U.S. patent application number 09/324042 was filed with the patent office on 2002-05-09 for exposure apparatus and control method for correcting an exposure optical system on the basis of an estimated magnification variation.
Invention is credited to YONEKAWA, MASAMI.
Application Number | 20020053644 09/324042 |
Document ID | / |
Family ID | 15906896 |
Filed Date | 2002-05-09 |
United States Patent
Application |
20020053644 |
Kind Code |
A1 |
YONEKAWA, MASAMI |
May 9, 2002 |
EXPOSURE APPARATUS AND CONTROL METHOD FOR CORRECTING AN EXPOSURE
OPTICAL SYSTEM ON THE BASIS OF AN ESTIMATED MAGNIFICATION
VARIATION
Abstract
In an exposure apparatus for projecting illumination light
irradiating a reticle on a wafer via an optical system, a mark
magnification variation .DELTA..beta..sub.M is calculated from the
displacement amounts of a plurality of position measurement marks
used for reticle alignment. From the mark magnification variation
.DELTA..beta..sub.M a shot magnification .DELTA..beta..sub.S is
estimated using an estimation equation having an aspect ratio A and
area ratio S of the exposure region as parameters:
.DELTA..beta..sub.S=c.multidot.A.sup.p.multidot.S.sup.q.multidot..DELTA..b-
eta..sub.M (where c, p, q are coefficients) The shot magnification
is corrected based on the estimation result using the magnification
correction function of a projection lens.
Inventors: |
YONEKAWA, MASAMI;
(UTSUNOMIYA-SHI, JP) |
Correspondence
Address: |
FITZPATRICK CELLA HARPER & SCINTO
30 ROCKEFELLER PLAZA
NEW YORK
NY
10112
US
|
Family ID: |
15906896 |
Appl. No.: |
09/324042 |
Filed: |
June 2, 1999 |
Current U.S.
Class: |
250/492.1 |
Current CPC
Class: |
G03F 9/7003 20130101;
G03F 7/70875 20130101 |
Class at
Publication: |
250/492.1 |
International
Class: |
A61N 005/00 |
Foreign Application Data
Date |
Code |
Application Number |
Jun 4, 1998 |
JP |
10-170546 |
Claims
What is claimed is:
1. A control method for an exposure apparatus for projecting
illumination light irradiating a reticle on a member to be exposed
via an optical system, comprising: the holding step of holding a
parameter corresponding to a shape of an exposure region on the
reticle; the measurement step of measuring a displacement of a mark
formed on the reticle; the estimation step of estimating a
magnification variation in the exposure region on the basis of the
displacement of the mark measured in the measurement step and the
parameter; and the correction step of correcting the optical system
on the basis of the magnification variation estimated in the
estimation step.
2. The method according to claim 1, wherein the mark is a position
measurement mark used to align the reticle.
3. The method according to claim 1, wherein the exposure region has
a rectangular shape, and the parameter includes at least either one
of an area ratio of the exposure region to a predetermined
reference square and aspect ratio of the exposure region.
4. The method according to claim 3, wherein the parameter includes
the aspect ratio of the exposure region, and letting
.DELTA..beta..sub.M be a mark magnification variation amount
measured in the measurement step, A be the aspect ratio, and
c.sub.1 and p be coefficients, the estimation step comprises
estimating, as the estimated magnification variation in the
exposure region, .DELTA..beta..sub.S given by
.DELTA..beta..sub.S=c.s-
ub.1.multidot.A.sup.p.multidot..DELTA..beta..sub.M
5. The method according to claim 3, wherein the parameter includes
the area ratio, and letting .DELTA..beta..sub.M be a mark
magnification variation amount measured in the measurement step, S
be the area ratio, and c.sub.2 and q be coefficients, the
estimation step comprises estimating, as the estimated
magnification variation in the exposure region, .DELTA..beta..sub.S
given by .DELTA..beta..sub.S=c.sub.2.multidot-
.S.sup.q.multidot..DELTA..beta..sub.M
6. The method according to claim 3, wherein the parameter includes
the aspect ratio and the area ratio, and letting
.DELTA..beta..sub.M be a mark magnification variation amount
measured in the measurement step, A be the aspect ratio, S be the
area ratio, and c, p, and q be coefficients, the estimation step
comprises estimating, as the estimated magnification variation in
the exposure region, .DELTA..beta..sub.S given by
.DELTA..beta..sub.S=c.multidot.A.sup.p.multidot.S.sup.q.multidot..DELT-
A..beta..sub.M
7. The method according to claim 1, wherein the correction step
comprises performing correction using a magnification correction
function of the projection lens.
8. The method according to claim 1, wherein said method further
comprises the determination step of determining whether the
magnification variation estimated in the estimation step exceeds an
allowable value, and the correction step comprises correcting the
optical system on the basis of the estimated magnification
variation when the estimated magnification variation is determined
to exceed the allowable value.
9. The method according to claim 8, wherein the allowable value in
the determination step is set on the basis of a required exposure
precision.
10. The method according to claim 1, further comprising the control
step of executing the estimation step, determination step, and
correction step every predetermined exposure amount.
11. The method according to claim 10, wherein the predetermined
exposure amount is set based on the number of recording members as
a unit.
12. The method according to claim 10, wherein the predetermined
exposure amount is set based on the number of exposure shots to the
exposure region as a unit.
13. The method according to claim 8, further comprising the
alignment step of aligning the reticle on the basis of the
displacement of the mark when the optical system-is corrected in
the correction step.
14. The method according to claim 13, wherein the alignment step
comprises aligning the reticle so as to minimize the sum of squares
of the displacement of the mark.
15. An exposure apparatus for projecting illumination light
irradiating a reticle on a member to be exposed via an optical
system, comprising: holding means for holding a parameter
corresponding to a shape of an exposure region on the reticle;
measurement means for measuring a displacement of a mark formed on
the reticle; estimation means for estimating a magnification
variation in the exposure region on the basis of the displacement
of the mark measured by said measurement means and the parameter;
and correction means for correcting the optical system on the basis
of the magnification variation estimated by said estimation
means.
16. The apparatus according to claim 15, wherein the mark is a
position measurement mark used to align the reticle.
17. The apparatus according to claim 15, wherein the exposure
region has a rectangular shape, and the parameter includes at least
either one of an area ratio of the exposure region to a
predetermined reference square and aspect ratio of the exposure
region.
18. The apparatus according to claim 17, wherein the parameter
includes the aspect ratio of the exposure region, and letting
.DELTA..beta..sub.M be a mark magnification variation amount
measured by said measurement means, A be the aspect ratio, and
c.sub.1 and p be coefficients, said estimation means estimates, as
the estimated magnification variation in the exposure region,
.DELTA..beta..sub.S given by .DELTA..beta..sub.S=c.s-
ub.1.multidot.A.sup.p.multidot..DELTA..beta..sub.M
19. The apparatus according to claim 17, wherein the parameter
includes the area ratio, and letting .DELTA..beta..sub.M be a mark
magnification variation amount measured by said measurement means,
S be the area ratio, and c.sub.2 and q be coefficients, said
estimation means estimates, as the estimated magnification
variation in the exposure region, .DELTA..beta..sub.S given by
.DELTA..beta..sub.S=c.sub.2.multidot.S.sup.q-
.multidot..DELTA..beta..sub.M
20. The apparatus according to claim 17, wherein the parameter
includes the aspect ratio and area ratio of the exposure region,
and letting .DELTA..beta..sub.M be a mark magnification variation
amount measured by said measurement means, A be the aspect ratio, S
be the area ratio, and c, p, and q be coefficients, said estimation
means estimates, as the estimated magnification variation in the
exposure region, .DELTA..beta..sub.S given by
.DELTA..beta..sub.S=c.multidot.A.sup.p.multi-
dot.S.sup.q.multidot..DELTA..beta..sub.M
21. The apparatus according to claim 15, wherein said correction
means performs correction using a magnification correction function
of the projection lens.
22. The apparatus according to claim 15, wherein said apparatus
further comprises determination means of determining whether the
magnification variation estimated by said estimation means exceeds
an allowable value, and said correction means corrects the optical
system on the basis of the estimated magnification variation when
the estimated magnification variation is determined to exceed the
allowable value.
23. The apparatus according to claim 22, wherein the allowable
value in said determination means is set on the basis of a required
exposure precision.
24. The apparatus according to claim 15, further comprising control
means for executing said estimation means, determination means, and
correction means every predetermined exposure amount.
25. The apparatus according to claim 24, wherein the predetermined
exposure amount is set based on the number of recording members as
a unit.
26. The apparatus according to claim 24, wherein the predetermined
exposure amount is set based on the number of exposure shots to the
exposure region as a unit.
27. The apparatus according to claim 22, further comprising
alignment means for aligning the reticle on the basis of the
displacement of the mark when the optical system is corrected by
said correction means.
28. The apparatus according to claim 27, wherein said alignment
means aligns the reticle so as to minimize the sum of squares of
the displacement of the mark.
Description
BACKGROUND OF THE INVENTION
[0001] 1. Field of the Invention
[0002] The present invention relates to an exposure apparatus for
sequentially projecting an electronic circuit pattern formed on a
reticle surface onto respective shots on a wafer surface via a
projection optical system in manufacturing a semiconductor element
such as an IC or LSI and, more particularly, to a semiconductor
manufacturing exposure apparatus having a function of, even if the
reticle absorbs exposure light to thermally expand during
projection exposure, immediately detecting this, converting it into
a shot magnification component, and quickly correcting the shot
magnification component using a magnification correction function
of a projection lens.
[0003] 2. Description of the Related Art
[0004] Recently, as semiconductor integrated circuit patterns such
as IC and LSI patterns continue to shrink in feature size, the
projection exposure apparatus is demanded for high resolution, high
overlay accuracy, and high throughput.
[0005] At present, the mass production line of each LSI
manufacturer tends to use an exposure apparatus having high
resolution for a critical layer and an exposure apparatus having
low resolution but high throughput for an uncritical layer in order
to increase COO (Cost Of Ownership). To cope with a process of the
Mix & Match method using different apparatuses, variations in
magnification and distortion in the shot must be suppressed in
addition to shifts in the shot matrix, magnification errors, and
rotation errors on the wafer.
[0006] In particular, as for variations in the shot, variations in
magnification and distortion by thermal expansion of the reticle
upon absorbing illumination light have recently surfaced. Since the
reticle is made of silica glass, the glass itself has an absorption
index of several % or less for exposure light (KrF 243 nm, i-line,
g-line, and the like) and a low linear expansion coefficient of 0.5
ppm/.degree. C., and thus thermal expansion of a conventional
reticle does not pose any problem. However, the Cr pattern absorbs
a large amount of exposure energy because of high luminance of an
illumination lamp for increasing throughput or a three-layered Cr
surface for preventing flare of an optical system. The silica glass
rises in temperature during a heat conduction process to thermally
expand.
[0007] To avoid this phenomenon, the temperature rise is avoided by
spraying conditioned air on the reticle. However, this is not
practical because the temperatures of the air and reticle cannot be
made equal, the apparatus becomes bulky, and the effects are low
for a reticle formed with a pellicle.
[0008] There is proposed another method (e.g., Japanese Patent
Laid-Open No. 4-192317) of estimating the thermal deformation
amount of the reticle by numerical calculation such as the
difference calculus or finite-element method on the basis of
various exposure parameters (reticle surface illuminance, pattern
density, and the like), and correcting the magnification or
distortion component using the correction means of a projection
optical system. However, this method is basically open-loop
correction. The thermal deformation amount of the reticle is
difficult to estimate by numerical calculation for the use of
various modified illuminations, phase shift reticles,
pellicle-formed reticles, or the like in an actual process, and the
use of a combination of them.
[0009] Accordingly, a so-called closed-loop correction method of
directly measuring and correcting thermal deformation of the
reticle is adopted. That is, thermal deformation of the reticle is
measured using an existing position measurement mark used to align
the reticle by a reticle stage, and is corrected based on the
result using the distortion correction means of the projection
optical system.
[0010] This method will be briefly explained with reference to
FIGS. 3 to 5. In FIG. 3, reference numeral 1 denotes a reticle; and
2 and 3, position measurement marks formed on the lower surface of
the reticle. In FIG. 4, reference numeral 4 denotes a base which
supports a reticle stage (not shown); and 5 and 6, reference marks
formed on the base at positions where they face the reticle
position measurement marks 2 and 3. In measuring thermal
deformation of the reticle, shifts (.DELTA.x, .DELTA.y) of the
position measurement marks 2 and 3 are measured with reference to
the reference marks 5 and 6, as shown in FIG. 5. From this result,
the magnification component generated on the reticle is corrected
using the distortion correction means of the projection optical
system.
[0011] This method is, however, unsatisfactory.
[0012] A process for an actual element does not always use the
maximum image field (e.g., 22 mm.times.22 mm) of the exposure
apparatus. The image field is often limited by a masking device to
a rectangular shape such as two or three chips by one shot
depending on the chip size. In this case, according to the above
method, since the edge of an actual shot is apart from the position
measurement mark, the magnification variation in the shot is
different from the magnification variation of the position
measurement mark.
[0013] FIG. 6 shows this state. In FIG. 6, reference numeral 7
denotes a rectangular actual element area. FIG. 7 schematically
shows the temperature distribution and deformation when exposure
light is incident on this area to attain a thermally steady state.
In FIG. 7, reference numeral 1a denotes a deformed reticle; and 2a
and 3a, displaced position measurement marks. If the circuit
pattern region has a rectangular shape long in the y direction,
like this example, the temperature distribution has an elliptical
shape long in the y direction, and thermal deformation along with
this also becomes prominent in the y direction. As for the shot
magnification variation, the y-direction magnification component
greatly changes. Despite of this, since the reticle position
measurement marks 2a and 3a are positioned near the edge of the
reticle, the magnification component calculated from the mark
displacement does not reflect an actual shot magnification.
[0014] When a circuit pattern region 8 is much smaller than the
maximum illumination region, as shown in FIG. 8, the temperature
distribution has a small concentric shape, as shown in FIG. 9. For
the same reason, the magnification component calculated from the
mark displacement does not reflect an actual shot magnification,
either.
[0015] As described above, the conventional method cannot
accurately monitor the shot magnification variation using the mark
magnification variation.
SUMMARY OF THE INVENTION
[0016] The present invention has been made to solve the above
problems, and has as its object to provide an exposure apparatus
and exposure control method capable of accurately estimating
magnification variations in an exposure region regardless of
changes in shape of the exposure region.
[0017] More specifically, when the shot magnification component
generated by thermal deformation of a reticle is to be calculated
from the displacement of a position measurement mark on an existing
reticle, the shot magnification component can be accurately
estimated even if the exposure region changes to an arbitrary size
by a masking device.
[0018] It is another object of the present invention to correct the
shot magnification component generated on the reticle from the
estimation result using the magnification correction means of a
projection lens.
[0019] It is still another object of the present invention to
accurately, easily estimate magnification variations in an exposure
region having an arbitrary shape by using either one of the aspect
ratio and area of the exposure region as shape information of the
exposure region.
[0020] It is still another object of the present invention to
increase the exposure precision by correcting the magnification by
the projection lens on the basis of the estimated magnification
variation.
[0021] It is still another object of the present invention to
execute magnification correction and minimize a decrease in
throughput by execution of magnification correction when the
estimated magnification variation reaches a level that poses
problems in a required exposure precision.
[0022] It is still another object of the present invention to
execute estimation of the magnification variation every
predetermined exposure amount in units of wafers or shots, thereby
minimizing a decrease in throughput by execution of magnification
correction.
[0023] Other features and advantages of the present invention will
be apparent from the following description taken in conjunction
with the accompanying drawings, in which like reference characters
designate the same or similar parts throughout the figures
thereof.
BRIEF DESCRIPTION OF THE DRAWINGS
[0024] The accompanying drawings, which are incorporated in and
constitute a part of the specification, illustrate embodiments of
the invention and, together with the description, serve to explain
the principles of the invention.
[0025] FIG. 1 is a flow chart showing exposure and reticle shot
magnification correction sequences according to the first
embodiment of the present invention;
[0026] FIG. 2 is a flow chart showing exposure and reticle shot
magnification correction sequences according to the second
embodiment of the present invention;
[0027] FIG. 3 is a plan view showing a reticle and reticle position
measurement mark;
[0028] FIG. 4 is a plan view showing a reticle stage base and
reference mark;
[0029] FIG. 5 is a sectional view showing the positional
relationship between the reticle and reticle stage base;
[0030] FIG. 6 is a plan view showing a reticle having a rectangular
circuit pattern region;
[0031] FIG. 7 is a plan view showing the thermal deformation and
temperature distribution of the reticle in FIG. 6;
[0032] FIG. 8 is a plan view showing a reticle having a small
square circuit pattern region;
[0033] FIG. 9 is a plan view showing the thermal deformation and
temperature distribution of the reticle in FIG. 8;
[0034] FIG. 10 is a schematic view showing the concept of an
apparatus to which the present invention is applied;
[0035] FIG. 11 is a graph showing a transient temperature rise of a
reticle having a high Cr pattern density;
[0036] FIG. 12 is a graph showing a transient temperature rise of a
reticle having a low Cr pattern density;
[0037] FIG. 13 is a plan view showing a masking size for
calculating a shot magnification estimation equation;
[0038] FIG. 14 is a plan view showing another masking size for
calculating the shot magnification estimation equation;
[0039] FIG. 15 is a flow chart showing a flow of manufacturing a
microdevice; and
[0040] FIG. 16 is a flow chart showing a detailed flow of the wafer
process.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0041] Preferred embodiments of the present invention will now be
described in detail in accordance with the accompanying
drawings.
[0042] According to a preferred embodiment of the present
invention, a shift between an existing reticle position measurement
mark used for reticle alignment and a paired reference mark is
always measured to calculate a mark magnification variation
.DELTA..beta..sub.M. An estimation equation for estimating an
actual shot magnification variation .DELTA..beta..sub.S from the
magnification variation .DELTA..beta..sub.M of the position
measurement mark is stored in a console. The shot magnification
variation .DELTA..beta..sub.S is always monitored by calculation
using the estimation equation. Since this estimation equation has
the aspect ratio and exposure area of the shot as parameters, an
actual shot magnification can be estimated with high precision even
for an arbitrary image field by a masking device.
[0043] If the shot magnification is determined to reach a level
that poses problems in an actual exposure process, the shot
magnification component generated on the reticle is corrected by a
projection lens using the magnification correction function of a
known projection lens, and then exposure operation restarts.
Subsequently, the shot magnification is estimated from the
measurement result of the reticle position measurement means to
always monitor the shot magnification. If the shot magnification is
determined to reach this level, the same operation is repeated.
[0044] With this processing, even when the reticle absorbs
illumination light and thermally expands, the thermal deformation
amount is measured using an existing reticle position measurement
means. The result is used in the estimation equation for estimating
the shot magnification. Consequently, the magnification can be
quickly corrected by the projection lens to always print a
high-quality circuit pattern on the wafer.
[0045] Prior to a description of the principal part of the present
invention, the arrangement of a projection exposure apparatus to
which the present invention is applied will be briefly described by
exemplifying a stepper with reference to FIG. 10.
[0046] In FIG. 10, reference numeral 50 denotes an illumination
system; 1, a reticle on which a circuit pattern is formed and
position measurement marks 2 and 3 are drawn; and 4, a base which
mounts a reticle stage (not shown) and has reference marks 5 and 6
at positions where they face the marks 2 and 3. The reticle 1 is
aligned with reference to the reticle stage base 4. The measurement
value of alignment is stored in a console unit 61 for controlling
the whole apparatus.
[0047] Reference numeral 51 denotes a projection lens for
projecting a reticle pattern image on a wafer 54; 52, a known lens
drive unit for correcting changes in imaging performance by air
pressure and temperature; 62, a mirror for deflecting probe light
for measuring wafer alignment by an off-axis TTL (Through The Lens)
method; and 59, a measurement system for the mirror 62.
[0048] Reference numerals 53 and 60 denote known focus/wafer tilt
detectors which irradiate the surface of the wafer 54 with a light
beam (53), and photoelectrically detect the reflected light (60),
thereby detecting the in-focus position of the projection lens 51
and the tilt of the wafer 54.
[0049] Reference numeral 55 denotes a wafer chuck for
vacuum-chucking the wafer 54; 56, a wafer stage capable of moving
coarsely and finely in the x, y, and 0 directions. The position of
the wafer stage 56 is always monitored by an interferometer mirror
57 and interferometer 58. The whole exposure apparatus including
these main units is controlled by the console unit 61.
[0050] A shot magnification correction method according to
embodiments of the present invention will be described.
[0051] As described above with reference to FIGS. 3 to 5, the
reticle position measurement mark and reference mark important in
the present invention will be explained. As is apparent from FIGS.
3 to 5, thermal deformation of the reticle is obtained by measuring
a change in distance between the reticle position measurement marks
2 and 3 relative to the distance between the reference marks 5 and
6. Note that the x and y coordinates of the positions of the
reticle position measurement marks 2 and 3 can be detected, and can
be expressed by (x.sup.(i).sub.R, y.sup.(i).sub.R) and
(x.sup.(i).sub.L, y.sup.(i).sub.L) where i is the number of
measurement operations.
[0052] Prior to exposure, the reticle 1 having the two position
measurement marks is aligned with reference to the reference marks
of the reticle stage base 4. In this state, exposure starts. During
intermittent exposure, the Cr pattern of the reticle absorbs
exposure light to rise in temperature and thermally deform. As
described above, since silica glass used as the base material of
the reticle has a low linear expansion coefficient of 0.5 ppm, when
the density (area ratio) of the Cr pattern is low, thermal
expansion of the reticle does not pose any problem. However, when
the density (area ratio) of the Cr pattern is high, the pattern
absorbs a large amount of exposure energy, and the silica glass
rises in temperature during a heat conduction process to thermally
expand.
[0053] FIGS. 11 and 12 show the reticle temperature plotted along
the time on the abscissa. FIGS. 11 and 12 respectively show the
reticle temperature when the Cr pattern density is high and low.
The temperature rise of the reticle traces a normal unsteady heat
conduction process. That is, when exposure starts at .tau.=0, the
temperature gradually rises to trace the curve of an error function
erf(.xi.). At given time .tau.=.tau..sub.0, the temperature reaches
a steady state at the maximum temperature T=T.sub.0. Exposure is
stopped at .tau.=.tau..sub.1, and then the temperature traces an
inverse curve to the rise curve to fall and reach the steady state
again. What is important in this process is an unsteady-state
section (0.ltoreq..tau..ltoreq..tau..sub.0) from the start of
exposure to the steady state. According to the present invention,
thermal expansion of the reticle is appropriately detected in this
section and converted into a shot magnification, and this shot
magnification is efficiently corrected using the
magnification-distortion correction function of the projection
lens.
[0054] The present invention attaches importance to the step of
estimating the shot magnification component generated by thermal
deformation of the reticle from the magnification component
calculated based on the displacement of an existing reticle
position measurement mark, i.e., an estimation equation calculation
method, which will be explained.
[0055] In an actual semiconductor process, the exposure region on
the reticle changes depending on a designed circuit pattern. If the
exposure region is unchanged, the total energy amount absorbed by
the reticle and the thermal deformation mount can be considered to
be proportional to each other. Therefore, assume that the total
absorption energy amount is constant.
[0056] The relationship between the aspect ratio of the exposure
region, mark magnification variation, and shot magnification
variation is examined by an experiment or simulation. As shown in
FIG. 13, the aspect ratio of the exposure region is arbitrarily
changed from 70 to 76 with a constant exposure area to obtain mark
magnification variations and actual shot magnification variations
at respective aspect ratios. More specifically, the mark
magnification variation is calculated from displacements
(.DELTA.x.sup.(i).sub.R, .DELTA.y.sup.(i).sub.R) and
(.DELTA.x.sup.(i).sub.L, .DELTA.y.sub.(i).sub.L) of the right and
left reticle position measurement marks a sufficient time after the
start of exposure (L is the distance between the marks):
.DELTA..beta..sup.(i).sub.M=(.DELTA.x.sup.(i).sub.R-.DELTA.x.sup.(i).sub.L-
)/L
[0057] The shot magnification variation .DELTA..beta..sup.(i).sub.S
at this time is calculated. The relationship between the mark and
shot magnification variations .DELTA..beta..sup.(i).sub.M and
.DELTA..beta..sup.(i).sub.S is examined at several aspect ratios
while the shot area is kept unchanged. The simulation results by
the present inventor reveal that, letting an aspect ratio A be
(breadth (x) size)/(length (y) size) of the shot, as the aspect
ratio increases, the mark magnification variation
.DELTA..beta..sup.(i).sub.M comes close to the shot magnification
variation .DELTA..beta..sup.(i).sub.S. Hence, when the aspect ratio
A is used as a parameter, the shot magnification can be given using
a power p of the aspect ratio A as a parameter:
.DELTA..beta..sup.(i).sub.S=c.sub.1.multidot.A.sup.p.multidot..DELTA..beta-
..sup.(i).sub.M (1)
[0058] where c.sub.1 is a coefficient.
[0059] The relationship between the shot area, mark magnification
variation, and shot magnification variation is examined by an
experiment or simulation. As shown in FIG. 14, the exposure region
area is arbitrarily changed from 80 to 84 with a constant aspect
ratio (=1) to obtain mark magnification variations and actual shot
magnification variations at respective areas. More specifically,
the mark magnification variation is calculated from displacements
(.DELTA.x.sup.(j).sub.R, .DELTA.y.sup.(j).sub.R) and
(.DELTA.x.sup.(j).sub.L, .DELTA.y.sup.(j).sub.L) of the right and
left reticle position measurement marks a sufficient time after the
start of exposure in the jth exposure region on the reticle (L is
the distance between the marks):
.DELTA..beta..sup.(j).sub.M=(.DELTA.x.sup.(j).sub.R-.DELTA.x.sup.(j).sub.L-
)/L
[0060] A shot magnification variation .DELTA..beta..sup.(j).sub.S
at this time is calculated. The relationship between the mark and
shot magnification variations .DELTA..beta..sup.(j).sub.M and
.DELTA..beta..sup.(j).sub.S is examined at several exposure areas
while the aspect ratio is kept unchanged. The simulation results by
the present inventor reveal that, letting an exposure area ratio S
be shot area/reference area (reference area: the area of a square
region at the mark position), as the area ratio increases to 1, the
shot edge and mark position come close to each other. When the
exposure area ratio S is used as a parameter, the shot
magnification can be given using a power q of the exposure area
ratio S as a parameter:
.DELTA..beta..sup.(j).sub.S=c.sub.2.multidot.S.sup.1.multidot..DELTA..beta-
..sup.(j).sub.M (2)
[0061] where C.sub.2 is a coefficient.
[0062] The shot magnification variation can be given by a
combination of equations (1) and (2) at an arbitrary aspect ratio A
and exposure area ratio S:
.DELTA..beta..sub.S=c.multidot.A.sup.p.multidot.S.sup.1.multidot..DELTA..b-
eta..sub.M (3)
[0063] In practice, the coefficients c, p, and q are calculated by
an experiment and simulation using the least squares method on the
basis of this model equation, thereby determining the estimation
equation.
[0064] A method of correcting a thermally expanded reticle by the
magnification correction means of the projection lens will be
explained. At certain time after the start of exposure, if the mark
magnification component .DELTA..beta..sub.M calculated from shifts
(.DELTA.x.sub.R, .DELTA.y.sub.R) and (.DELTA.x.sub.L,
.DELTA.y.sub.L) from the initial coordinates of the respective
reticle position measurement marks is generated, the shot
magnification .DELTA..beta..sub.S is estimated using equation (3).
In an actual sequence, the magnification component
.DELTA..beta..sub.S generated on the reticle 1 is controlled by an
open loop of calculating a lens drive amount corresponding to the
magnification component by the console 61 shown in FIG. 10 and
inputting the command value to the magnification correction unit 52
to correct the magnification component.
[0065] The above description concerns the concept of the method of
correcting the shot magnification generated by thermal deformation
of the reticle according to the present invention.
[0066] Detailed embodiments of the present invention will be
described below with reference to the accompanying drawings.
[0067] First Embodiment
[0068] An exposure sequence according to the present invention will
be explained with reference to FIG. 1.
[0069] The sequence starts (step 10), and a reticle 1 is loaded
onto the reticle stage and chucked (step 11) The reticle 1 is
aligned using two reticle position measurement marks (2, 3) with
reference to reference marks (5, 6) formed on a reticle stage base
4. The measurement coordinates of the reticle position measurement
marks (2, 3) at this time are set to (.DELTA.x.sup.(0).sub.R,
.DELTA.y.sup.(0).sub.R) and (.DELTA.x.sup.(0).sub.L,
.DELTA.y.sup.(0).sub.L), which are used as initial measurement
values (step 12). The first wafer 54 is conveyed from a convey
system onto a wafer chuck 55, chucked by the wafer chuck 55, and
aligned (step 13). The first wafer 54 is exposed (step 14).
[0070] In loading the reticle in step 11, parameters (aspect ratio
A and area ratio S) representing the shape of the shot region of
this reticle are set in a console 61. These parameters may be set
manually or automatically by making the parameters on the reticle
and reading them by the apparatus.
[0071] After exposure for all shots is complete, the wafer is
replaced. During this replacement, the positions of the position
measurement marks (2, 3) of the reticle 1 are measured to calculate
shifts ((.DELTA.x.sup.(1).sub.R, .DELTA.y.sup.(1).sub.R) and
(.DELTA.x.sup.(1).sub.L, .DELTA.y.sup.(1).sub.L)) from the initial
mark positions of the reticle position measurement marks. Based on
these shifts, the mark magnification variation
.DELTA..beta..sup.1).sub.M is calculated (step 15). From this
magnification variation, the shot magnification variation is
estimated using equation (3), i.e.,
.DELTA..beta..sub.S=c.multidot.A.sup.p.multidot.S.sup.1.multidot..DELTA..b-
eta..sub.M (3)
[0072] (step 16). This shot magnification variation
.DELTA..beta..sup.(1).sub.S is compared with a shot magnification
variation allowable value .DELTA..beta..sub.S0 calculated in
advance from the line width and overlay accuracy of a target
process (step 17).
[0073] If the calculated shot magnification variation
.DELTA..beta..sup.(1).sub.S is equal to or larger than the shot
magnification variation allowable value .DELTA..beta..sub.S0, the
exposure sequence shifts to a reticle magnification correction
sequence (steps 18 and 19). If .DELTA..beta..sup.(1).sub.S is
smaller than .DELTA..beta..sub.S0, exposure operation is continued
until exposure of all wafers is complete (step 20). At this time,
the reticle immediately after the start of exposure hardly
thermally deforms, exposure operation is determined to be
continued, and the sequence returns to step 13 to process the
second wafer. The execution interval of a series of shot
magnification variation estimation processes (steps 15 to 17) may
change depending on the process. When the pattern density of the
reticle is high, and the exposure energy and duty are large, the
estimation processes are done every wafer. When the pattern density
of the reticle is low, and the exposure energy and duty are small,
the estimation processes may be done in units of 5 or 10 wafers. In
the first embodiment, the estimation processes are done every wafer
in order to always monitor thermal deformation of the reticle.
[0074] Note that the unit number of wafers for which the estimation
processes are executed may be manually set in the console 61.
[0075] As this loop (i.e., exposure operation for the wafer) is
repeated, the reticle temperature gradually rises. If the shot
magnification variation .DELTA..beta..sup.(i).sub.S estimated in
step 16 is determined by the ith mark measurement to be larger than
the shot magnification variation allowable value
.DELTA..beta..sub.s0, the sequence advances from step 17 to step 18
to shift to magnification correction operation. In step 18, reticle
alignment is performed for an expanded reticle, errors are assigned
to the respective marks so as to minimize the sum of squares of
errors caused by the expansion, and the reticle is aligned in the
x, y, and .theta. directions by the reticle stage. A magnification
correction unit 52 of the projection lens operates in accordance
with a magnification correction value of the projection lens
corresponding to the generated shot magnification variation
.DELTA..beta..sup.(i).sub.S, thereby correcting the magnification
of the lens (step 19).
[0076] If the magnification is ideally corrected, no reticle
magnification variation is generated on the wafer, and the
alignment measurement values of the reticle marks at this time are
newly set as initial mark positions. The initial mark positions are
used to measure the respective mark positions of the reticle at the
restart of exposure. The sequence returns to the main sequence, a
new wafer is loaded onto the stage and aligned (step 13), and the
sequence of the present invention restarts. In this fashion, every
time the estimated shot magnification variation
.DELTA..beta..sup.(i).sub.S exceeds the shot magnification
variation allowable value .DELTA..beta..sub.S0, the reticle
magnification is corrected. If all wafers to be processed are
exposed, this process ends (step 21).
[0077] The first embodiment of the present invention has been
described. An application of the present invention to an actual
process will be explained. The present invention is applied until
the reticle reaches a thermally steady state. For example, when the
pattern density of the reticle is high, and the exposure energy and
duty are large, the reticle temperature also greatly changes, as
shown in FIG. 11, and reticle magnification correction operation is
executed at times .tau.2, .tau.3, .tau.4, and .tau.5. That is,
correction is done at a short correction interval immediately after
the start of exposure because the reticle temperature abruptly
rises, and at a long correction interval very close to the steady
state because the temperature hardly changes. When the pattern
density of the reticle is low, and the exposure energy and duty are
small, the reticle temperature hardly changes, as shown in FIG. 12,
and thus reticle magnification correction operation is done only
once at time .tau.6.
[0078] Second Embodiment
[0079] The second embodiment of the present invention will be
described.
[0080] This embodiment is characterized by a sequence of correcting
the shot magnification variation on the reticle in units of shots.
This will be explained with reference to FIG. 2.
[0081] Steps 30 to 33 are the same as steps 10 to 13 in the first
embodiment. A sequence of monitoring the shot magnification
variation on the reticle in units of shots, which is a feature of
the second embodiment, will be explained. In step 34, the first
shot of the first wafer is exposed. While the wafer stage moves to
the second shot and is aligned at a predetermined position, shifts
((.DELTA.x.sup.(1).sub.R, .DELTA.y.sup.(1).sub.R) and
(.DELTA.x.sup.(1).sub.L, .DELTA.y.sup.(1).sub.L)) from the initial
mark positions of position measurement marks (2, 3) of a reticle 1
are measured. Based on these measurement values, the mark
magnification variation .DELTA..beta..sup.(1).sub.S is calculated.
From this magnification variation, the shot magnification variation
is estimated using equation (3), (step 16). The estimated shot
magnification variation .DELTA..beta..sup.(1).sub.S is compared
with a shot magnification variation allowable value
.DELTA..beta..sub.S0 calculated in advance from the line width and
overlay accuracy of a target process (step 37).
[0082] If .DELTA..beta..sup.(1).sub.S is equal to or larger than
.DELTA..beta..sub.S0 the sequence shifts to a reticle magnification
correction sequence (steps 38, 39, and 40). If
.DELTA..beta..sup.(1).sub.- S is smaller than .DELTA..beta..sub.S0,
exposure operation is continued until exposure of all shots on all
wafers is complete (step 41). At this time, the reticle immediately
after completion of exposure of the first shot hardly thermally
deforms, exposure operation is determined to be continued, and the
sequence returns to step 34 to perform exposure for the second
shot. The execution interval of a series of shot magnification
variation estimation processes in steps 35 to 37 may change
depending on the process. When the pattern density of the reticle
is high, and the exposure energy and duty are large, the estimation
processes are done every shot exposure. When the pattern density of
the reticle is low, and the exposure energy and duty are small, the
estimation processes may be done in units of 10 or 20 shots. In the
second embodiment, the estimation processes are done every shot
exposure in order to always monitor thermal deformation of the
reticle every shot.
[0083] Note that the unit number of exposure shots for which the
estimation processes are executed may be manually set in a console
61.
[0084] As this loop (i.e., exposure operation for each shot region)
is repeated, the reticle temperature gradually rises. If the
estimated shot magnification variation .DELTA..beta..sup.(i).sub.S
is determined by the ith mark measurement to be larger than the
shot magnification variation allowable value .DELTA..beta..sub.S0,
the sequence shifts to magnification correction operation (steps 38
to 40). In step 38, exposure operation is paused, reticle alignment
is performed for an expanded reticle, errors are assigned to the
respective marks so as to minimize the sum of squares of errors
caused by the expansion, and the reticle is aligned in the x, y,
and .theta. directions by the reticle stage (step 39).
[0085] A magnification correction unit 52 of the projection lens
operates in accordance with a magnification correction value of the
projection lens corresponding to the generated shot magnification
variation .DELTA..beta..sup.(i).sub.S, thereby correcting the
magnification of the lens (step 40).
[0086] If the magnification is ideally corrected, no reticle
magnification variation is generated on the wafer, and the
alignment measurement values of the reticle marks at this time are
newly set as initial mark positions. The initial mark positions are
used to measure the respective mark positions of the reticle at the
restart of exposure. The sequence returns to the main sequence, and
the wafer stage is aligned at a next shot position to restart the
exposure sequence. In this way, every time the estimated shot
magnification variation .DELTA..beta..sup.(i).sub.S exceeds the
shot magnification variation allowable value .DELTA..beta..sub.S0,
the reticle magnification is corrected. If all shots are processed,
a new wafer is conveyed onto the wafer stage and exposed again.
When the shot magnification variation is monitored for each shot,
correction operation is repeated as needed, and all wafers have
been exposed, this process ends (step 43).
[0087] The second embodiment is especially suitable for a process
suffering a large shot magnification variation of the reticle
because the shot magnification variation generated on the reticle
is monitored every shot, when the shot magnification variation
exceeds the variation allowable value, the magnification of the
projection lens is corrected, and thus the transmittance of a
reticle pattern such as a contact hole is very low.
[0088] Note that the above embodiments estimate the shot
magnification variation using equation (3) in order to cope with
changes in both the aspect ratio and area of the exposure shot
region. Alternatively, the shot magnification variation may be
estimated using equation (1) if the area of the exposure shot
region is constant in the process, or equation (2) if the aspect
ratio is constant.
[0089] The above embodiments have exemplified the stepper. However,
the present invention is not limited to the stepper and can also be
applied to any apparatus so long as it uses the reference mark in
reticle alignment, such as a scanner disclosed in Japanese Patent
Laid-Open No. 09-246168 in which the reference mark is formed on
the reticle stage.
[0090] (Embodiment of Device Manufacturing Method)
[0091] An embodiment of a device manufacturing method using the
above-mentioned exposure apparatus or exposure method will be
described.
[0092] FIG. 15 shows a flow of manufacturing a microdevice
(semiconductor chip such as an IC or LSI, liquid crystal panel,
CCD, thin-film magnetic head, micromachine, or the like). In step
101 (circuit design), a device pattern is designed. In step 102
(mask manufacturing), a mask having the designed pattern is
manufactured. In step 103 (wafer manufacturing), a wafer is
manufactured using a material such as silicon or glass. In step 104
(wafer process) called a pre-process, an actual circuit is formed
on the wafer using the prepared mask and wafer by lithography. In
step 105 (assembly step) called a post-step, the wafer manufactured
in step 104 is formed into a semiconductor chip. The post-step
includes an assembly step (dicing and boding), packaging step (chip
sealing), and the like. In step 106 (inspection), inspections such
as an operation confirmation test and durability test are done for
the semiconductor device formed in step 105. The semiconductor
device is complete through these steps and shipped (step 107).
[0093] FIG. 16 shows a detailed flow of the wafer process. In step
111 (oxidization), the wafer surface is oxidized. In step 112
(CVD), an insulating film is formed on the wafer surface. In step
113 (electrode formation), an electrode is formed on the wafer by
deposition. In step 114 (ion implantation), ions are implanted in
the wafer. In step 115 (resist processing), the wafer is coated
with a photosensitive agent. In step 116 (exposure), the circuit
pattern of a mask is printed and exposed on the wafer by the
above-described exposure apparatus having the alignment device. In
step 117 (developing), the exposed wafer is developed. In step 118
(etching), the portion of the pattern except for the developed
resist image is etched away. In step 119 (resist removal), the
unnecessary resist after etching is removed. These steps are
repeatedly executed to form multiple circuit patterns on the
wafer.
[0094] By using the manufacturing method of this embodiment, a
high-integration-degree device, which is difficult to manufacture
by the conventional method, can be manufactured at low cost.
[0095] As has been described in detail, according to the present
invention, even if the reticle is irradiated with exposure light
and thermally deforms to influence the shot overlay accuracy, the
mark magnification variation is calculated from the displacement
amount of the reticle position measurement mark using the
estimation equation by the present inventor, thereby estimating the
shot magnification variation. The shot magnification variation can
always be monitored to quickly correct the shot magnification on
the reticle using the magnification correction function of the
projection-lens.
[0096] In the present invention, since the displacement amount of
the reticle is directly measured by the reticle position
measurement mark, the shot magnification on the reticle can be
estimated without any complicated process using a modified
illumination, phase shift register, and the like. When the shot
magnification variation of the reticle is determined to exceed the
allowable value, the reticle is realigned, and a thermal
deformation error is assigned to each mark. Therefore, even if
asymmetrical thermal deformation of the reticle occurs, the
magnification correction function of the projection lens can be
effectively used.
[0097] As a result, a high-quality circuit pattern can always be
printed on the wafer without depending on the shot magnification
variation along with thermal deformation of the reticle. The
present invention can be relatively easily practiced only by
rewriting software without modifying the apparatus because an
existing reticle position measurement mark and the magnification
correction function of a known projection lens are used.
[0098] As many apparently widely different embodiments of the
present invention can be made without departing from the spirit and
scope thereof, it is to be understood that the invention is not
limited to the specific embodiments thereof except as defined in
the claims.
* * * * *