JPH05243121A - Surface position setting device - Google Patents

Abstract

PURPOSE:To maintain a fixed range of allowable deviation at all times even when the sensitivity of detection is changed by a method wherein the sensitivity such as the changing rate inclination and the like of the level of deviation signal against the actual amount of positional deviation is computed, and a correction operation is conducted in accordance with the changing rate or the inclination based on the tolerance value on the changing level of the deviation signal. CONSTITUTION:A processor 118 outputs an instruction to a drive control roller 116 to change a tilt angle of a plane parallel 12 by a very small amount within the prescribed range. When the plane parallel 12 is tilted at a very small specific amount the measured value (value corresponding to the angular position of plane parallel 12 is read in and stored in a memory 108. The level of detection signal FS is converted into a digital value every time of inversion timing, and it is stored in the memory successively. Then, the sensitivity at the zero point of the detection signal is computed by the method of least squares, and an arithmetic operation for correction is conducted by comparing the measured sensitivity with the standard sensitivity.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mask, a reticle, a mask used in the process of manufacturing a semiconductor device, a liquid crystal display device or the like.
The present invention relates to a device for positioning the surface of a substrate such as a wafer or a glass plate with respect to a predetermined reference plane.

[0002]

2. Description of the Related Art Conventional surface position setting devices of this type are widely used for proximity / gap setting, focus adjustment, leveling adjustment, etc. In many cases, a pattern on a mask or reticle is used as a photosensitive substrate. It is incorporated in a so-called exposure apparatus, which is an apparatus for transferring to a predetermined area on a wafer or plate. In order to set the surface position, it is necessary to measure the height position of the entire surface of the target substrate to be aligned or a predetermined local portion with high accuracy. FIG. 1 shows a conventional configuration of an oblique incident light type surface position setting device that detects a height position of a local portion on a substrate as a deviation amount from a virtual surface (reference surface) serving as a reference. This configuration is equivalent to that disclosed in Japanese Patent Laid-Open No. 58-113706, for example, and the surface position of the wafer W is set in FIG. In FIG. 1, a projection optical system 1 image-projects a reticle pattern onto a wafer W. The pattern image of the reticle is formed with the highest contrast in the best image plane (best focus plane) perpendicular to the optical axis AX.

A Z stage 20 mounts a wafer W on the XY
Finely move on the stage 21 in the optical axis AX direction (Z direction),
The surface of a specific shot area on the wafer W is matched with the best focus surface. Now, in order to detect the height position of the surface of the wafer W, that is, the deviation amount of the surface of the shot area in the Z direction from the best focus surface, a projector LSU as an oblique-incidence light type auto-focus detection system and a connection for projection. An image lens system L ₁ , an image forming lens system L ₂ for receiving light, and a light receiver RVU are provided.

The light projector LSU projects an image-forming light beam from the oblique direction onto the surface of the wafer W through the lens system L ₁ , and the light receiver RSU.
The VU receives the specularly reflected light beam from the wafer W via the lens system L ₂ . Then, the light receiver RVU outputs a photoelectric signal that changes according to the light receiving position of the reflected light beam to the focus error detection circuit FD.
Output to. Normally, the imaging light flux from the light projector LSU is projected in the vicinity of the position of the optical axis AX of the projection optical system PL, and when the regular reflection light flux is received so as to match the detection center point in the light receiver RVU, the wafer W The surface (strictly speaking, the partial surface on which the light flux from the projector LSU is projected) is set so as to match the best focus surface.

The error detection circuit FD calculates a deviation signal of a level proportional to the amount of positional deviation in the Z direction of the wafer surface with respect to the best focus surface, based on the signal from the light receiver RVU, and drives the Z stage 20 ( Hereinafter, it will be referred to as Z-DRV) 18. The Z-DRV 18 servo-controls the Z stage 20 so that the level of the deviation signal becomes a predetermined target value (zero or a constant value). At this time, Z-
The DRV 18 is also provided with a circuit that detects how much the difference signal level differs from the target value and determines whether the difference is within the allowable range. This is because the servo control is stabilized and the speed is increased by setting the allowable range with a slight width with respect to the target value.

[0006]

In the apparatus shown in FIG. 1, the level change of the deviation signal output from the focus error detection circuit FD is caused by the position change of the surface of the wafer W in the Z direction in the vicinity of the best focus surface. It is proportional to. The proportional constant corresponds to the surface position detection sensitivity of the oblique incident light type detection system in FIG. 1, and the level change amount on the deviation signal when the position deviation amount of the wafer in the Z direction is ΔZ (μm). Is the voltage ΔV, the detection sensitivity (slope) is defined by ΔV / ΔZ.

If the value of ΔV / ΔZ is large, the detection sensitivity is high, and the responsiveness at the time of servo control of the Z stage 20 is enhanced, but the stability of the servo system may be problematic. On the other hand, if the detection sensitivity becomes low, there is a problem that the precision of the servo system becomes low. In this type of servo system, the response, stability, and drive-in accuracy are all set to be the best, but when you look at the surface position detection systems in multiple exposure devices, the detection sensitivity itself is Not all of them are always available. That is, there are many machine differences. Therefore, even if the permissible range with respect to the target value during servo control is set to be the same on the level of the deviation signal, due to the difference in detection sensitivity (inclination, change rate), Z
The allowable width on the actual position deviation of the direction will differ for each machine. Therefore, there is a problem in that even if the result of performing exposure by performing focusing with one exposure apparatus is good, the condition of the parameter setting for focusing at that time may not be reproduced as it is with another exposure apparatus.

This is a very serious problem in accuracy control during device manufacturing in a manufacturing line using a large number of exposure apparatuses. Further, in recent years, one shot area on the wafer W (area exposed at one time by the projection optical system PL)
A multipoint autofocus system (hereinafter referred to as a multipoint AF system) that simultaneously detects positional deviations of a large number of points in the Z direction is also considered. In this case, since the deviation signal is output by the focus error detection circuit for each of a large number of measurement points set in the shot area, if the detection sensitivity (ΔV / ΔZ) at each measurement point is not uniform, The allowable position deviation range when controlled by the servo system is different for each measurement point, which causes the same problem that the advantage of using the multipoint AF system is reduced by half.

Therefore, the present invention can maintain a constant allowable deviation range even when the detection sensitivity (inclination, rate of change) of the system for detecting the surface position is different or when the detection sensitivity changes over time. An object of the present invention is to provide a surface position setting device that can be used.

[0010]

Therefore, in the present invention,
The sensitivity of the deviation signal obtained by the oblique incident light type surface position detection system or the detection system of the height position of multiple measurement points, that is, the rate of change of the deviation signal level with respect to the actual position deviation amount, the sensitivity such as inclination A means for calculating and a means for correcting the permissible value on the level change of the deviation signal according to the calculated change rate or inclination are provided.

Further, in the multipoint AF system, the functions of the above-mentioned calculation means and correction means are applied to each of the multipoint measurement points.

[0012]

In the present invention, since the level change characteristic of the deviation signal representing the position deviation of the surface is detected and the slope (rate of change) on the level change is calculated, it is considered that there is a change over time. The accurate slope (rate of change) is always obtained. Then, in order to make the position deviation allowable range that changes according to the inclination constant, the allowable range on the level change characteristic of the deviation signal is corrected, so that it is the same for any exposure apparatus regardless of the machine number difference. It becomes possible to unify the focus tracking accuracy. Further, in the multi-point AF system, the positional deviation allowable range for each measurement point can be made uniform, so that focus adjustment in consideration of a focus offset set individually for each measurement point, for example, image plane tilt It is possible to perform optimum focus adjustment or leveling adjustment in consideration of field curvature.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The structure of a surface position setting device according to an embodiment of the present invention will now be described. In the following embodiment, in order to facilitate comparison with the prior art shown in FIG. Take a focus detection system as an example. Hereinafter, an oblique incident light type AF system will be described with reference to FIG. 2, but here, multipoint AF is adopted. The multipoint AF is one in which measurement points for measuring the positional deviation of the wafer W in the optical axis direction (so-called focal point deviation) are provided at a plurality of positions within the projection visual field of the projection lens PL.

In FIG. 2, illumination light IL, which is non-photosensitive to the resist on the wafer W, illuminates the slit plate 1. The light passing through the slit of the slit plate 1 obliquely irradiates the wafer W via the lens system 2, the mirror 3, the diaphragm 4, the projection objective lens 6, and the mirror 6. At this time, if the surface of the wafer W is on the best imaging plane Fo, the slit plate 1
The image of the slit is formed on the surface of the wafer W by the lens system 2 and the objective lens 5. The angle between the optical axis of the objective lens 5 and the wafer surface is set to about 5 to 12 degrees, and the center of the slit image of the slit plate 1 is the optical axis A of the projection lens PL.
X is located at the intersection with the wafer W.

The slit image light flux reflected by the wafer W passes through the mirror 7, the light receiving objective lens 8, the lens system 9, the vibrating mirror 10, and the inclinable parallel plane plate (plane parallel) 12, and the light receiving slit plate. It is re-imaged on 14. The vibrating mirror 10 microvibrates a slit image formed on the light-receiving slit plate 14 in a direction orthogonal to its longitudinal direction, and the plane parallel 12 forms a slit on the slit plate 14 and a reflected slit image from the wafer W. The relative relationship with the vibration center is shifted in the direction orthogonal to the slit longitudinal direction. And vibrating mirror 10
Is oscillated by a mirror drive unit (M-DRV) 11 driven by a drive signal from an oscillator (OSC.) 16.

When the slit image vibrates on the light-receiving slit plate 14 in this manner, the light flux transmitted through the slit of the slit plate 14 is received by the array sensor 15. This array sensor 15 is one in which the longitudinal direction of the slit of the slit plate 14 is divided into a plurality of minute regions, and individual photoelectric cells are arranged in each minute region. Here, an array sensor of a silicon photodiode or a phototransistor is used. Shall be used. Then, the signals from the respective light receiving cells of the array sensor 15 are selected via the selector circuit 13,
Alternatively, they are grouped and input to the synchronous detection circuit (PSD) 17. This PSD 17 has an OSC. An AC signal having the same phase as the drive signal from 16 is input, and synchronous rectification is performed based on the phase of this AC signal. At this time PSD
Reference numeral 17 includes a plurality of detection circuits for individually synchronously detecting respective output signals of a plurality of light receiving cells selected from the array sensor 15, and the respective detection output signals FS are provided to the main control unit (MCU) 30. Is output to. Detection output signal FS
Is a so-called S-curve signal, and becomes zero level when the slit center of the light-receiving slit plate 14 and the vibration center of the reflection slit image from the wafer W coincide with each other.
Is a positive level when is displaced upwards from that state,
It is at a negative level when the wafer W is displaced downward. Therefore, the wafer W whose output signal FS becomes zero level
The height position of is detected as the focal point. However, in such an oblique incident light system, there is no guarantee that the height position of the wafer W at the in-focus point (the output signal FS is zero level) always matches the best image plane Fo. That is, in the oblique incident light system, the system has a virtual reference plane determined by the system itself, and when the wafer surface coincides with the reference plane, PS
Since the D output signal FS becomes zero level, the reference surface and the best image forming surface Fo are set so as to coincide with each other as much as possible at the time of manufacturing the device, but they coincide with each other for a long period of time. There is no guarantee. Therefore, if the virtual reference plane is displaced in the optical axis AX direction by inclining the plane parallel 12 in FIG. 2, the reference plane and the best imaging plane Fo can be matched (or the positional relationship can be defined). ..

In FIG. 2, the MCU 30 inputs the output signal KS from the photoelectric sensor 45 to calibrate the oblique incident light type multipoint AF system, the function to set the inclination of the plane parallel 12, and the multipoint AF. Based on each output signal FS of the system, the motor 19 for driving the Z stage 20
Signal DS to the circuit (Z-DRV) 18 that drives the
And a function of outputting a command to a drive unit (including a motor and its control circuit) 22 that drives the XY stage 21.

FIG. 3 shows the projection field If of the projection lens PL,
It is the figure which looked at the positional relationship with the projection slit image ST of AF system on the wafer surface. The projection visual field If is generally circular, and the pattern area PA of the reticle R is rectangular included in the circle. The slit image ST is formed on the wafer with an inclination of about 45 ° with respect to each of the moving coordinate axes X and Y of the XY stage 21. Therefore, both optical axes AFx of the light projecting objective lens 5 and the light receiving objective lens 8 extend in the direction orthogonal to the slit image ST on the wafer surface. Further, the center of the slit image ST is set so as to substantially coincide with the optical axis AX. With such a configuration, the slit image ST is set to extend as long as possible within the projection area of the pattern area PA. Generally, on the surface of the wafer onto which the pattern area PA is projected, a shot area to be superposed thereon is already formed. Within the shot region, the variation of the uneven portion increases every time the device manufacturing process progresses, and a large uneven change may exist in the longitudinal direction of the slit image ST. Particularly when a plurality of chips are arranged in one shot area, a scribe line for separating each chip is formed in the shot area so as to extend in the X direction or the Y direction. In the extreme case, a step difference of 2 μm or more may occur between the point on the chip and the point on the chip. Since which part in the slit image ST the scribe line is located in is known in advance from the designed shot arrangement, the chip size in the shot, etc., the slit image S
It is naturally understood from which part of the longitudinal direction of T the reflected light should be selected on the array sensor 15.

FIG. 4 shows an example of a specific configuration of the light-receiving slit plate 14 and the array sensor 15.
4 is a chrome layer (light shielding) deposited on the entire surface of the glass substrate,
A transparent slit is formed in a part of it by etching. The slit plate 14 is fixed to a holding metal piece 14A, and the metal piece 14A is fixed to a printed circuit board 15A such as ceramics holding the array sensor 15. As a result, the slits of the slit plate 14 are in parallel with the one-dimensional array of light receiving cells of the array sensor 15 and are in close contact therewith. Thus, the slit plate 14 and the array sensor 15
It is better to contact with or as close as possible to the slit plate 1
An imaging lens system is provided between 4 and the array sensor 15,
The slit plate 14 and the array sensor 15 may be optically conjugated. Although the length of the slit image ST on the wafer shown in FIG. 3 varies depending on the diameter of the projection visual field If, the diameter of the visual field If is 32 with the 1/5 reduction projection lens.
When it is around mm, it is desirable to make it about 1 to 1/3 times the diameter.

Next, the device configuration according to the first embodiment of the present invention will be described with reference to the circuit block of FIG. 5. FIG. 5 shows an array sensor 15 corresponding to one measurement point of a multipoint AF system.
The case where a photoelectric signal from the above one light receiving cell is processed is shown as a representative. Therefore, in the circuit of FIG.
The selector 13 shown in is omitted. Now, the photoelectric signal (AC) from one light receiving cell 15n on the array sensor 15 is converted into a voltage by the IV conversion amplifier 100, and then the PSD 102 (one circuit in the PSD 17 of FIG. 2).
Input to and the reference frequency (vibration frequency) from OSC16
The signal is synchronously detected. Since the output signal of the PSD 102 includes the reference frequency component, the detection signal (S curve signal) FS is obtained by cutting and smoothing the reference frequency component by the next low pass filter (LPF) 104. The level of the detection output signal FS is converted into a digital value by the analog-digital converter (ADC) 106 and stored in the memory (RAM) 108. A
The counter circuit 110 performs the conversion timing of the DC 106 and the address creation for the memory 108.
The counter circuit 110 is the plane parallel 12 in FIG.
Is to count up / down pulses from an encoder 114 coupled to a motor 112 for controlling the slope of the. The motor 112 is drive-controlled by a drive controller 116 that reads the count value of the counter circuit 110. The drive amount of the motor 112, that is, the current position of the plane parallel 12 (count value of the counter circuit)
Information on how much to change from processor 1
Given by 18. The processor 118 reads the level change characteristic (S curve waveform) of the detection signal FS stored in the memory 108, calculates the slope, and sets the allowable range. Further processor 118
Has a function of outputting a digital value of an offset voltage to be added to the detection signal FS to the digital-analog converter (DAC) 120, and a function of opening / closing a switch SW for applying the detection signal FS to the adder 122. ing. The adder 122 outputs a voltage obtained by adding the offset voltage from the DAC 120 and the detection signal FS as a signal DS to the Z-DRV 18. The DAC 120 and the adder 1
The switch SW ₂ between the Z stage 20 and the switch 22 does not add an offset to the detection signal FS itself,
When an offset is added to the drive-in target value within V18, the state shown in the drawing is switched.

Now, in the Z-DRV 18, a differential amplifier 124 and a power amplifier 12 for driving the motor 19 are schematically provided.
6 and the like are incorporated, and the difference between the signal DS from the adder 122 and the target value TS is calculated by the differential amplifier 124. The output of the differential amplifier 124 is sent to the power amplifier 126 via the switch SW ₄ , and the motor 19 is driven. When the switch SW ₄ is switched from the illustrated state, the output of the power amplifier 126 becomes zero, and the Z movement of the Z stage 20 is stopped. Further, since the target value TS is at the zero level via the switch SW ₃ in the illustrated state, the detection signal FS is at the zero level (zero point of the S curve) in a state where no offset is added to the detection signal FS. Then, the motor 19 is driven. When both the switches SW ₂ and SW ₃ are switched from the illustrated state, the offset voltage from the DAC 120 becomes the target value TS, so the level of the detection signal FS becomes the target value TS.
The motor 19 is driven so as to match the offset voltage of.

By the way, the switch SW ₄ is opened when the level of the signal DS is within a predetermined allowable range ± ΔV with respect to the target value TS, and is operated so as to be closed at any time outside the allowable range. To do. The window comparator 130 and the DAC 128 are provided for such operation. The DAC 128 inputs a digital value corresponding to the allowable value ΔV from the processor 118, converts the digital value into an analog voltage, and sends the analog voltage to the window comparator 130. In the window comparator 130, a circuit that generates two values, that is, a value obtained by adding the allowable value ΔV to the target value TS (TS + ΔV) and a value obtained by subtracting the allowable value ΔV from the target value TS (TS−ΔV), as reference voltages. Is provided and the input signal D
The S level has two values (TS + ΔV) and (TS−ΔV)
When the switch SW ₄ is in the middle of the period, a signal for switching the switch SW ₄ from the illustrated state to open it is generated. This signal is also sent to the processor 118 at the same time, and is used to know that the servo movement of the Z stage 20 is completed.

Next, the operation of the apparatus according to the first embodiment of the present invention will be described with reference to FIGS. 2 to 5. As is apparent from the configuration of FIG. 5, this first embodiment is similar to that of FIG. It can be similarly applied to the conventional device shown. First of all,
In the configuration shown in (1), the reference plate FM provided at a position different from the wafer W on the Z stage 20 is positioned at the light beam projection position of the focus detection system (multipoint AF system). The reference plate FM has a size substantially equal to or larger than the projection visual field of the projection lens PL, and is attached so that its surface is as perpendicular as possible to the optical axis AX of the projection lens PL.

The switch SW _{1 in} FIG. 5 is closed as shown, the switch SW ₂ is switched from the position shown in the figure, and the switch SW ₃ is set in the position shown in FIG. Type multi-point AF light receiving cell 1
Focusing is temporarily performed on the measurement point corresponding to 5n. When the focusing on the reference plate FM is completed, the surface of the reference plate FM coincides with the virtual reference plane determined by the oblique incidence AF system having the measurement point defined by the light receiving cell 15n.

After that, the processor 118 switches the switch SW.
Open ₁ to prevent the detection signal FS from being output to the Z-DRV 18. By this, of course, the switch SW ₄
Also remains open, the movement of the Z stage 20 in the Z direction is prohibited. Next, the processor 118 outputs a command to the drive controller 116 to change the inclination angle of the plane parallel 12 by a fixed minute amount within a predetermined range. However, the variable range of the tilt angle of the plane parallel 12 needs to be set before and after the angular position of the plane parallel 12 at the time when the focusing on the reference plate FM is completed.

In this way, when the motor 112 is driven by the drive controller 116 and the plane parallel 12 is tilted by a fixed minute amount, the count value of the counter circuit 110 at that time (the value corresponding to the angular position of the plane parallel 12) is stored in the memory. It is read and stored in 108.
Further, the counter circuit 110 sends to the ADC 106 a plurality of times (m
The ADC 106 converts the level of the detection signal FS into a digital value and stores it in the memory 108 sequentially at each conversion timing. Therefore, with respect to one angular position data obtained from the counter circuit 110, m pieces of data obtained by digitally sampling the level of the detection signal FS m times at fixed time intervals (on the order of μs or ms) are stored in the memory 108. ..

When the above is repeatedly executed within the designated tilt drive range of the plane parallel 12, the data of the level change characteristic (S curve waveform) of the detection signal FS is stored in the memory 108. These data are averaged by the processor 118 level value data of the m detection signals FS sampled for each angular position of the plane parallel 12 by the processor 118 to be one data, and are sequentially re-stored in the memory 108. To be done. The level change characteristic of the detection signal FS thus obtained is as shown in FIG. 6 as an example. In FIG. 6, a broken line connecting the plotted points of each sampling data represents the envelope of the signal FS, the vertical axis represents the level of the detection signal FS, and the horizontal axis represents the angular position of the plane parallel 12, where the angular position is Z. It also uniquely corresponds to the position of the stage 20 in the Z direction.

As is apparent from FIG. 6, the S-curve signal has a point-symmetrical waveform with respect to the zero point of the level, and has a straight line portion centered on the zero point. This straight line is AF
The detection signal FS (deviation signal) detected by the system and the actual amount of positional deviation (deviation of the object surface with respect to the reference surface) are in a range in which a substantially linear relationship (linear relationship) is maintained, and the servo system is exclusively in this range. To use.

Therefore, when the level change characteristic of the detected signal FS is taken into the memory 108, it may be limited to only its linear range. Next, the processor 118 calculates the sensitivity (change rate of the signal level with respect to the change of the position deviation) ΔA at or near the zero point of the detection signal FS from the characteristic of FIG. 6 by using the least square method or the like. In FIG. 6, the slope of the straight line k of the first-order differential equation drawn on the waveform of the detection signal FS with the zero point as the contact point is obtained.

The sensitivity (gradient) ΔA has a unit of V / μm, where the detected signal FS is represented by voltage (V) and the positional deviation is represented by μm. By the way, since the value of the standard sensitivity (or the sensitivity determined by the previous calibration) ΔAr is stored in the processor 118 in advance, the processor 1
18 compares the measured sensitivity ΔA with the standard sensitivity ΔAr. At this time, if there is a character display device for monitoring the measurement sequence in the control rack of the exposure apparatus main body or the console, the value of the sensitivity ΔA measured together with the standard sensitivity ΔAr may be displayed on the display device.

The meaning of the standard sensitivity ΔAr is shown in FIG.
A brief explanation will be given with reference to. In FIG. 7, the vertical axis represents the level of the detection signal FS, the horizontal axis represents the position deviation of the Z stage 20 (or the wafer W, the reference plate FM) in the Z direction,
It is assumed that the zero point coincides with the best focus plane.
When the straight line Kr passing through the zero point in FIG. 7 is the standard sensitivity ΔAr, the servo drive range of the Z stage 20 at the time of focusing, that is, the standard sensitivity ΔAr corresponding to the allowable range Δα of the position deviation that can be considered to be in focus is obtained. The allowable voltage range is set to ± ΔVr. This allowable voltage range ± ΔVr is calculated from the DAC128 in FIG.
This corresponds to the window width set in the comparator 130.

In such a state, it is assumed that the actual level change characteristic of the detected signal FS has a steeper slope of the straight line portion than the straight line Kr of the standard sensitivity ΔAr as in FS ₁ . In this case, the allowable voltage range ± ΔVr set for the servo system is the allowable range Δα ₁ in terms of the position deviation of the Z stage 20, which is smaller than the standard range Δα. For this reason, when the servo system operates stably with the characteristic FS ₁ , the follow-up accuracy is improved, so that no inconvenience is caused.

On the contrary, the characteristic FS₂The slope of the straight line part like
When becomes smaller than the slope of standard sensitivity ΔAr
Occurs. That is, the characteristic FS₂Allowable voltage range above ±
The allowable range of position deviation with respect to ΔVr is Δα₂become,
This is of course Δα ₂> Δα,
The driving accuracy becomes poor. For this reason
Wafer W against the allowable range Δα
The position may be misaligned. That is,
The surface position of the wafer W changes by Δα
₂In this case, the variation may occur and Δα may jump out. this
Focus on each shot area on the wafer W by
Even if the focus is set, the focus accuracy varies depending on the shot area.
I get bigger. Focus accuracy variation
Is the fidelity of the projected pattern image transferred in the shot area.
The reproducibility of each shot of the degree (image quality) will be compromised.

Therefore, the processor 118 compares the measured sensitivity ΔA with the standard sensitivity ΔAr, and if there is a difference, performs the following correction calculation. First, for the measured sensitivity ΔA, if the allowable range on the position deviation is set to Δα equal to the standard value, the following two equations hold. 2 · ΔVr / Δα = ΔAr, 2 · ΔVx / Δα = ΔA where ΔVx is the allowable voltage range on the level characteristic to be obtained. Therefore, by solving the above two equations, the relationship of ΔVx = ΔA · Δα / 2 = ΔA · ΔVr / ΔAr (Equation 1) is obtained.

The processor 118 corrects and calculates a new allowable voltage range ΔVx corresponding to the sensitivity ΔA based on this equation, stores it internally, and a value relating to the window width to be set in the window comparator 130 via the DAC 128. To ΔVx. As is clear from Equation 1, when ΔA> ΔAr, ΔVx> ΔVr, and when ΔA <ΔAr, ΔVx <ΔVr.

As described above, the AF detection signal F
When the calibration of S, that is, the allowable range according to the sensitivity of the focus deviation signal is completed, the processor 118
Returns the switch SW ₁ to the position shown and returns to the normal focusing operation. By the way, in the circuit configuration shown in FIG. 5, two methods are possible for servo control at the time of focusing. One is switch SW ₃
Is set to the position shown in the drawing and the target value TS is set to zero,
The other is a case where the switches SW ₂ and SW ₃ are switched from the positions shown in the figure to set the target value TS to an offset voltage from the DAC 120.

When the target value TS is set to zero, the detection signal F
This is the point where the level change characteristic of S is always OV. Therefore, the offset voltage is added to the detection signal FS via the DAC 120 and the switch SW ₂ , and, for example, the characteristic FS in FIG.
_{Even if 1} or FS ₂ is parallel-shifted by the offset amount in the vertical axis direction, the position where the level becomes zero on the characteristic in the parallel-shifted state is the target position of the Z stage 20. Therefore, assuming that the position in the Z direction, which is the zero point on the characteristic before the offset is added, coincides with the best focus plane, the position to be focused on the wafer surface is intentionally changed from the best focus plane by adding the offset. Can be shifted. The shift amount is uniquely determined by the offset voltage and the sensitivity ΔA of the detection signal.

This is exactly the same even when the offset voltage is applied to the target value TS, and the difference is that the servo system is settled when the level of the detection signal FS matches the offset voltage as the target value. is there. Note that tilting the plane parallel 12 can give a focus offset, but in this case, unlike the offset giving a voltage, the level change characteristic of the detection signal FS is moved in parallel in the direction of the position deviation (Z direction). It will be.
That is, regarding the characteristic FS ₁ or FS ₂ in FIG. 7, it means that the characteristic FS ₁ or FS ₂ is parallel-shifted in the horizontal axis direction in the graph of FIG. 7.

As described above, in the first embodiment, the DAC 128 and the window comparator 130 switch the switch SW ₄ to stop the servo system when it is within the allowable range. However, such a function is performed by the software in the processor 118. May be executed as a target. In that case, AD
The conversion of C106 is always repeated at high speed,
The digital value is sequentially read by the processor 118,
Tolerance range ΔVr or Δ stored digitally
A comparison operation is performed with Vx. The function of the servo system differential amplifier 124 is that the target value T stored digitally is
It is replaced with software that sequentially calculates the difference between S and the digital value from the ADC 106. Then, the calculated difference is D /
The power amplifier 126 is configured to be applied via the A converter.

In this way, when the servo system is stopped upon detecting that the allowable range ΔVr or ΔVx is entered, the difference is calculated by the software that calculates the difference between the target value TS and the digital value of the ADC 106. Is forcibly set to zero and output to the amplifier 126. By the way, the allowable voltage range ± ΔVr shown in FIG. 7 is a considerably small value in an actual device, and the allowable range Δα is also extremely small. As an example, in a projection exposure apparatus used for manufacturing a 16 Mbit DRAM, the effective depth of focus of the projection lens PL is only about ± 0.7 μm, and therefore the allowable range Δα is ±
It is around 0.3 μm. In addition, the range of the straight line portion of the detection signal FS of the oblique incident light type AF system is ±
It is around 3 to 6 μm.

Next, a second embodiment of the present invention will be described. The second embodiment is basically the same as the first embodiment, but the inclination of the detection signal (S curve signal) corresponding to each measurement point of the multipoint AF system of the first embodiment is measured. The difference is that a structure for calibrating the allowable range of the position deviation amount from the target value in the Z direction is added to each measurement point. Therefore, the configurations of FIGS. 2, 3 and 4 are the same in the second embodiment as well.
Some of the configurations are different.

FIG. 8 schematically shows an example of the connection relationship between the array sensor 15 and the selector circuit 13 shown in FIG. As shown in FIGS. 3 and 4 described above, the array sensor 15 is one in which a plurality of light receiving cells are arranged one-dimensionally in the longitudinal direction of the slit image ST on the wafer W. Therefore, the light-receiving cell array on the array sensor 15 is divided into 5 equal parts and divided into 5 parts.
Two groups Ga, Gb, Gc, Gd and Ge are set. Each group includes several light receiving cells.
In this embodiment, as shown in FIG. 8, each of the five groups outputs a photoelectric signal according to reflected light,
The photoelectric signals from each group are five individually prepared synchronous detection circuits (PSD) 102A, 102B, 102C,
The five detection signals FSa, FSb, FSc, F applied to the low pass filter 102D and 102E, respectively.
Sd and FSe are obtained at the same time. Also, in FIG.
The AF processing circuit for each measurement point is set to CGa, CGb, C
Gc, CGd, and CGe are the same in configuration, and only the circuit CGa is shown in detail as a representative.

Five light receiving cells 15A ₁ , 15A ₂ and 15A are included in the group Ga of light receiving cells of the array sensor 15.
₃ , 15A ₄ and 15A ₅ are included, which are wired in series with each of the five switches in the selector circuit 13A and further connected in parallel to each other to connect the IV conversion amplifier 1
00A input. Five switches in the selector circuit 13A is alternatively a command from MCU30 shown in FIG. 2, or any of a plurality (2-5) each is closed, any of the five light-receiving cells 15A ₁ to 15A ₅ The photoelectric signal from one of them or the added signal of the signals from each of the arbitrary plurality (2 to 5) is IV converted. The feedback circuit of the IV conversion amplifier 100A is provided with five resistors and four switches for correcting the gain according to the number of light receiving cells selected from the five light receiving cells, as illustrated. The four switches are appropriately switched by a command from the MCU 30 according to the selection pattern of the light receiving cells by the selector circuit 13A. Here, the four switches are the gain switch 13B. The signal from the IV conversion amplifier 100A is the PSD 102A and the LPF 104.
Detected by A, the circuit CGa outputs a detection signal FSa. Hereinafter, the other circuits CGb to CGe are configured in exactly the same manner and output the detection signals FSb to FSe, respectively.

With the above configuration, when the selector circuit 13A selects any one of the five light receiving cells in the group Ga and closes the switch, the switch 13B causes the combination in the feedback loop of the IV conversion amplifier 100A. It is switched so that the resistance becomes the highest, that is, the gain of the IV conversion becomes the highest. In the configuration of FIG. 8, any two, three, four, or five of the light receiving cells can be connected in parallel. Therefore, the gain correction of the IV conversion by switching the switch 13B is the highest gain. To ½, ⅓, ¼, and ⅕, respectively. As described above, by making it possible to select the light receiving cells even within one group, it is possible to finely set the measurement points corresponding to the pattern structure in the shot region on the wafer W. When some of the light receiving cells in the group are used in parallel, the area as the measurement point is the slit image ST.
However, it is advantageous in terms of averaging effect by adding photoelectric signals, improvement of S / N ratio, and the like.

In the circuit configuration of FIG. 8, five circuits CGa ...
When the selection patterns by the respective selector circuits 13A of CGe can be independently set, it becomes possible to use a method of arbitrarily selecting 5 points out of 25 apparently 25 measurement points in the slit image ST. As described above, when the detection signals FSa to FSe corresponding to each of the five measurement points are obtained, the slit image S is generated due to the difference in the zero point positions of those signals.
It is possible to know the inclination of the surface of the wafer W and the local unevenness in the longitudinal direction on T. That is, it becomes possible to correct (level) the inclination of the wafer by obtaining the flatness of the wafer surface. However, in order to obtain the inclination and the local unevenness of the wafer surface with respect to the reference plane (the moving plane of the XY stage 21) with high accuracy, the detection signal FSa-
It is also necessary to know the inclination (sensitivity) of the straight line portion of FSe accurately in advance. Further, when the Z stage 20 is servo-controlled based on only the detection signal from any one of the five measurement points to perform focusing, the allowable deviation range Δα varies for each detection signal. By selecting an arbitrary measurement point, the effect peculiar to the multi-point AF system that the best focus can be set on a specific portion in the shot area is halved.

Therefore, in front of the ADC 106 shown in FIG. 5, a multiplexer circuit for selecting any one of the five detection signals FSa to FSe is provided, or for each of the five detection signals FSa to FSe. AD
By providing C, the sensitivities ΔAa, ΔAb, ΔAc, ΔAd, and ΔAe of the five detection signals FSa to FSe are obtained in the same manner as in the first embodiment. Then, in the same manner as the above (Formula 1), based on the standard allowable deviation range Δα defined by the standard sensitivity ΔAr, the allowable voltage ranges ΔVxa, ΔVx
b, ΔVxc, ΔVxd, and ΔVxe may be calculated.

Alternatively, one of five measurement points is focused
Sensitivity obtained for one measurement point, eg slit image ST
The sensitivity ΔAc obtained at the measurement point of the group Gc located at the center of may be regarded as the standard sensitivity, and the allowable voltage range may be calibrated corresponding to each of the remaining four sensitivities ΔAa, ΔAb, ΔAd, and ΔAe. in this case,
The allowable deviation range of the AF system corresponding to each measurement point other than the measurement point of interest in the slit image ST is aligned with the allowable deviation range of the AF system corresponding to the measurement point of interest.

By the way, when focusing the wafer W by using the multi-point AF system as described above, as a practical operation, the focus shift at each of the five measurement points is measured, and the shift amounts are averaged. , Or it should be determined as one deviation amount by performing a statistical calculation. Therefore, FIG. 9 shows an example of a signal processing system suitable for the purpose.

FIG. 9 is a diagram in which most of the analog processing parts in the circuit shown in FIG. In FIG. 9, the same members as those in FIG. 5 are designated by the same reference numerals. The five detection signals FSa to FSe are input to an analog multiplexer circuit (MPX) 130 that performs switching according to a command from the processor 118, and one of them is ADC1.
It is output to 06. ADC 106 is processor 118
The conversion timing is controlled by, and the converted digital value is sequentially read by the processor 118, and further written in the memory 108 at the time of calibration within the allowable range. Further, the processor 118 reads the count value from the counter circuit 110 similar to that shown in FIG. 5, and writes it in the memory 108 as necessary.

Further, the processor 118 outputs to the DAC 132 a digital value representing a deviation voltage corresponding to the amount by which the Z stage 20 should be moved for focusing, based on the measured values of the five measurement points. As a result, the differential amplifier 124 and the power amplifier 126 are moved to the Z stage 20.
The Z motor 20 is servo-controlled to set the Z stage 20 to the target position.

With the above configuration, when calibrating the AF system for each measurement point, the plane parallel 12 is driven and controlled so as to be tilted by a certain amount, and then the counter circuit 1
The count value from 10 is stored in the memory 108, and
Switching is performed so that the signals FSa to FSe are sequentially output from the MPX 130, and the digital value from the ADC 106 is read and stored in the memory 108 at each switching. Then, the tilting of the plane parallel 12 by a certain amount is repeated again. Then, the waveform data as shown in FIG. 6 may be created in the memory 108 for each measurement point.

On the other hand, when focusing on the wafer W (or the reference plate FM), switching by the MPX 130 and conversion timing of the ADC 106 are performed at high speed so that the levels of the detection signals FSa to FSe are sequentially read at high speed. Further, the processor 118 temporarily stores a plurality of level values obtained in some of the repetition cycles of reading the level of each detected signal, and performs averaging for each detected signal by software. FIG. 10 is a schematic hardware representation of this situation.
The processor 118 has five stack memory units in software for five detection signals, and FIG. 10 shows one of the stack memory units MST. The stack memory unit MST is, for example, for the detection signal FSa, and every time the level of the signal FSa is read from the ADC 106, the digital value DF _{n of the} level is written in one address and read one cycle before. Digital value DF
_{n− 1} , the digital value DF _n−2 read two cycles ago is sequentially shifted to the adjacent address. When a new digital value is written in the stack memory unit MST, the adder ADD outputs some digital values DF _n , DF _n-1 , DF _n-2.
... is added to obtain the average value. The average value shifts the addition result data by 1 bit in the LSB (least significant bit) direction when the number of digital values to be added is two, and the addition result when the number of digital values to be added is four. The data can be easily obtained by shifting the data of 2 bits in the LSB direction.

By the above operation, the processor 1
Since the level values of the five detection signals FSa to FSe are taken into the circuit 18 in almost real time, the processor 11
8 calculates the position deviation amount from the best focus surface at each measurement point based on those level values. This calculation is started immediately based on each level of the signals FSa to FSe when the shot area to be exposed is moved by the XY stage 21 and stopped at a predetermined position.

However, in order to execute this calculation, it is premised that the sensitivities ΔAa to ΔAe of the detection signals FSa to FSe are obtained. The processor 118 reads the level value (voltage) V of each of the detected detection signals FSa to FSe.
Based on a, Vb, Vc, Vd, Ve and the sensitivities ΔAa to ΔAe, the positional deviation amounts ΔZa, Δ in the Z direction at the respective measurement points.
Zb, ΔZc, ΔZd, and ΔZe are calculated. This operation only needs to multiply the level value and the sensitivity.

Next, the processor 118 determines the weight for each of the measured values ΔZa to ΔZe according to the weighting coefficient preset by the pattern structure (step, etc.) in the shot area, and then simply averages the five measured values. , Minimum 2
The average deviation Δ from the best focus plane of the wafer surface is processed by processing with a power approximation or an arithmetic expression such as a spline function.
Calculate Zav. Then, the processor 118 obtains the difference (ΔZc−ΔZav) between the current measurement value ΔZc and the average deviation amount ΔZav of the detection signal FSc of the AF system having a measurement point at the center of the shot area, and further detects the detection corresponding to the difference. The level difference on the signal FSc is obtained. Detection signal F
Since the sensitivity of Sc is ΔAc, the level difference is ΔA
The level value Vc ′ obtained from c · (ΔZc−ΔZav) and added (or subtracted) to this level Vc of the current detection signal FSc is the target value. That is, the processor 118 changes the level of the detection signal FSc from Vc to Vc '.
The drive data that changes to is output to the DAC 132 in FIG. As a result, the Z stage 20 moves in the Z direction, the processor 118 sequentially monitors the level of the detection signal FSc, and when the level reaches the target value Vc ′ within the allowable voltage range set at the time of calibration, the DAC is executed.
The drive data to 132 is set to zero and the Z stage 20 is stopped.

Next, a third embodiment of the present invention will be described.
Here, assuming that the sensitivity of the AF system corresponding to each measurement point of the multipoint AF system has been confirmed and the allowable range has been calibrated, how to give an offset to each measurement point will be described. As shown in FIG. 3, in a multipoint AF system having a large number of measurement points in the projection visual field If, the height of the wafer surface at which the AF system signal corresponding to each measurement point becomes a zero point is set as much as possible by the projection lens PL. It is made to match the best focus plane (reticle conjugate plane). However, the best focus surface is not exactly an ideal plane due to the optical performance of the projection lens PL, the flatness of the pattern surface of the reticle R, and the like. Even if it is an ideal plane, there is no guarantee that it is strictly orthogonal to the optical axis AX of the projection lens PL depending on the holding state of the reticle R and the like. Therefore, if the best focus points (heights) are found at innumerable points in the pattern projection image by the projection lens and a plane including them is assumed, the optical axis AX is overall.
It is tilted with respect to the ideal horizontal plane perpendicular to, and is partially curved with respect to the ideal horizontal plane.

Therefore, the true best focus position is obtained at high speed for each position of each measurement point of the multipoint AF system, and the deviation amount between the Z direction position detected by the AF system as a zero point and the true best focus position is calculated. It is necessary to know each measurement point. If the amount of deviation is known for each measurement point, it is given as an offset value to each AF system, so that a multi-point AF system that also takes into account the unevenness and inclination of the wafer surface, as well as the inclination and curvature of the best focus surface. It will be possible.

FIG. 11 shows a TT for detecting the true best focus position at an arbitrary position within the field of view If of the projection lens PL with the reticle for actual device exposure mounted.
The configuration of an L-FC (through-the-lens focus check) system is shown. In FIG. 11, the reticle R having the circuit pattern area PA for actual device manufacturing formed on the lower surface is held by a reticle holder (not shown). Aperture surface (pupil surface)
The optical axis AX of the projection lens PL, which is schematically shown in a front group and a rear group across the EP, passes through the center of the reticle R, that is, the center of the pattern area PA, perpendicularly to the reticle pattern surface. The Z stage 2 on which the wafer W is placed and is moved by a very small amount (for example, within ± 100 μm) in the optical axis AX direction.
A reference plate FM is fixed to a part of 0 at a height position substantially equal to the surface of the wafer W. On the reference plate FM, as shown in FIG. 12, a plurality of transmissive slits extending in the X direction are arranged in the Y direction at a constant pitch, and a mark ISy,
A mark ISx in which a plurality of transmissive slits extending in the Y direction are arranged at a constant pitch in the X direction, and an oblique slit ISa that is 45 ° with respect to each of the X and Y directions are formed. These slit marks ISx, ISy,
ISa is formed by depositing a chromium layer (light-shielding layer) on the entire surface of a quartz reference plate FM and engraving it as a transparent portion.

In FIG. 11, a mirror M1, an objective lens 40 for illumination, and an optical fiber 41 are provided below the reference plate FM (inside the Z stage 20), and the illumination light from the exit end of the optical fiber 41 is an objective. The slit mark IS of the reference plate FM is condensed by the lens 40.
Irradiate x, ISy, and ISa from the back side. A beam splitter 42 is provided on the incident end side of the optical fiber 41, and the exposure illumination light IE is introduced into the optical fiber 41 via a lens system 43. The illumination light IE is reticle R
It is desirable to obtain it from a light source for illumination (a mercury lamp, an excimer laser, etc.), but a dedicated light source may be prepared separately.
However, when using a different light source, the same wavelength as the exposure illumination light,
Alternatively, it is necessary to use illumination light having a wavelength extremely close to that.

The illumination condition of the reference plate FM by the objective lens 40 is matched as much as possible with the illumination condition of the projection lens PL at the time of pattern projection. That is, the numerical aperture (NA) of the illumination light on the image side of the projection lens PL and the numerical aperture (NA) of the illumination light from the objective lens 40 to the reference plate FM are substantially matched. Now, with such a configuration, the illumination light I
When E is introduced into the optical fiber 41, an image light flux that travels backward to the projection lens PL is generated from the marks ISx, ISy, ISa on the reference plate FM. In FIG. 11, it is assumed that the Z stage 20 is set so that the reference plate FM is located slightly below the best focus surface (reticle conjugate surface) Fo of the projection lens PL. At this time, the reference plate F
The image light flux L1 generated from one point on M passes through the center of the pupil plane EP of the projection lens PL, is condensed in a plane Fr slightly shifted downward from the pattern plane of the reticle R, and then diverges to form the pattern of the reticle R. It reflects on the surface and then returns to the original optical path.
Here, the surface Fr is at a position conjugate with the reference plate FM with respect to the projection lens PL. If the projection lens PL is a telecentric system on both sides, the emission marks ISx, I on the reference plate FM
The image light fluxes from Sy and ISa are regularly reflected on the lower surface (pattern surface) of the reticle R and again marked with marks ISx, ISy, and I.
It returns so as to overlap with Sa. However, when the reference plate FM is deviated from the best focus surface Fo as shown in FIG. 11, the marks ISx, ISy, IS on the reference surface FM.
When the blurred reflection image of a is formed and the reference plate FM coincides with the surface Fo, the surface Fr also coincides with the pattern surface of the reticle R, and each mark IS on the reference plate FM.
Sharp reflected images of x, ISy, and ISa are formed so as to be superimposed on the respective marks. FIG. 13 schematically shows the relationship between the emission mark ISx and the reflected image IMx when the reference plate FM is defocused. In the projection lens PL that is telecentric on both sides, the reflected image IMx is thus the light emission mark ISx that is its own source.
Projected on. When the reference plate FM is defocused, the reflected image IMx becomes larger than the shape size of the mark ISx, and the illuminance per unit area also decreases.

Then, of the reflected image formed on the reference plate FM, the light flux of the image portion not shielded by the original marks ISx, ISy, ISa is received by the optical fiber 41 via the mirror M ₁ and the objective lens 40, and the beam Splitter 42,
The photoelectric sensor 45 receives light through the lens system 44. The light receiving surface of the photoelectric sensor 45 is arranged substantially conjugate with the pupil surface (Fourier transform surface) EP of the projection lens PL.

In a conventionally known TTL-FC system of this type, a light emitting type mark, a light receiving type mark, or a simple reflection type reference mark is arranged on the image plane side of the projection lens, and a reticle is formed in the XY plane. An electrical signal for contrast detection is obtained by scanning the mark optically or electrically. However, in the configuration of FIG. 11, it is not necessary to scan the reference plate FM in the XY plane to obtain an electric signal representing the contrast change as in the conventional case, and the Z stage 20 is only moved in the vertical direction (Z direction). A contrast signal can be obtained.

FIG. 14 shows the signal level characteristic of the output signal KS of the photoelectric sensor 45, and the horizontal axis represents the position of the Z stage 20 in the Z direction, that is, the height position of the reference plate FM in the optical axis AX direction. In FIG. 14, FIG. 14A shows a light emission mark I.
1 shows signal levels when Sx, ISy, and ISa are back-projected on the chrome portion in the pattern surface of the reticle R, and FIG.
4 (B) shows the signal level when back projected onto the glass portion (transparent portion) in the pattern surface. Normally, the chrome portion of the reticle is vapor-deposited on a glass (quartz) plate with a thickness of about 0.3 to 0.5 μm, and the reflectance of the chrome portion is naturally much larger than the glass portion reflectance. However, since the reflectance at the glass portion is not completely zero, the level is considerably small as shown in FIG. 14 (B), but detection is possible. Generally, the reticle for manufacturing an actual device has a high pattern density,
It is considered that the probability that all the back-projected images of the light emission marks ISx, ISy, ISa (within a region of 1 to several millimeters on the reticle) are simultaneously applied to the glass portion (transparent portion) in the reticle pattern is extremely small.

In any case, when the surface of the reference plate FM is moved in the optical axis direction so as to cross the best focus surface Fo, the signal level becomes the maximum value at the position Zo in the Z direction. Therefore, the position of the Z stage 20 in the Z direction (detection signal F
S level) and the output signal KS are measured at the same time, and the signal K
The position of the best focus surface Fo can be obtained by detecting the Z direction position when the level of S becomes maximum, and this detection method can detect the best focus surface Fo at any position in the reticle R. Become. That is, it is not necessary to use a pattern (mark) formed at a specific position in the reticle R as in the prior art, and as long as the reticle R is set on the object side of the projection lens PL, it is always in the projection visual field. The absolute focus position (best image plane Fo) can be measured at the position. In addition, as described above, the chrome layer of the reticle R has a thickness of 0.3 to 0.5 μm, and the detection error of the best imaging plane Fo caused by this thickness is that the projection magnification of the projection lens PL is reduced to ⅕. Then, (0.3-0.5)
× (1/5) ² = 0.012-0.02 μm, which is almost negligible.

Therefore, after the light emitting mark group shown in FIG. 12 is positioned near the measurement point of the multipoint AF system in the projection field If, the level of the signal KS and the level of the detection signal FS are changed while moving the Z stage 20. After the changes and the changes are simultaneously captured by the AD converter and stored in the memory (RAM), FIG.
As shown in FIG. 3, the two waveforms are compared by the processor, and the level value ΔFS on the detection signal FS corresponding to the position T ₁ of the peak point on the level change of the signal KS is obtained. It is the offset amount (specific error) from the absolute position of the AF system corresponding to the measurement point.

Therefore, the TTL-FC system of FIG.
Offset amount ΔFSa, Δ for each of the two measurement points
FSb, ΔFSc, ΔFSd, ΔFSe are obtained and stored. In principle, this operation is performed once every time the reticle R is exchanged. However, in a general-purpose IC exposure stepper, the same reticle may be mounted all day or over several days. It is desirable to periodically remeasure the offset amount.

When the above offset amount measurement is executed,
Sensitivity ΔAa to Δ of the detection signals FSa to FSe at each measurement point
Since the measurement of Ae and the setting of the permissible deviation range Δα have already been performed, the multipoint AF system can execute the focusing in consideration of the image plane inclination and the image plane curvature.
FIG. 16 is a graph in which one example of the offset tendency when field curvature exists is exaggerated. This Figure 1
6, the horizontal axis represents the position of the slit image ST in the longitudinal direction, and the vertical axis represents the offset amount at each of the five measurement points MPa to MPe.

When this offset tendency is stored in advance,
When performing the focusing by measuring the displacement of the wafer surface in the Z direction at each of the five measurement points MPa to MPe and obtaining an average approximate plane, or when measuring any one of the five measurement points At the time of pinpoint AF in which the wafer surface is focused on the basis of the detection signal at the point, the offset amount thereof is corrected by addition or subtraction with respect to each level of the detection signal FS, and then from the best focus surface. It is treated as a true misregistration amount. For example, in FIG. 16, for the measurement point MPc, the detection signal FS
The point where c is positively offset by ΔFSc is the true best focus position.

Further, since the allowable deviation range Δα is calibrated to be constant at all measurement points MPa to MPe, the allowable deviation range of a constant width is centered on the curve of the offset tendency shown by the broken line in FIG. Will exist. That is, regarding the measurement point Cc, when the permissible deviation range is ± Δα / 2, the best focus position is recognized when the level value of the detection signal FSc is between ΔFSc + Δα / 2 and ΔFSc−Δα / 2. Will be done.

As described above, when the actual focusing is performed after the offset amount as the systematic error with respect to the true best focus surface is obtained, the detection signals FSa to FSa are obtained.
If the permissible deviation range Δα for each Se is set to be constant, extremely accurate focusing can be achieved in any of the averaging AF and the pinpoint AF in which the approximate surface is specified.

As shown in FIG. 16, the projection image plane itself is curved,
If it is tilted, the approximate plane of the image plane is determined in advance by using the least-squares method or the spline function, and thereafter, a focusing method for matching the wafer surface with the approximate plane can also be performed. is there. By the way, in the projection method of the slit image ST shown in FIG. 3 described above, the slit image ST is inclined by 45 ° with respect to the XY coordinate axes, and therefore the multipoint AF system lacks the ability to specify a surface. Therefore, another set of multipoint AF systems may be arranged in the 135 ° direction with respect to the X and Y axis directions, and the slit image ST ′ as shown in FIG. 17 may be projected simultaneously. This makes it possible to set nine measurement points (circles in FIG. 17) including the four corners and the center of the pattern area PA (shot area on the wafer). For that purpose, 5 points are selected on each of the two sets of array sensors, and a total of 10 (or 9) synchronous detection circuits are required.

Further, the multipoint AF system projects the minute slit image or the spot image obliquely onto a plurality of predetermined measurement points within the projection visual field If, and reflects slits (spots) from the respective measurement points. The present invention is also applicable to a method in which an image is collectively received by a two-dimensional image pickup device, for example, a CCD, and a deviation signal is output corresponding to a deviation amount of a plurality of received images from a reference pixel position on each CCD. The same effect can be obtained. Further, the present invention can be similarly applied to a multi-point AF system of a type in which spot light is projected onto four points on a wafer via only a projection lens.

[0073]

As described above, according to the present invention, the permissible deviation range that occurs depending on the variation in the detection sensitivity (tilt) of the oblique incident light type AF system corresponding to each measurement point of the fixed point AF system or the multipoint AF system Is always set to a constant value, it is possible to set the surface position such as focus adjustment between different devices with the same accuracy. Also, since the allowable deviation range can always be kept constant with respect to systematic errors due to drift over time, in an apparatus that performs long-term exposure operations, the focus setting conditions intended by the operator can be operated with good reproducibility. Can be made

[Brief description of drawings]

FIG. 1 is a diagram showing the configuration of a conventional surface position setting device (oblique incidence light type AF system).

FIG. 2 is a diagram showing the overall configuration of an oblique incident light type AF system according to an embodiment of the present invention.

FIG. 3 is a plan view showing a state in which a slit image is projected by an oblique incident light type AF system.

FIG. 4 is a perspective view showing a structure of an array sensor as a photoelectric detector of an oblique incidence type AF system.

FIG. 5 is a block diagram showing a signal processing system of an oblique incidence light type AF system according to an embodiment of the present invention.

FIG. 6 is a waveform diagram showing an example of a detection signal (deviation signal).

FIG. 7 is a waveform diagram for explaining the difference in the slope (sensitivity) of the detection signal.

FIG. 8 is a block diagram showing a configuration of a signal processing system suitable when the present invention is applied to a multipoint AF system.

FIG. 9 is a block diagram showing a modified example of a configuration of a signal processing system and a control system suitable for a multipoint AF system.

FIG. 10 is a block diagram schematically showing averaging during signal reading.

FIG. 11 is a diagram showing a configuration of a through-the-lens type focus check (TTL-FC) system.

FIG. 12 is a plan view showing the shape of a TTL-FC light emitting mark.

FIG. 13 is a plan view showing the relationship between a light emitting mark and a reflection image.

FIG. 14 is a waveform diagram showing an example of a photoelectric signal obtained by the TTL-FC system.

FIG. 15 is a waveform diagram showing a state of offset measurement of an oblique incidence AF system using a photoelectric signal from the TTL-FC system.

FIG. 16 is a diagram schematically showing an offset tendency when a projected image has a curvature.

FIG. 17 is a plan view showing a modification of the arrangement of measurement points in a multipoint AF system.

[Explanation of symbols for main parts]

 PL Projection lens AX Optical axis W Wafer R Reticle 5 Projection objective lens 8 Receiving objective lens 12 Plane parallel 15 Array sensor 17 Synchronous detection circuit (PSD) 18 Z stage drive unit (Z-DRV) 20 Z stage 21 XY stage

Claims (6)

[Claims] 1. A light projecting means for projecting a light flux from a diagonal direction onto a predetermined portion of the substrate surface, and a reflected light flux of the light flux on the substrate surface is received, and a photoelectric signal is output according to a change in the light receiving position. Light receiving means, an arithmetic circuit that outputs a deviation signal according to a deviation amount of the substrate surface with respect to a predetermined reference surface based on the photoelectric signal, and the substrate as the reference surface based on the deviation signal. In a surface position setting device provided with a substrate moving means for moving and setting to a predetermined position in a vertical direction, a level change of the deviation signal caused when the substrate surface and the reference surface are relatively displaced, Level change detecting means for detecting within a range having a substantially linear relationship with the deviation amount; and the level change characteristic at a point where the substrate surface and the reference plane substantially match each other based on the characteristic of the detected level change. Up And inclination calculation means for calculating an inclination value;
A correction unit that corrects an allowable range set on the level change characteristic for controlling the substrate moving unit according to a difference between the calculated inclination value and the reference value. Surface position setting device. 2. The correcting means determines an offset corresponding to a deviation between a target surface on which the substrate surface is to be set and the reference surface, from at least the light receiving means, the arithmetic circuit, or the substrate moving means. The apparatus of claim 1 including offset generating means for providing one. 3. A light projecting means for projecting a light beam from a plurality of predetermined portions on the surface of the substrate in an oblique direction, and a reflected light beam from each of the plurality of predetermined portions on the surface of the substrate are individually received, and Based on each of the plurality of photoelectric signals and a plurality of photoelectric elements that output a photoelectric signal according to the change, based on each of the plurality of photoelectric signals, each deviation amount of a plurality of predetermined portions on the substrate surface with respect to a predetermined reference plane An arithmetic circuit for individually outputting a deviation signal according to the above, and at least one of the plurality of deviation signals
In a plane position setting device comprising a moving means for moving the substrate in a direction perpendicular to the reference plane or in a direction inclined to the reference plane based on one deviation signal, each of a plurality of predetermined portions on the substrate surface. The characteristic of the level change of each deviation signal generated when the reference surface is relatively displaced in the direction perpendicular to the reference surface is individually detected within a range having a substantially linear relationship with the deviation amount. Level change detection means; inclination calculation means for calculating respective inclination values on the plurality of level change characteristics at points where each of a plurality of predetermined portions on the substrate surface substantially coincides with the reference surface; In order to align the allowable range of the deviation amount for each predetermined portion, the allowable level change range set for each of the plurality of level change characteristics is set according to the inclination value calculated for each of the plurality of predetermined parts. Individually Surface position setting device being characterized in that a positive correcting means. 4. A moving means for holding a substrate to be positioned substantially parallel to a predetermined reference plane and moving it in a direction perpendicular to the reference plane, and a plurality of measurement points set on the surface of the substrate. And measuring the amount of positional deviation from each of the reference planes of the plurality of measurement points with respect to a direction perpendicular to the reference plane,
Based on at least one deviation signal of a plurality of deviation signals output for each measuring point, a multipoint measuring means for outputting a deviation signal corresponding to the deviation amount for each measuring point, A surface position setting device including control means for controlling; change rate detecting means for detecting a rate of change in level of the deviation signal with respect to change in the deviation amount at each of the plurality of measurement points; On the basis of the rate of change of each of the detected measurement points, the deviation signal is set so that the allowable deviation amounts of the measurement points on the substrate from the reference surface are substantially the same when the position is set by the moving means. And a permissible range setting means for individually setting the permissible range on the level change of each of the plurality of measurement points. 5. A projection exposure apparatus for projecting an image of a pattern of a mask substrate onto a photosensitive substrate through a projection optical system, wherein the surface position setting device according to claim 4 is provided, and the projection optical system is provided. A surface position setting device for a projection exposure apparatus, wherein one of the mask substrate and the photosensitive substrate is set by the moving means with the surface perpendicular to the optical axis of the system as the reference surface. 6. The apparatus according to claim 5, wherein the projection exposure apparatus has a Z stage that holds the photosensitive substrate and moves the photosensitive substrate in an optical axis direction of the projection optical system, as the moving unit. Further, the Z stage is provided with a reference plate on which a reference mark is formed for measuring inclination or curvature of the projection image plane of the projection optical system with respect to the reference plane with the mask attached. Surface position setting device.