 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a stage apparatus for correcting the inclination
of a body, and particularly to a tilting apparatus capable of supporting
thereon and inclining a photosensitive substrate to which the pattern of a
mask is transferred, which apparatus is suitable for an exposure apparatus
used in the photolithography process for manufacturing, for example,
semiconductive elements, liquid crystal display elements, thin film
magnetic heads, etc.
2. Related Background Art
In an exposure apparatus used in the photolithography process, the pattern
of a mask or a reticle is transferred onto a photosensitive substrate (a
wafer to which photoresist is applied or a glass plate or the like). In
this case, the work of setting the whole or a local area of the surface of
the photosensitive substrate in parallelism to a plane which is the
reference, i.e., so-called levelling (tilting), is effected. For example,
in the proximity system wherein a mask and a photosensitive substrate are
made proximate to each other with a gap of the order of 10 .mu.m to 300
.mu.m, the pattern surface of the mask and the surface of the
photosensitive substrate are made parallel to each other over the whole
area thereof. Also, in the projection system wherein the image of the
pattern of a mask or a reticle is imaged on a photosensitive substrate
through a projection optical system, the image plane of the projection
optical system and the surface of the photosensitive substrate are made
parallel to each other.
Such levelling work is done to obtain a good quality of transfer image (the
quality of pattern profile formed on the photosensitive substrate) on the
whole area of the surface of the photosensitive substrate to which the
pattern of the mask or reticle is transferred. Particularly, in an
exposure apparatus of the projection type which is provided with a
projection optical system of high numerical aperture and high resolving
power, the depth of focus is markedly small as compared with the size of
the projection field (image field) of the projection optical system. This
depth of focus .DELTA.F is represented as .DELTA.F=.lambda./(2NA.sup.2) by
the wavelength .lambda. of illuminating light for exposure and the
numerical aperture NA of the projection optical system.
Assuming here that .lambda.=365 [nm] (the i-line of a mercury lamp) and
NA=0.6, the actually effective depth of focus .DELTA.F is about 0.5 .mu.m
(1 .mu.m in terms of the width in the direction of the optical axis of the
projection optical system) relative to the best focus position. In
contrast, the size of the pattern image projected differs depending on the
exposure apparatus, but is of the order of 15 mm.times.15 mm in a typical
wafer stepper. This, when considered in terms of the thickness of a plate
material of 1 m square, corresponds to only 66.7 .mu.m. Because of such
very small depth of focus, in a projection exposure apparatus, it is
necessary to effect levelling accurately with the localized waviness or
the like of the surface of the photosensitive substrate taken into account
to thereby keep the degree of parallelism thereof to the projection image
plane.
To effect such levelling, it is necessary to accurately measure the amount
of inclination of the whole or a local area of the surface of the
photosensitive substrate from a fiducial plane. Various measuring systems
therefor have heretofore been proposed, and as typical ones of them, there
are known (A) Japanese Patent Application Laid-Open No. 58-103136, (B)
U.S. Pat. No. 4,084,903, (C) U.S. Pat. No. 4,558,949 and (D) U.S. Pat. No.
4,383,757.
In publication (A), there is disclosed a system in which the height
position of each of a plurality of points (three or more points) on a
photosensitive substrate is measured by a gap sensor such as an air
micrometer and on the basis of the measured values and the coordinates
value of each measuring point in XY plane, the approximate plane
expression of the surface of the photosensitive substrate is specified by
the least square method, whereafter the amount of deviation of the
approximate plane from a fiducial plane is determined by the coordinates
positions of the three drive points of a levelling mechanism for the
photosensitive substrate, and the amount of deviation in the height
direction at each of those drive points is corrected.
U.S. patent (B) discloses a system in which the nozzles of an air
micrometer are disposed at four locations around the lower portion of the
barrel of a projection optical system and two of those four nozzles are
provided on the X-axis and the remaining two are provided on the Y-axis,
and the distance (gap) between the surface of a photosensitive substrate
and the barrel is measured by each nozzle, and the amount of inclination
.alpha..sub.y of the photosensitive substrate about the Y-axis is found
from the difference in back pressure between the two nozzles on the X-axis
and the amount of inclination .alpha..sub.x of the photosensitive
substrate about the X-axis is found from the difference in back pressure
between the two nozzles on the Y-axis, and by the use of these amounts of
inclination .alpha..sub.x and .alpha..sub.y, a holder for the
photosensitive substrate is vertically moved by three piezo elements to
thereby correct the inclination.
In U.S. patent (C), there is disclosed a system in which collimated light
is applied from an oblique direction to a local area on a photosensitive
substrate to which a projected image by a projection optical system is
transferred, and the reflected light thereof (a parallel beam of light) is
imaged into a spot-like shape by a condensing lens and is received by a
four-division photoelectric element, and the average amount of inclination
of the local area on the photosensitive substrate is detected from a
variation in the received position of the spotlight on the four-division
photoelectric element.
U.S. patent (D) discloses a system in which a light spot is projected onto
each of a plurality of points on a photosensitive substrate through a
projection optical system, and the reflected light thereof is re-imaged
through the projection optical system and from a variation in the contrast
of this re-imaged image, the focus error at the projected point of each
light spot on the photosensitive substrate in the direction of the optical
axis of the projection optical system (Z-direction) is detected, and a
holder for holding the photosensitive substrate is inclined by three drive
units (servo-motors) so that the focus error may be substantially zero at
each projected point.
Besides the above-described conventional techniques (A) to (D), what is
necessary for levelling is the structural precision and stability of the
levelling mechanism itself, and a conventional technique therefor is
disclosed in U.S. patent (D). As the levelling mechanism, there is also
known a system as disclosed in (E) U.S. Pat. No. 4,770,531 wherein drive
points in Z-direction are provided at locations trisecting the
circumference of a levelling table and at each drive point, the levelling
table and the base thereof are coupled together by a doughnut-shaped leaf
spring to thereby enhance the lateral rigidity of the levelling table.
There is further known in (F) U.S. Pat. No. 4,504,144 a levelling control
system in which a focus error is measured at each of three points in a
local area on a photosensitive substrate to which the pattern image of a
mask or a reticle is transferred, and from the result of the measurement,
the amounts of inclination (.theta.x, .theta.y) in the direction of the
X-axis and the direction of the Y-axis in a field and the overall focus
error amount (f) are found as analog values and by the use of the
coordinates position (analog value) of a field on a wafer and the
coordinates positions (analog values) of three drive points on a wafer
levelling table, besides the amounts of inclination (.theta.x, .theta.y)
and the focus error amount (f), the amount of movement of a servo-motor at
each of the three drive points (the amount of correction of the drive
point in Z-direction) is calculated by an analog calculation circuit.
Among the conventional techniques as described above, the levelling control
system disclosed in publication (A) has suffered from the inconvenience
that the heights at three or more measuring points on the photosensitive
substrate are measured by the gap sensor such as an air micrometer and
therefore the time required for the measurement of the heights is long and
the responsiveness of the levelling operation is bad. In a control system
wherein the responsiveness of the levelling operation is bad, it is
difficult to apply to an exposure apparatus a scanning system particularly
such as a step and scan system in which a reticle and a photosensitive
substrate are synchronously scanned to thereby successively transfer the
pattern images of the reticle onto the photosensitive substrate.
The levelling control system disclosed in U.S. patent (B), which uses an
air micrometer, is similar to publication (A) in that the responsiveness
is bad. Further, in U.S. patent (B), the angles of inclination about two
axes are first calculated from the focus positions at four points and the
amount of drive of the driving element is newly calculated from those
angles of inclination, and this has led to the inconvenience that the
calculation time becomes longer and the responsiveness becomes worse.
Also, in U.S. patent (C), only the average angle of inclination is measured
and the focus position (the height) is not measured, and this has led to
the inconvenience that a sensor for the detection of the focus position
becomes discretely necessary and the control mechanism is complicated. In
U.S. patent (D), the focus position at each point on the photosensitive
substrate is measured through the projection optical system, and this
leads to the possibility that this system cannot be applied to an exposure
apparatus which does not use a projection optical system and that an
illumination optical system or the like for applying illuminating light to
the reticle becomes complicated. On the other hand, the system in which
the lateral rigidity is enhanced as disclosed in U.S. patent (E) suffers
from no special inconvenience.
Further, in the levelling control system disclosed in U.S. patent (F), as
in U.S. patent (B), two steps of calculation in which the angles of
inclination about two axes and the focus position are first calculated
from the focus positions at three points and the amount of drive of the
driving element is newly calculated from those angles of inclination and
the focus position are effected, and this has led to the inconvenience
that the calculation time becomes longer and the responsiveness becomes
worse. Also, only the focus positions at three points are measured, and
this leads to the inconvenience that when for example, there is a
measuring point which is peculiarly high or low in the focus position, the
errors of the angle of inclination of the exposed surface of the
photosensitive substrate with respect to the fiducial plane and the focus
position become great.
In this connection, when the pattern images of the reticle are to be
transferred to each shot area of the photosensitive substrate, unevenness
(level difference) is sometimes created in each shot area by the process
hitherto. Patterns of various line widths are mixedly included in the
pattern images of the reticle and therefore, to form a clear-cut image on
the whole surface of each shot area, it is desirable that the area on the
photosensitive substrate to which one of the pattern images of the reticle
which is narrowest in line width is transferred be adjusted to the
fiducial plane. In the conventional levelling control system, however, it
has been impossible to preponderantly adjust a desired area on the
photosensitive substrate to the fiducial plane.
SUMMARY OF THE INVENTION
It is the object of the present invention to provide a stage apparatus
capable of adjusting the angle of inclination and/or the height position
(a position in a direction perpendicular to the surface of a substrate) of
a substrate in order to adjust the surface of the substrate to a fiducial
plane, and particularly a stage apparatus suitable, for example, for an
exposure apparatus of the step and repeat type or the step and scan type
of which high responsiveness is required.
The stage apparatus according to the present invention has a table for
holding a substrate thereon, a stage supporting the table thereon and
two-dimensionally movable in a predetermined plane, a driving member for
displacing each of the three fulcrums of the table in a direction
perpendicular to the predetermined plane, a first detector for detecting
the position of the table in the predetermined plane, a second detector
for detecting the deviation between the surface of the substrate and a
predetermined fiducial plane in the direction perpendicular to the
predetermined plane at each of at least three measuring points on the
substrate, a calculator for calculating any residual deviation in the
direction perpendicular to the predetermined plane at each of the three
fulcrums of the driving member on the basis of the results of the
detection by the first and second detectors, and a controller for
controlling the amounts of displacement of the three fulcrums of the
driving member on the basis of the calculated residual deviation, the
integrated value of this residual deviation and the differentiated value
of this residual deviation. It is desirable that at this time, weights
W.sub.1 to W.sub.n be given to the deviation at each of at least three
measuring points on the substrate and the calculator calculate the
residual deviation at each of the three fulcrums of the driving member so
that a residual error component calculated with those weights given to the
result of the detection by the second detector may be minimum.
Now, in the calculator of the present invention, the operation of finding
the residual deviation (control deviation) at each fulcrum of the driving
member from the output of the second detector is based on the following
principle.
First, as shown in FIG. 2 of the accompanying drawings, the plane in which
the substrate is moved is defined as xy plane and an axis perpendicular to
this xy plane is defined as the Z-axis. The second detector (16A to 16C)
is fixed to the Z-axis, the coordinates of the measuring points P.sub.1 to
P.sub.n thereof are defined as (x.sub.i, y.sub.i) (i=1 to n), and the
measured value by the second detector, i.e., the deviation between the
surface of the substrate and the predetermined fiducial plane in the
direction of the Z-axis at each measuring point P.sub.i (hereinafter
referred to as the "defocus amount") is defined as .DELTA.f.sub.i. Also,
the surface of the substrate approximated, for example, by the least
square method (the approximate plane) is defined by the following
expression by the use of coefficients A, B and C:
Z=A.multidot.x+B.multidot.y+C (1)
Next, the weighting by a weight coefficient W.sub.i is effected to the
measured value .DELTA.f.sub.i at each measuring point P.sub.i and the
residual error component S by the square sum of the weighting like the
following expression is defined.
##EQU1##
Here, as a method of setting the weight coefficient W.sub.i, when the
present invention is applied to an exposure apparatus of the collective
exposure type such as a stepper, for example, the value of the weight
coefficient to a measuring point in or near the area on the substrate
(wafer) to which a pattern of the narrowest line width on a mask is
transferred can be made great. Also, when the present invention is applied
to an exposure apparatus of the scanning exposure type, the weight
coefficient W.sub.i is a function of the relative coordinates of the
two-dimensional coordinates (X, Y) of the table (14) holding the substrate
and the Z-axis (if there is a projection optical system, the optical axis
thereof), i.e., W.sub.i (X, Y). When the weighting is not effected,
W.sub.i in expression (2) can be W.sub.i =1 (i=1 to n).
The values of the coefficients A, B and C which determine the approximate
plane are then determined so that the residual error component S may be
minimum. That is, the coefficients A, B and C are given by the following
expression by the use of a matrix Q and a variable vector F (which will be
described later):
##EQU2##
Here, the matrix Q is a matrix of 3 columns.times.3 rows as shown below.
##EQU3##
The numerators q.sub.11 to q.sub.33 and denominator q.sub.o of the elements
of this matrix are given by the following expressions by the use of the
coordinates (x.sub.i, y.sub.i) and the weight coefficient W.sub.i :
##EQU4##
Also, the variable vector F in expression (3) is defined by the following
expression:
##EQU5##
Here, the coordinates (x.sub.i, Y.sub.i) of each measuring point P.sub.i of
the second detector are a constant and therefore, when as in the
collective exposure system such as a stepper, the weight coefficient
W.sub.i is a constant, the matrix Q is a constant matrix. On the other
hand, in the case of the scanning exposure system, the weight coefficient
W.sub.i is a function of the two-dimensional coordinates (X, Y) of the
table and therefore, the matrix Q also is a function of the coordinates
(X, Y). Also, the variable vector F is determined by the defocus amount at
each measuring point P.sub.i outputted by the second detector, the
coordinates position of each measuring point P.sub.i of the second
detector and the weight coefficient W.sub.i.
In the present invention, with the predetermined fiducial plane (e.g. the
image plane of a projection optical system) as a plane of z=0, the surface
of the substrate (an approximate plane) is made coincident with the
fiducial plane (the plane of z=0) by the levelling operation and the
focusing operation. That is, the control deviation of the three fulcrums
(16A to 16C) of the drive member (15) is nothing but the distance in the
direction of the Z-axis between the approximate plane of the surface of
the substrate and the fiducial plane (the plane of z=0) at the position of
each fulcrum and therefore, the control deviations (residual deviations)
e.sub.1, e.sub.2 and e.sub.3 at the fulcrums (16A to 16C) are given by the
following expression:
##EQU6##
Here, the matrix R is defined by the following expression:
##EQU7##
Here, (X.sub.1, Y.sub.1), (X.sub.2, Y.sub.2) and (X.sub.3, Y.sub.3) are the
coordinates positions of the three fulcrums (16A-16C) with the Z-axis
(when there is a projection optical system, the optical axis thereof) as
the reference, as shown in FIG. 3 of the accompanying drawings.
Accordingly, the matrix R is a matrix varying with the movement of the
table (14) and further, the substrate (12). That is, when the coordinates
of the three fulcrums (16A to 16C) with the center OW of the substrate
(12) as the reference are defined as (X.sub.10, Y.sub.10), (X.sub.20,
Y.sub.20) and (X.sub.30, Y.sub.30), the matrix R of expression (8) can be
expressed as follows:
##EQU8##
Here, (X, Y) is the coordinates of the center 0W of the substrate (12) with
the Z-axis as the reference, and is detected by the first detectors (19X,
19Y, 20X, 20Y) for measuring the coordinates position of the table (14) in
xy plane. In the manner described above, there is obtained the relation
between the defocus amount .DELTA.f.sub.i outputted by the second detector
and the control deviations of the three fulcrums (16A to 16C) of the drive
member (15).
Next, drive speed command values u.sub.1, u.sub.2 and u.sub.3 for
displacing the fulcrums (16A to 16C) are found by a controller (22) so
that the control deviations e.sub.2, e.sub.2 and e.sub.3 obtained in this
manner may be made zero. When the so-called PID control system which is
proportion, integration and differentiation control is applied, the drive
speed command values are given by the following expression by the use of
integration gain k.sub.I, proportion gain k.sub.p and differentiation gain
k.sub.D :
##EQU9##
By effecting this control, the control deviations e.sub.1, e.sub.2 and es
can be made zero. That is, the approximate plane of the surface of the
substrate (12) by expression (1) coincides with the fiducial plane of z=0.
The approximate plane of the surface of the substrate (12) is determined
by the weight coefficient W.sub.i and thus, if the weight coefficient
W.sub.i to the measuring point P.sub.i of the second detector set within
or near the area on the substrate (12) in which it is particularly desired
to suppress the defocus amount small is set to a great value, optimum
control therefor will be effected. Also, in the aforedescribed control,
the focusing operation and the levelling operation are performed at a time
without being distinguished from each other and therefore, control is
effected at a high speed as compared with a case where these operations
are executed continuously (time-serially). Further, the control applied
here is PID control and therefore has both of the high-speed adaptability
by the effect of the differentiation term and the characteristic that the
steady deviation by the effect of the integration term is zero.
As described above, in the present invention, the three fulcrums of the
drive member are displaced to thereby effect the levelling operation and
height adjustment (focusing operation) and further, by the calculator, the
residual deviation at each fulcrum is directly found from at least three
deviations detected by the second detector. Thus, according to the present
invention, as compared with the system for calculating the angles of
rotation about the x-axis and about the y-axis and height, the calculation
time is shortened and moreover, the three fulcrums are driven at a time,
whereby the levelling operation and height adjustment (focusing operation)
are executed at a time and therefore, the throughput of the levelling and
height adjusting process is improved.
Further, in the present invention, the operation of each fulcrum is
controlled by the PID control system and therefore, there can be
constructed a servo-system which is excellent in the responsiveness to any
change in the control target and free of steady deviation, and a high
imaging performance can be obtained when the present invention is applied
as the focusing and levelling stage of an exposure apparatus. Also, when a
weight is given to each of the three or more measuring points of the
second detector, for example, the weight of the measuring point near a
desired location on the substrate is made great, whereby the height
deviation (defocus amount) at that measuring point can be made small, that
is, the optimization of the height deviation on the substrate becomes
possible.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 schematically shows the construction of the focusing and levelling
mechanism of the projection exposure apparatus of FIG. 4.
FIG. 2 is a view used for the illustration of the defocus amount of the
surface of a wafer in FIG. 1 and the time plane of a projection optical
system.
FIG. 3 is a perspective view used for the illustration of the relations
among the z-axis of FIG. 1, the measuring points of a multipoint AF sensor
and three fulcrums.
FIG. 4 shows a projection exposure apparatus to which is applied a stage
apparatus according to an embodiment of the present invention.
FIG. 5 shows an example of the distribution of the focus positions on the
wafer by the multipoint AF sensor in FIG. 4.
FIG. 6 is a cross-sectional view taken along the line A--A of FIG. 5.
FIG. 7 shows an example of the construction of one fulcrum in FIG. 4.
DESCRIPTION OF THE PREFERRED EMBODIMENT
A stage apparatus according to an embodiment of the present invention will
hereinafter be described with reference to the drawings. This embodiment
is one in which the present invention is applied to a stage apparatus for
effecting the levelling and focusing of a wafer in a projection exposure
apparatus.
FIG. 4 shows the construction of a projection exposure apparatus used in
the present embodiment. In FIG. 4, exposure light IL emitted from an
illuminating device including a light source, an optical integrator and an
aperture stop or the like passes through a first relay lens 2, a reticle
blind (variable field stop) 3, a second relay lens 4, a mirror 5 and a
condenser lens 6 and illuminates a pattern area 8 on a reticle 7 with
substantially uniform illuminance. The plane on which the reticle blind 3
is disposed is conjugate with the pattern forming surface of the reticle
7, and by varying the position and shape of an opening in the reticle
blind 3, the position and shape of the illuminated area on the reticle 7
are set. As the light source in the illuminating device 1, use is made of
a super-high pressure mercury lamp, an excimer laser source, a YAG laser
harmonic generating device or the like.
The light passed through the pattern area 8 on the reticle 7 enters a
projection optical system PL, which projects the image of a pattern in the
pattern area 8 onto a shot area 13A on a wafer 12 to which photoresist is
applied. The z-axis is taken in parallelism to the optical axis AX of the
projection optical system PL, the x-axis is taken in a direction parallel
to the plane of the drawing sheet of FIG. 4 in a two-dimensional plane
perpendicular to the optical axis AX, and the y-axis is taken in a
direction perpendicular to the plane of the drawing sheet of FIG. 4. The
reticle 7 is held on a reticle stage 9, and a main control system 10 for
generally controlling the operation of the entire apparatus effects the
adjustment of the position of the reticle stage 9 through a drive system
11.
On the other hand, a wafer 12 is held on a table (wafer holder) 14, which
in turn is placed on a focusing and levelling stage 15 through three
fulcrums 16A to 16C movable in the direction of the z-axis. The focusing
and levelling stage 15 is placed on an XY stage 17 two-dimensionally
movable on a base 18. The focusing and levelling stage 15 effects the
adjustment of the position (focus position) of the wafer 12 on the wafer
holder 14 in Z-direction through the three fulcrums 16A to 16C and also
effects the adjustment of the angle of inclination of the wafer 12. The XY
stage 17 positions the focusing and levelling stage 15, the wafer holder
14 and the wafer 12 in x-direction and y-direction.
The x-coordinates of the wafer holder 14 are always monitored by a movable
mirror 19X fixed to the upper end of the wafer holder 14 and a laser
interferometer 20X disposed outside (for example, on the base 18), and as
shown in FIG. 1, the y-coordinates of the wafer holder 14 are always
monitored by a movable mirror 19Y and an outside laser interferometer 20Y,
and these detected x- and y-coordinates are supplied to the main control
system 10 and a plane position calculating system 33 which will be
described later.
Turning back to FIG. 4, the main control system 10 controls the operation
of the XY stage 17 through a drive system 21. Also, on the basis of
residual deviation (control deviation) calculated by the plane position
calculating system 33, a focusing and levelling control system 22 of the
so-called PID (proportion, integration, differentiation) control type
controls the amounts of movement of the three fulcrums 6A to 16C of the
focusing and levelling stage 15 in z-direction through amplifiers 23A to
23C.
Description will now be made of the construction of a multipoint focus
position detecting system (hereinafter referred to as the AF sensor) for
detecting the position (focus position) of the exposed surface (e.g. the
front surface) of the wafer 12 in z-direction. In the present embodiment,
the number of the measuring points of the focus position on the wafer 12
by the multipoint AF sensor is nine. Accordingly, in the apparatus of FIG.
4, nine AF sensors of the same construction disclosed, for example, in
U.S. Pat. Nos. 4,558,949 and 4,650,983 are disposed as the multipoint AF
sensor, but only three AF sensors 25A1, 25A2 and 25A3 on this side are
shown in FIG. 4. First, in the central AF sensor 25A2, detection light
non-photosensitive to photoresist which is emitted from a light source
26A2 illuminates a slit pattern on a light transmitting slit plate 27A2,
and the image of the slit pattern is projected onto the central (on the
optical axis AX) measuring point P2 of the shot area 13A on the wafer 12
obliquely to the optical axis AX of the projection optical system PL
through an objective lens 28A2. The reflected light from the measuring
point P2 is condensed on a vibration slit plate 30A2 through a condensing
lens 29A2, and the image of the slit pattern projected onto the measuring
point P2 is re-imaged on the vibration slit plate 30A2.
The light passed through a slit opening in the vibration slit plate 30A2 is
received by a photoelectric detector 31A2, and a photoelectrically
converted signal from the photoelectric detector 31A2 is supplied to an
amplifier 32A2. The amplifier 32A2 synchronously detects the
photoelectrically converted signal from the photoelectric detector 31A2 by
the driving signal of the vibration slit plate 30A2, and amplifies the
obtained signal to thereby produce a focus signal (S curve signal) varying
substantially linearly within a predetermined range relative to the focus
position of the surface of the wafer at the measuring point P2, and
supplies this focus signal to the plane position calculating system 33.
Likewise, the AF sensor 25A1 projects the image of the slit pattern onto a
measuring point P1 set on the -x-direction side (the left side as viewed
in the plane of the drawing sheet) relative to the measuring point P2, and
the reflected light from this measuring point P1 is received by a
photoelectric detector 31A1 and a photoelectrically converted signal from
the photoelectric detector 31A1 is supplied to an amplifier 32A1. The
amplifier 32A1 supplies the plane position calculating system 33 with a
focus signal corresponding to the focus position of the surface of the
wafer at the measuring point P1. Likewise, the AF sensor 25A3 projects the
image of the slit pattern onto a measuring point P3 set on the
+x-direction side (the right side as viewed in the plane of the drawing
sheet) relative to the measuring point P2, and the reflected light from
this measuring point P3 is received by a photoelectric detector 31A3 and a
photoelectrically converted signal from the photoelectric detector 31A3 is
supplied to an amplifier 32A3. The amplifier 32A3 supplies the plane
position calculating system 33 with a focus signal corresponding to the
focus position of the surface of the wafer at the measuring point P3.
Here, the calibration of the AF sensors 25A1 to 25A3 is effected so that
the intensities (voltage values) of the focus signals outputted from the
amplifiers 32A1 to 32A3 may become zero when at the measuring points P1 to
P3, the surface of the wafer is coincident with the imaging plane of the
projection optical system PL (the surface onto which the reticle pattern
is projected). Accordingly, the intensity of each focus signal corresponds
to the amount of deviation in Z-direction (the direction of the optical
axis AX) between the surface of the wafer and the imaging plane of the
projection optical system PL at the measuring points P1 to P3. Therefore,
the amount of deviation corresponding to the intensity of the focus signal
obtained by the ith AF sensor 25Ai (i=1, 2, . . . , 9) can be regarded as
a defocus amount .DELTA.f.sub.i.
FIG. 5 shows the distribution of the nine measuring points of the
multipoint AF sensor in FIG. 4 on the wafer. In FIG. 5, 3.times.3
measuring points P.sub.1 to P.sub.9 are set in the shot area 13A on the
wafer at a predetermined pitch in x-direction and y-direction, and the
image of the slit pattern is projected by the AF sensors 25Ai (only the AF
sensors 25A1 to 25A3 being shown in FIG. 4) corresponding to the
respective measuring points Pi. The defocus amount .DELTA.f.sub.i (focus
signal) of the surface of the wafer (shot area) at each measuring point Pi
measured by the AF sensor 25Ai is supplied to the plane position
calculating system 33 of FIG. 4.
Also, in the present embodiment, a weight coefficient Wi is allotted to the
defocus amount .DELTA.f.sub.i at each measuring point Pi measured by the
multipoint AF sensor. When for example, the line width of the pattern
image of the reticle transferred to the shot area 13A is substantially
equal everywhere, the values of the weight coefficients Wi may all be 1.
Also, when for example, a pattern image of the narrowest line width is
transferred to areas 34A and 34B surrounding the measuring points P.sub.4
and P.sub.6, respectively, the values of the weight coefficients W.sub.4
and W.sub.6 allotted to the defocus amounts at the measuring points
P.sub.4 and P.sub.6 may be set to values greater than 1, and the value of
the weight coefficient W.sub.5 allotted to the defocus amount at the
central measuring point P.sub.5 s may be set to a value smaller than 1. It
is to be understood that the values of the weight coefficients allotted to
the defocus amounts at the remaining six measuring points are set to e.g.
1. Thus, even if for example, the surface of the wafer along the line A--A
of FIG. 5 is wavy as exaggeratedly shown in FIG. 6, when the focusing
operation and levelling operation according to the present embodiment are
executed, the image plane of the projection optical system PL will be set
to a plane 46B relative near to the surfaces of both of the partial areas
34A and 34 | | |
