JP2580651B2 - Projection exposure apparatus and exposure method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a step-and-repeat projection exposure apparatus for manufacturing a semiconductor IC, a liquid crystal display, a photomask, or the like.

[Conventional technology]

Conventionally, in this type of apparatus, an autofocus mechanism that detects a focus at a wafer printing position (exposure position) and controls the Z direction (vertical position) in the optical axis direction of the projection optical system is essential. Recently, the depth of focus has become shallow due to the high resolution of the projection lens, and there has been a problem in that the resolution within the exposure area due to the unevenness and inclination of the wafer surface and the uniformity of the line width are reduced. For this reason, a mechanism (generally called auto-leveling) for detecting and controlling the horizontal position of the wafer for each exposure position has been proposed. For example, as disclosed in Japanese Patent Application Laid-Open No. 58-113706, a horizontal position detecting device is a device in which a grazing incidence type collimator type leveling detection system and a grazing incidence type focus detection system are integrally combined. Are known.

By the way, in recent years, there is a tendency to shorten the wavelength of an exposure light source to achieve high resolution in order to manufacture an IC having a higher degree of integration. For example, use of a KrF excimer laser (wavelength λ = 248.5 nm) as a light source has been considered. In the case of a projection exposure system using this KrF excimer laser, TTL (Through The Lens) alignment (alignment) using non-exposure light realizes an alignment optical system that corrects chromatic aberration because there is no appropriate light source near this wavelength. It is actually difficult to do. Even if alignment is performed using the excimer laser light itself, it must be solved because the resist is exposed, and the excimer laser itself is a pulsed light source, and the output varies greatly from pulse to pulse, causing problems in accuracy. There are many points that have to be done.
For this reason, in a projection exposure apparatus using a DUV (far-ultraviolet) light source, an off-axis type using a microscope that is arranged at a fixed distance from a projection lens and exclusively detects an alignment mark on a wafer. Alignment is effective.

This is because an off-Axis type wafer mark detection optical system has almost no restrictions on the exposure wavelength and the detection method, and alignment with high reproducibility can be expected.

[Problems to be solved by the invention]

However, in a conventional projection exposure apparatus having an off-axis type alignment apparatus, a mechanism for performing auto-focusing and auto-leveling only at an exposure position (a field of view of a projection lens) is employed. The off-axis type wafer mark detection optical system causes defocus due to unevenness, and after detecting the position of the alignment mark on the wafer, moves each exposure area (local shot area) of the wafer to the exposure position However, since the exposure is performed by performing the auto-leveling, there is a problem that the positional deviation in the horizontal direction due to the leveling operation deteriorates the alignment accuracy.

The present invention has been made in view of such conventional problems, and has as its object to minimize the defocusing of a mark detection optical system and a decrease in alignment accuracy due to leveling after wafer position detection. .

[Means to solve the problem]

In order to solve the above problems, in the present invention, an auto-focusing or auto-leveling mechanism is provided in conjunction with an off-Axis type mark detection device so that even at a mark detection position, auto-focusing and auto-leveling can be performed. did.

(Operation)

First, the effect of providing the auto-focus mechanism in the off-axis type wafer mark detection device is, of course, to prevent the out-of-focus of the wafer with respect to the detection device due to unevenness of the wafer surface. This eliminates the accuracy degradation due to the defocus inherent in the detection device, and when there is a parallel displacement (so-called telecentric displacement) between the optical axis of the mark detection device and the optical axis of the projection lens, alignment of the wafer caused by the defocus. This has the effect of eliminating mark position measurement errors.

Next, the operation of attaching the auto-leveling mechanism to an off-Axis type wafer mark detection device will be described. Here, the position (shot coordinate) of each exposure region of the layer at the previous stage formed on the wafer to be subjected to the overlay exposure is (x _i ^L , y _i ^L ) (i = 1, 2,... N). The suffix L on the shoulder indicates that the layer was printed by operating the leveling for each exposure area even in the exposure of the layer at the previous stage. Similarly, assuming that the position of each exposure area when printing is performed without operating the leveling in the previous layer is (x _i , y _i ) (i = 1, 2,..., N), the coordinates (x _i ^L , y _i) ^L ) and the coordinates (x _i , y _i ) are different due to a displacement error due to the leveling operation. off-Axis
For mold alignment, usually, the positions of the marks attached to several exposure areas on the wafer are measured, and based on the numerical values, the array positions of the other exposure areas are calculated by interpolation, and the stages are sequentially moved by that numerical value. Move and perform exposure at the exposure position.
The position of the stage, of course, is precisely monitored by a laser interferometer. Now, in a state where the position multiplied by the leveling of the measured exposure region ^{_{(X j L, Y j L}} ), in a state not to apply leveling and (X _j, Y _j). In general, the exposure process for each layer is usually performed by another exposure apparatus. However, since the irregularities on the upper surface of the wafer holder are so small that they can be ignored, X _i ^L = x _i ^L and Y _i ^{L at} i = j. = Y _i ^L
And X _i ≠ x _i ^L and Y _i ≠ y _i ^L. Now (x _i ^L , y _i ^L )
Since the is to interpolation ^{_{calculation, (X j L, Y j}} L) from ^{_{(x i L, y i L}} ) of the seek it is (X _j, Y _j) from (x _i ^L,
The accuracy is better than determining y _i ^L ). Therefore, when the position measurement is performed after the leveling operation is performed, higher positioning accuracy can be expected.

Of course, each time the position of a wafer mark is measured at each measurement position on the wafer, the measurement mark is moved to the exposure position to operate the leveling, and then the off-Axi
It is sufficient to measure the position of the measurement mark with an s-type mark detection device, but this requires a very long time and has a disadvantage that the throughput is reduced. In this regard, if an off-Axis type wafer mark detection device is provided with a leveling mechanism, such a disadvantage does not occur.

〔Example〕

FIG. 1 is a diagram showing the overall configuration of a projection exposure apparatus according to an embodiment of the present invention. Exposure proof light IL having a uniform illuminance distribution irradiates reticle R having a circuit pattern or the like, and reticle R is horizontally held by reticle holder RH. The projection lens PL projects the pattern of the reticle R onto a local area on the wafer W, ie, a shot area to be exposed at one time. Projection lens PL is the case of this embodiment, the principal ray of the imaging of the reticle side and wafer side are designed together so as to be parallel to the optical axis AX _1. A resist layer is coated on the surface of the wafer W, and is arranged such that the surface of the wafer W and the best image forming plane of the pattern image formed by the projection lens PL substantially coincide with each other during exposure.

The wafer W is fixed on the wafer holder WH by vacuum suction, and is flattened and corrected so that desired flatness is maintained. The wafer holder WH is held on a leveling stage LS that can be tilted in an arbitrary direction under conditions that minimize horizontal displacement. Leveling stage LS is provided on the Z stage ZS moving up and down in the direction of the optical axis AX _1, the lowest fixed to the Z stage ZS 2 places the leveling drive unit 10a, 10b and 1
It is inclined in an arbitrary direction by the fixed points at the two places. The Z stage ZS is a Z stage that moves two-dimensionally in the XY direction for step-and-repeat or alignment.
It is provided via a direction drive mechanism 12. The vertical movement of the Z stage ZS with respect to the stage XYS is performed by the motor 14, and the two-dimensional movement of the stage XYS is
This is performed by the motor 16.

On two sides (only one side in FIG. 1) of the Z stage ZS, a movable mirror 18 having reflection surfaces orthogonal to each other in the XY plane is fixed, and the movable mirror 18 receives a laser beam from a laser interferometer 20. Irradiated vertically. The center line of the laser beam of the interferometer 20 is determined so as to be located substantially in the same plane as the surface of the wafer W (strictly, the best image forming plane of the projection lens PL). Therefore, the interferometer 20 sets the distance to the reflecting surface of the movable mirror 18 to 0.
The current position (coordinate value) of the stage XYS is sequentially detected by measuring with a resolution of about 01 to 0.02 μm. In one place on the Z stage ZS, there is a reference reflector FP having a size of about the effective visual field area (for example, 20 mmφ) of the projection lens PL.
Has been fixed. Chromium or the like is uniformly deposited on the surface of the reference reflection plate FP to maintain a constant surface accuracy.
Note that a fiducial mark (for example, a shape similar to a mark on a wafer) serving as a reference for various alignments is formed on the surface of the reflection plate FP.

Now, in the field of view of the projection lens PL on the wafer W,
Illumination light (non-exposure wavelength) from a light transmission system 22 for autofocus and autoleveling is applied obliquely, and the reflected light is incident on a light reception system 24. Since the configurations of the light transmitting system 22 and the light receiving system 24 are disclosed in detail in Japanese Patent Application Laid-Open No. 58-113706, detailed description thereof is omitted here. An imaging light flux for projecting an image, and the surface of the wafer W
The light receiving system 24 outputs a focus signal F _sg1 representing the focus when the image coincides with the best image forming plane. Focus signal F
_sg1 is an output voltage (a so-called S-curve voltage) corresponding to the defocus direction and amount in a state where the focus is defocused back and forth by a predetermined amount from the in-focus state. Furthermore, light transmission system
The illumination light for auto-leveling from 22 is
A parallel light beam is obliquely applied to the wafer W in a size including the exposure area for the shot, and the light receiving system 24 sets a relative position between the average surface of the exposure area surface and the best imaging plane of the projection lens PL. And outputs a leveling signal _Lsg1 relating to the inclination direction and amount.

Also at a position separated by a predetermined distance from the optical axis AX ₁ of the projection lens PL is an alignment mark on the wafer W, or Fi wafer alignment system for detecting a fiducial mark (mark detection optical system) WGA is arranged You. Non-exposure wavelength illumination light (white light, etc.) 26
Is directed toward the wafer W (or the reflector FP), and a telecentric objective lens 28 that receives reflected light from the wafer W (or the reflector FP) and a mark (or a fiducial mark) on the wafer W A photoelectric receiver 30 for photoelectrically detecting the reflected light is provided. Further, the photoelectric receiver 30 has an index mark 30 as a reference for alignment.
a is built-in. In this embodiment, the matching state between the index mark 30a and the mark on the wafer W is detected.
When the illumination light 26 is a laser beam that projects a spot light (or a slit-shaped sheet beam) stationary on the wafer W along the optical axis AX ₂ of the objective lens 28, the spot light itself serves as an index. The index mark 30a is unnecessary. Then, the photoelectric receiver 30 outputs an alignment signal _Asg1 corresponding to the amount of deviation between the index mark 30a and the wafer mark. In the case of the method without the index mark 30a, the photoelectric receiver 30 scans the wafer W in the X or Y direction with respect to the spot light (or the sheet beam), and aligns the profile of the mark with the waveform corresponding to the information. outputs a signal a _sg1, position or for spot light mark on the basis of the stage position measurement value by the interferometer 20 and the waveform, the position with respect to the optical axis AX ₁ of the projection lens PL is measured.

A light transmitting system 32 and a light receiving system 34 are provided in order to operate the auto-focusing and the auto-leveling of the wafer W even when the wafer surface is observed by the wafer alignment system WGA. The light transmitting system 32 and the light receiving system 34 may have exactly the same configuration as the light transmitting system 22 and the light receiving system 24 described above. And, the light receiving system 34 is similar to the light receiving system 24, a focus signal F _sg2 corresponding to the optical axis AX ₂ direction of relative displacement between the surface of best focus plane and the wafer W of the projection lens PL (or reflective plate FP) And a leveling signal _Lsg2 relating to the direction and amount of the relative inclination between the best imaging plane and the surface of the wafer W (reflector FP). The best imaging plane of the objective lens 28 and the best imaging plane of the projection lens PL (a conjugate plane with the reticle R) are made as coincident as possible.

Above, focus signal F _sg1 , F _sg2 , leveling signal L
Each of _sg1 and _Lsg2 is input to the main control system 42 via the selector circuit 40. The selector circuit 40 selects which of the detection systems performs auto-focusing and leveling. One of the selected focus signals F _sg0 ,
Outputs the leveling signal _Lsg0 . The main control system outputs a switching signal SS to the selector circuit. This switching is broadly divided according to the sequence of the equipment, and the wafer alignment system
When various marks are detected by the WGA, the signals F _sg2 and L
_sg2 is selected as the signals F _sg0 and L _sg0 , and the signals F _sg1 and L _sg2 are selected during the exposure operation on the wafer W. Main control system
42 is the motor based on the selected focus signal _Fsg0
14 to output a signal DS _z for driving the wafer W (or the reflection plate F).
Adjust the height position of the surface of the P), a drive source 10a further selected on the basis of the leveling signal L _sg0, signal DS _a for driving the 10b, and outputs a DS _b, adjusting the tilt of the surface of the wafer W . The main control system 42 receives the measurement values from the interferometer 20 and
Output of drive signal to motor 16 and alignment signal A
_sg1 is also input, and the device is controlled overall.

Next, the operation of this embodiment will be described with reference to FIG. FIG. 2 shows a planar arrangement relationship between the projection lens PL and the objective lens 28 of the wafer alignment system WGA, and further shows a positional relationship with the wafer W at the time of alignment and at the time of exposure. Also, four shot areas SA are shown on the wafer W in FIG. 2 as a representative. Also, alignment marks WM ₁ and WM ₂ are provided in the scribe line area (width of about 50 to ₁₀₀ μm) associated with each of the shot areas SA. Now,
Assuming that the optical axis AX ₂ of the wafer alignment system WGA, that is, the distance between the index mark 30a and the center of the pattern projection image of the reticle R has been accurately obtained in advance, the main control system 42 becomes as shown in FIG. First, the reflector FP is moved below the alignment system WGA. Here, the main control system 42 determines the inclination direction and amount of the surface of the reflector FP detected by the light receiving system 34 as a signal L _sg2
Is stored based on Next, the reflection plate FP is moved below the projection lens PL, and here the main control system 42 similarly stores the inclination direction and the amount of the reflection plate FP detected by the light receiving system 24 based on the signal _Lsg1 . With this operation, a relative detection offset between the two leveling sensors is obtained, and calibration is performed.

Specifically, it is desired that the common reflector FP detected by the two leveling sensors (the light receiving system 34 and the light receiving system 34) be detected as the same reference plane. Therefore, at least one of the two leveling sensors, for example, the light receiving system 34,
By providing parallel plate glass or the like in a part of the optical path of the set of the light transmitting system 32 and changing the angle of the parallel plate glass with respect to the optical path, the light receiving system 34 can be equivalently inclined to the virtual plane detected as zero inclination. To Then, the angle of the parallel flat glass is adjusted so that the previously detected relative detection offset becomes zero. With this operation, calibration between the two leveling sensors is performed. As described above, for example, a method disclosed in Japanese Patent Application Laid-Open No. 62-58624 can be used as it is as a method of adjusting the inclination of the virtual reference plane by inserting parallel flat glass into the optical path of the leveling sensor.

Further, the leveling offset adjusting mechanism using the parallel plate glass may be provided in the combination of the light receiving system 24 and the light transmitting system 22. In the case where the reflector FP shown in FIG. 1 is provided in a part of the leveling stage LS, the reflector FP is firstly used by using the signal _Lsg1 from the light receiving system 24 to generate a virtual reference plane. Leveling stage LS so that it is parallel to
Is adjusted and fixed at that angle. Next, the reflection plate FP is detected by the light receiving system 34, and the angle of the parallel flat glass is servo-controlled by driving the motor based on the signal _Lsg2 . That is, the angle of the parallel flat glass is corrected so that the light receiving system 34 detects the reflection plate FP having the fixed inclination as zero inclination.
This calibrates the relative detection offset between the two leveling sensors to zero.

Next, the wafer W to be subjected to overlay exposure is moved below the wafer alignment system WGA, and a mark detection operation is performed as shown on the left side in FIG. Alignment system marks WM ₁ accompanying the upper side of one shot area SA in Figure 2 WGA
The detection is shown by the index mark 30a.
The main control system 42 based on the measurements from the alignment signal A _sg1 and interferometer 20, is detected and stored mark position (coordinate values) as the index marks 30a and the mark WM ₁ matches.
Similarly detected and stored for other positions of the mark WM ₂ on the wafer W. The above-described mark detection operation is similarly performed by moving the stage XYS for each of a plurality of shot areas on the wafer W to obtain a plurality of mark position information. When detecting the marks MW ₁ and MW ₂ on the wafer W in this manner, the main control system 42 switches the selector circuit 40 to select the signals F _sg2 and L _sg2 . Accordingly, the main control system 42 servo-controls the motor 14 in response to the selected signal _Fsg2, and the surface of the local region including the mark of the wafer W is optically connected to the best image _forming surface (or the index mark 30a) of the objective lens 28. The height of the Z stage ZS is successively adjusted so as to coincide with each other. At the same time, the main control system 42 responds to the selected signal _Lsg2 so that the driving sources 10a and 10b
And the inclination of the leveling stage LS is sequentially adjusted so that the best image forming plane of the equivalent projection lens PL and the surface of the local area including the mark on the wafer are optically parallel to each other.

The main control system 42 _confirms that the auto focus and the leveling are correctly servo-locked in response to the focus signal F _sg2 and the leveling signal L _sg2 , and then detects the mark position based on the alignment signal A _sg1. Do.

Next, the main control system 42 obtains the regularity of the arrangement of the center points of the shot areas SA on the wafer W by a statistical calculation method based on the measured plurality of mark position information, and determines the stage to be connected at the time of exposure. Find the XYS coordinate value (shot address).

Then, the main control system 42 moves the wafer W under the projection lens PL as shown on the right side in FIG.
Overlay exposure of the pattern of the reticle R is performed for each SA. At this time, the main control system 42 switches the selector circuit 40 to select the focus signal F _sg1 and the leveling signal L _sg1 .
Thus, the height position of the Z stage ZS is servo-controlled by the motor 14 in response to the signal _Fsg1 via the main control system 42, and the inclination of the leveling stage LS is servo-controlled by the drive sources 10a and 10b in response to the signal _Lsg1. Is done. This allows
The best imaging plane of the projection lens PL and the wafer surface are parallelized and matched for each shot area, and optimal image quality and alignment accuracy are obtained. As described above, when the mark detection is performed by the off-axis method, extremely high-precision alignment can be achieved by keeping the wafer surface parallel to the best imaging plane of the projection lens PL.

As described above, in the present embodiment, each focus sensor or leveling sensor directly detects the positional relationship between the best imaging plane of each lens PL and 28 and the wafer surface via the projection lens PL or the objective lens 28, a so-called method. TTL (through the
The same effect can be obtained by using a (lens) method. In addition, focus sensors and leveling sensors inject gas at a constant pressure from a small nozzle onto the wafer surface, and detect the back pressure to detect the distance between the nozzle and the wafer surface.
A similar effect can be obtained by using a so-called air micrometer method. However, for leveling, it is necessary to provide micro nozzles at least at three places and average the intervals at each nozzle position to form a wafer surface.

In addition, the autofocus and leveling sensor of the present embodiment irradiates the illumination light to the wafer surface obliquely from one direction, but irradiates the illumination light from each of two orthogonal axis directions, and tilts the wafer surface, Alternatively, separate light receiving systems (two-segment sensor, PSD, etc.) with separate focus planes in both axial directions
May be detected.

Further, in this embodiment, by controlling the attitude of the wafer surface by using both the leveling signal L _sg2 and focus signal F _sg2 from off-Axis scheme wafer alignment optical system WGA light receiving system 34 provided in the mark Although the detection is performed, it may be sufficient to perform the posture control (one of the Z direction and the leveling) using at least one of the signals.

〔The invention's effect〕

As described above, according to the present invention, in a projection exposure apparatus that performs off-axis type alignment and performs leveling of a photoreceptor (a wafer or the like), it is possible to perform off-axis type detection without lowering throughput. It is possible to prevent the optical system from defocusing due to irregularities on the surface of the photoconductor. Further, an effect of minimizing a decrease in the positioning accuracy due to the leveling operation can be obtained.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a projection exposure apparatus according to an embodiment of the present invention, and FIG. 2 is a plan view showing a planar arrangement with a projection lens wafer alignment optical system and a positional relationship of a wafer. [Explanation of Signs of Main Parts] R: Reticle, PL: Projection lens, W: Wafer, LS: Leveling stage, ZS: Z stage, WGA: Wafer alignment system, 22, 32: Light transmission System, 24, 34: light receiving system, 40: selector circuit, 42: main control system.

Claims (7)

(57) [Claims] 1. A pattern of a mask is exposed to each of local areas of a photoreceptor via a projection optical system, and is exposed to the photoreceptor by a mark detection optical system disposed at a predetermined distance from the projection optical system. In a projection exposure apparatus that detects a formed alignment mark and relatively aligns the mask and the photoconductor, holding the photoconductor and tilting the surface of the photoconductor in an arbitrary direction Attitude adjusting means; a first leveling sensor for detecting first information relating to a relative inclination between the surface of the photoreceptor positioned within a field of view of the projection optical system and a predetermined first reference plane; A second leveling sensor for detecting second information relating to a relative inclination between the surface of the photoreceptor positioned within a visual field of the system and a predetermined second reference plane; and any one of the first information and the second information Choose one or the other,
Control means for controlling the attitude adjusting means based on the selected information. 2. The control means selects the second information when the mark detection optical system detects a mark on the photoconductor, and changes the pattern of the mask via the projection optical system to the photoconductor. 2. The apparatus according to claim 1, further comprising an information switching unit that selects the first information when the exposure is performed upward. 3. The apparatus according to claim 1, wherein the first leveling sensor is connected to the first surface of the photoconductor without passing through the projection optical system.
Either an oblique incidence method for detecting a relative inclination with respect to a reference surface, or a through-the-lens method for detecting a relative inclination between the surface of the photoconductor and the first reference surface via the projection optical system 3. Apparatus according to claim 1 or claim 2, characterized in that it is configured as one or the other. 4. An oblique incidence method, wherein the second leveling sensor detects a relative inclination between the surface of the photoconductor and the second reference surface without passing through the mark detection optical system.
Or a through-the-lens system for detecting a relative positional relationship between the surface of the photoconductor and the second reference surface via the mark detection optical system. The device according to claim 1 or 2. 5. An alignment system disposed at a predetermined distance from said projection optical system prior to projecting and exposing a pattern of a mask on a photosensitive substrate mounted on a moving stage via a projection optical system. An exposure method for detecting a mark formed on the photosensitive substrate by using the moving stage and moving the moving stage to an exposure position based on the detection result, wherein a surface of a reference reflecting member provided on a part of the moving stage Moves the moving stage so that the sensor is positioned within a field of view of the projection optical system, and sends first information on a relative positional relationship between a surface of the reference reflecting member and a predetermined first reference surface to a first sensor. Moving the moving stage so that the surface of the reference reflection member is located within the field of view of the alignment system, and Detecting second information about a relative positional relationship with a second reference plane by a second sensor; and detecting the first sensor and the second sensor based on a difference between the first information and the second information. The first sensor and the second sensor so that a relative detection offset between the first sensor and the second sensor is corrected.
Calibrating at least one of the sensors; and, after detecting the calibration, detecting the mark on the photosensitive substrate by the alignment system.
The position of the photosensitive substrate is adjusted based on information detected by the sensor, and when the pattern of a mask is exposed on the photosensitive substrate by the projection optical system, based on the information detected by the first sensor. Adjusting the attitude of the photosensitive substrate. 6. An oblique incidence method wherein said first sensor detects a relative positional relationship between a surface of said reference reflecting member or a surface of said photosensitive substrate and a predetermined reference surface without passing through said projection optical system. Or via the projection optical system, it may be configured as one of a through-the-lens system that detects a relative positional relationship between the surface of the reference reflection member or the surface of the photosensitive substrate and a predetermined reference surface. The method of claim 5, characterized in that: 7. An oblique incidence system for detecting a relative positional relationship between the surface of the reference reflection member or the surface of the photosensitive substrate and a predetermined reference surface without passing through the alignment system. Or, through the alignment system, is configured by one of a through-the-lens system that detects the relative positional relationship between the surface of the reference reflective member or the surface of the photosensitive substrate and a predetermined reference surface A method according to claim 5, wherein