment marker attached to the first exposure area C1
SY1 and SX1 (see FIG. 3) are captured with sufficient accuracy. Here, the light of the step alignment detection system in the x direction
The alignment mark SX1 is scanned by the spot LB.
(Step 205), the position in the X direction is determined. Further
The light spot of the step alignment detection system in the Y direction
The alignment mark SY1 is scanned by the
Step 206), the position in the y direction is determined, and eventually the first exposure area C
Make the exposure optical axis AX (see Fig. 1) coincide with the center P1 of 1
Can be. As soon as this alignment is completed, the first exposure
Exposure (printing) of the light region C1 is performed. Next, the pre
It is determined whether or not the number of contacts has reached a predetermined number (n = 14 in FIG. 3).
Refused. If the predetermined number n is not reached, the next exposure
After stepping into the area (step 209), the next exposure
Return to step 205 for scanning the area alignment mark SX2
Alignment (similar) is performed and immediately
(Step 207), the number of prints reaches a predetermined number n.
Steps from step 205 to step 209
-The AND repeat exposure operation is repeated. In step 208, the number of prints reaches a predetermined number n.
Then, the process proceeds to step 107, where the wafer 7 is unloaded and
It is determined whether or not the number of wafers has reached a predetermined number N (S
Step 108). If the predetermined number N is not reached, a new
The wafer 7 is placed on the stage (step 101), and
Pre-alignment step (102-104) and fine
The alignment steps (201 to 206) are repeated.
When the number of wafers is equal to the predetermined number N, the above operation is stopped.
You. Further, in the above sequence, the step 203
After the global alignment is completed,
Identify the shot array with even higher precision and for each shot
Eliminates the need for alignment and improves throughput
For this reason, the extension disclosed in Japanese Patent Application Laid-Open No. 61-44429 (en
Hansement) Exposure by global alignment
You can do it. In that case, several shots on the wafer 7
The positions of the marks SXn and SYn attached to the
Detect and detect each shot on the wafer 7 based on statistical means.
And the exposure can be performed with minimum error.
, The design shot during step and repeat
Modify the position and based on this modified shot position
The stage 8 may be stepped. Therefore in FIG.
After step 203, the enhancement global
Mark position measurement and alignment of shots for alignment
A positive (arithmetic) is executed and then the modified shot
Step 106 is executed based on the sequence data.
Should be changed to In addition, the exposure position detecting device of the present invention includes an off-actuator.
You can also use the cis type global alignment system.
it can. As shown in FIG. 2, in this embodiment, two gross
There is only a global alignment system.
Off-axis alignment that detects alignment marks
If a system is provided, these off-states in the x and y directions
Axis type global alignment system, shown in Fig. 10.
A sequence similar to the alignment sequence shown
Can be run immediately. In the embodiment, a fine alignment system is provided.
Optical information from the edge of the wafer
However, the present invention is not limited to this.
No. Furthermore, the light of specular reflection light as light information from the wafer edge
A change in the amount may be detected. As described above, according to the present invention, a fine alignment apparatus is provided.
The position of the notch is detected using the light beam projected from the
Pre-alignment is performed.
Alignment with same alignment and pre-alignment
Pre-stage
Of course, there is no need to provide
It is possible to transport wafers easily and quickly.
Become. Furthermore, alignment with sub-micron accuracy is essential.
Uses a fine alignment system that has the ability to perform
To perform pre-alignment.
The accuracy of the font is also reliable enough,
Ein alignment can be performed quickly.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a perspective view showing an embodiment of the present invention. FIG. 2 is a configuration diagram of a reduction projection type exposure apparatus incorporating the embodiment of FIG. FIG. 3 is a plan view of the wafer showing a pattern and an alignment mark which are exposed and printed in a matrix on a first layer of the wafer by the reduction projection type exposure apparatus of FIG. 2; FIG. 4 is a plan view showing shapes of the alignment mark of FIG. 3 and a light spot for scanning the alignment mark. FIG. 5 is a plan view showing a positional relationship between detection centers of the step alignment detection system and the global alignment detection system shown in FIG. 1; FIG. 6 is a layout diagram showing a specific configuration of a photoelectric detection unit of the step alignment detection system shown in FIG. FIG. 7 is a plan view of the scattered light spatial filter shown in FIG. 6; 8 is a plan view of the diffracted light spatial filter shown in FIG. FIG. 9 is a circuit block diagram showing a specific configuration of a signal processing circuit of a step alignment detection system. FIG. 10 is a flowchart showing the procedure of pre-alignment and fine alignment. FIG. 11 shows a flat rotational deviation angle Δθ in pre-alignment.
FIG. FIG. 12 is an explanatory diagram for obtaining a center position (position in the x direction) of a flat in pre-alignment. FIG. 13 is an explanatory diagram for obtaining a rotation angle θ by a global alignment detection system. [Description of Signs of Main Parts] 6: Projection lens, 7: Wafer, 8: Stage, F
… Flat, LA, LB… Light spot, 21… Laser light source, 22… Beam expander, 23… Cylindrical lens, 24, 70… Beam splitter, 26A, 26B…
… Semi-transmissive mirror, 27A, 27B… Condenser lens, 29A, 29B… Field stop, 30A, 30B… Reflector, 31,41… Photoelectric detector, 71,7
4… Spatial filter, 73,77a, 77b …… Photoelectric device

Continuation of front page    (56) References JP-A-55-123142 (JP, A)                 JP-A-57-55730 (JP, A)                 JP-A-61-22626 (JP, A)                 JP-A-61-219807 (JP, A)                 JP-A-57-65429 (JP, A)                 JP-A-60-189951 (JP, A)                 JP-A-61-134021 (JP, A)                 JP-A-53-105376 (JP, A)                 JP-A-61-131441 (JP, A)                 Shokai Sho 57-88125 (JP, U)

Claims (1)

(57) [Claims] Prior to projecting a reticle pattern image through a projection optical system onto a photosensitive substrate having a notch in a part of the circumference, a predetermined moving direction of a stage that moves two-dimensionally, and the notch Placing the photosensitive substrate on the stage so that the notch direction of the photosensitive substrate substantially coincides with the stage, having a predetermined positional relationship with the optical axis of the projection optical system, and aligning the alignment mark formed on the photosensitive substrate. Detecting light information generated at different points of the notch portion using a light beam from the fine alignment device that can be detected; and rotating the notch portion in the predetermined movement direction based on a change in the light information. Determining the amount of deviation; and positioning the photosensitive substrate such that the amount of rotational deviation is offset when the amount of rotational deviation is greater than or equal to an allowable range; The emissions alignment device, a photosensitive substrate alignment method characterized by comprising the steps of: detecting the alignment mark formed on the photosensitive substrate placed on the stage. 2. 2. The method according to claim 1, wherein the fine alignment device detects the notch and the alignment mark via the projection optical system.