U.S. patent application number 10/807603 was filed with the patent office on 2004-10-21 for metrology system using optical phase.
This patent application is currently assigned to KLA-Tencor Corporation. Invention is credited to Nikoonahad, Mehrdad, Smith, Ian, Vaez-Iravani, Mehdi, Zhao, Guoheng.
Application Number | 20040207849 10/807603 |
Document ID | / |
Family ID | 31978959 |
Filed Date | 2004-10-21 |
United States Patent
Application |
20040207849 |
Kind Code |
A1 |
Nikoonahad, Mehrdad ; et
al. |
October 21, 2004 |
Metrology system using optical phase
Abstract
Misalignment error between two periodic structures such as two
overlay targets placed side-by-side is measured. The two structures
are illuminated by coherent radiation and the positive and negative
diffraction beams of the input beam by the two structures are
detected to discover the optical phase difference between the
positive and negative diffraction beams. The misalignment between
the two structures may then be ascertained from the phase
difference.
Inventors: |
Nikoonahad, Mehrdad; (Menlo
Park, CA) ; Zhao, Guoheng; (Milpitas, CA) ;
Smith, Ian; (Los Gatos, CA) ; Vaez-Iravani,
Mehdi; (Los Gatos, CA) |
Correspondence
Address: |
PARSONS HSUE & DE RUNTZ LLP
655 MONTGOMERY STREET
SUITE 1800
SAN FRANCISCO
CA
94111
US
|
Assignee: |
KLA-Tencor Corporation
|
Family ID: |
31978959 |
Appl. No.: |
10/807603 |
Filed: |
March 23, 2004 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10807603 |
Mar 23, 2004 |
|
|
|
09639495 |
Aug 14, 2000 |
|
|
|
6710876 |
|
|
|
|
Current U.S.
Class: |
356/401 |
Current CPC
Class: |
G03F 7/70633
20130101 |
Class at
Publication: |
356/401 |
International
Class: |
G01B 011/00 |
Claims
What is claimed is:
1. A method for detecting misalignment of two periodic structures
placed next to each other so that they are periodic substantially
along a first line, comprising: illuminating a portion of each of
the two structures using radiation that is substantially coherent,
each of said portions having a dimension along the first line
larger than the period of the corresponding structure; detecting
diffracted radiation signals from the illuminated portions of the
structures to provide at least one output signal; and determining
from the at least one output signal a misalignment between the
structures.
2. The method of claim 1, further comprising dividing a
substantially coherent beam of radiation into a first and a second
beam that are substantially coherent, wherein said illuminating
employs radiation from the first beam to illuminate one of the
structures and radiation from the second beam to illuminate the
other structure.
3. The method of claim 2, wherein said dividing includes passing
said substantially coherent beam of radiation to a birefringent or
acoustooptic element.
4. The method of claim 3, further comprising modulating the first
beam at a first frequency before radiation from the first beam is
employed to illuminate said one of the structures.
5. The method of claim 2, further comprising modulating the second
beam at a second frequency before radiation from the second beam is
employed to illuminate said other one of the structures.
6. The method of claim 5, wherein said detecting detects at a
frequency that is proportional to a difference between the first
and second frequencies, or between multiples thereof.
7. The method of claim 5, wherein said modulating includes
splitting said coherent beam into a third and a fourth beam,
modulating the third and fourth beams by different frequencies and
combining the modulated third and fourth beam into a combined
beam.
8. The method of claim 7, wherein said splitting into the third and
fourth beams employs a polarizing beam splitter.
9. The method of claim 7, wherein said combined beam is
substantially coherent and is divided into said first and second
beams by the dividing.
10. The method of claim 1, wherein said illuminating uses radiation
that contains a component at a second frequency, and said detecting
detects at a frequency that is proportional to a difference between
and/or sum of the first and second frequencies, or a difference
between and/or sum of multiples thereof.
11. The method of claim 2, further comprising: combining the
diffracted radiation signals from the illuminated portions of the
two structures; and supplying the combined diffracted radiation
signals to one or more detectors.
12. The method of claim 1 1, wherein said dividing or combining
employs a birefringent or acoustooptic element.
13. The method of claim 12, further comprising causing relative
motion between the element and the structures, and wherein said
determining includes calculating a phase difference between the
diffraction radiation signals that are detected from the two
structures when said relative motion is caused.
14. The method of claim 11, wherein said combining combines a first
order diffraction signal from each of the two structures.
15. The method of claim 14, wherein said combining combines a
positive first order diffraction signal from one of the two
structures with a positive first order diffraction signal from the
other of the structure and combines a negative first order
diffraction signal from one of the two structures with a negative
first order diffraction signal from the other of the structure.
16. The method of claim 1, further comprising causing a change in
the relative phase between radiation illuminating one of the two
structures and radiation illuminating the other of the two
structures, or between the diffracted radiation signals from one of
the two structures and the diffracted radiation signals from the
other of the two structures.
17 The method of claim 1, wherein said illuminating employs a
single beam to illuminate the two portions of the two structures,
with a part of the beam illuminating the portion of one structure,
and another part of the beam illuminating the portion of the other
structure.
18. The method of claim 1, wherein said two periodic structures are
also periodic substantially along a second line transverse to the
first line, and wherein said determining determines misalignments
between the structures along the first and second lines.
19. A apparatus for detecting misalignment of two periodic
structures placed next to each other so that they are periodic
substantially along a first line, comprising: a source providing a
radiation beam that is substantially coherent to illuminate a
portion of each of the two structures, each of said portions having
a dimension along the first line larger than the period of the
corresponding structure; two or more detectors, each detector
detecting a diffracted radiation signal from the illuminated
portion of each structure to provide an output signal; and a
processor determining from said output signals any misalignment
between the structures.
20. The apparatus of claim 19, further comprising a first device
dividing a substantially coherent beam of radiation into a first
and a second beam that are substantially coherent, wherein
radiation from said first beam illuminates one of the structures
and radiation from the second beam illuminates the other of the
structures.
21. The apparatus of claim 20 wherein said first device is a
birefringent or acoustooptic element.
22. The apparatus of claim 21, further comprising an instrument
that causes the first device to move along a second direction
substantially normal to the first line when said at least one
detector detects the diffracted signals, and wherein said processor
calculates a phase difference between the output signals provided
by the two or more detectors when said relative motion is
caused.
23. The apparatus of claim 20, wherein said first device has an
optical axis at an angle of about 45.degree. to one side of the
prism.
24. The apparatus of claim 20, further comprising a second device
combining the diffracted radiation signals from the illuminated
portions of the two structures and supplying the combined
diffracted radiation signals to the detectors.
25. The apparatus of claim 24, wherein said first or second device
includes a common birefringent or acoustooptic element.
26. The apparatus of claim 25, further comprising an instrument
causing relative motion between the element and the structures, and
wherein said processor includes calculating a phase difference
between the output signals when said relative motion is caused.
27. The apparatus of claim 24, wherein said second device combines
a first order diffraction signal from each of the two
structures.
28. The apparatus of claim 27, wherein said second device combines
a positive first order diffraction signal from one of the two
structures with a positive first order diffraction signal from the
other of the structure and combines a negative first order
diffraction signal from one of the two structures with a negative
first order diffraction signal from the other of the structure.
29. The apparatus of claim 19, further comprising an instrument
that causes a change in the relative phase between radiation
illuminating one of the two structures and radiation illuminating
the other of the two structures, or between the diffracted
radiation signals from one of the two structures and the diffracted
radiation signals from the other of the two structures.
30. The apparatus of claim 19, further comprising a first modulator
modulating the first beam at a first frequency before radiation
from the first beam is employed to illuminate said one of the
structures.
31. The apparatus of claim 30, further comprising a second
modulator modulating the second beam at a second frequency before
radiation from the second beam is employed to illuminate said other
one of the structures.
32 The apparatus of claim 31, further comprising a phase detector
detecting the outputs at a frequency that is proportional to a
difference between the first and second frequencies, or between
multiples thereof.
33. The apparatus of claim 30, further comprising a second device
that splits said coherent beam into the first and second beams, and
combines the modulated first and second beams into a combined
beam.
34. The apparatus of claim 33, wherein said second device includes
a polarizing beam splitter.
35. The apparatus of claim 33, wherein said combined beam is
substantially coherent and is divided into said first and second
beams by the first device.
36. The apparatus of claim 19, wherein said radiation beam contains
a component at a second frequency, and said detectors detect a
frequency that is proportional to a difference between and/or sum
of the first and second frequencies, or a difference between and/or
sum of multiples thereof.
37 The apparatus of claim 19, wherein said source provides a single
beam to illuminate the two portions of the two structures, with a
part of the beam illuminating the portion of one structure, and
another part of the beam illuminating the portion of the other
structure.
38. The apparatus of claim 19, wherein said two periodic structures
are also periodic substantially along a second line transverse to
the first line, and wherein said processor determines misalignments
between the structures along the first and second lines.
39. A method for controlling lithographical processing of
substrates by detection of misalignment of two periodic structures
placed next to each other so that they are periodic substantially
along a first line, comprising: illuminating a portion of each of
the two structures using radiation that is substantially coherent,
each of said portions having a dimension along the first line
larger than the period of the corresponding structure; detecting
diffracted radiation signals from the illuminated portions of the
structures to provide at least one output signal; determining from
the at least one output signal a misalignment between the
structures; and adjusting a lithographical instrument in response
to the misalignment.
40. The method of claim 39, wherein said adjusting adjusts a
lithographical stepper.
41. A apparatus for controlling lithographical processing of
substrates by detection of misalignment of two periodic structures
placed next to each other so that they are periodic substantially
along a first line, comprising: a source providing a radiation beam
that is substantially coherent to illuminate a portion of each of
the two structures, each of said portions having a dimension along
the first line larger than the period of the corresponding
structure; two or more detectors, each detector detecting a
diffracted radiation signal from the illuminated portion of each
structure to provide an output signal; and a lithographical
instrument; a processor determining from said output signals any
misalignment between the structures and providing information
concerning the misalignment to the lithographical instrument.
42. A method for providing yield information by detection of
misalignment of two periodic structures placed next to each other
so that they are periodic substantially along a first line,
comprising: illuminating a portion of each of the two structures
using radiation that is substantially coherent, each of said
portions having a dimension along the first line larger than the
period of the corresponding structure; detecting diffracted
radiation signals from the illuminated portions of the structures
to provide at least one output signal; determining from the at
least one output signal a misalignment between the structures; and
providing yield related information in response to the
misalignment.
43. A apparatus for providing yield related information by
detection of misalignment of two periodic structures placed next to
each other so that they are periodic substantially along a first
line, comprising: a source providing a radiation beam that is
substantially coherent to illuminate a portion of each of the two
structures, each of said portions having a dimension along the
first line larger than the period of the corresponding structure;
two or detectors, each detector detecting a diffracted radiation
signal from the illuminated portion of each structure to provide an
output signal; and one or more processors determining from said
output signals any misalignment between the structures and
providing yield related information.
Description
BACKGROUND OF THE INVENTION
[0001] This invention relates in general to metrology systems for
measuring periodic structures such as overlay targets employed in
photolithography in a research or production environment, and, in
particular, to a metrology system employing optical phase for
detecting misalignment of such structures.
[0002] Overlay error measurement requires specially designed marks
to be strategically placed at various locations, normally in the
street area between dies, on the wafers for each process. The
alignment of the two overlay targets from two consecutive processes
is measured for a number of locations on the wafer and the overlay
error map across the wafer is analyzed to provide feedback for the
alignment control of lithography steppers.
[0003] A key process control parameter in the manufacturing of
integrated circuits is the measurement of overlay target alignment
between successive layers on a semiconductor wafer. If the two
overlay targets are misaligned relative to each other, the
electronic devices fabricated will malfunction and the
semiconductor wafer will need to be reworked or discarded.
[0004] Typically, conventional overlay targets are box-in-box
targets and bar-in-bar targets. The box-in-box target typically has
a 10 .mu.m inner box and a 20 .mu.m outer box. The outer box is
printed on the substrate (or previous process layer) and the inner
box is resist printed on the current layer. Overlay error is
reported as the mis-position of the inner mark with respect to the
outer mark. A bar-in-bar target also has a 10 .mu.m inner target on
the current layers and a 20 .mu.m outer target on the previous
layers. However, the box edge is replaced with a narrow bar 2 .mu.m
wide. The box-in-box targets are more compact; however, the
bar-in-bar targets provide better measurement performance. Overlay
targets may comprise grating structures on top of the wafer or
etched into the surface of the wafer. For example, one overlay
target may be formed by etching into the wafer while another
adjacent overlay target may be a photoresist layer at a higher
elevation over the wafer.
[0005] Conventional systems for detecting overlay target
misalignment typically employs an electronic camera that images the
"box-in-box target." The accuracy of the conventional system is
limited by the accuracy of the line profiles in the target, by
aberrations in the illumination and imaging optics and by the image
sampling in the camera. Such methods are complex and they require
full imaging optics. Vibration isolation is also required and it
may be difficult to integrate such systems into process
equipment.
[0006] An improvement to the conventional method is described in
U.S. Pat. No. 6,023,338. This patent discloses a method where two
overlay target structures are placed next to each other and two
radiation beams are scanned in two separate paths across portions
of both structures. The intensity of the radiation reflected along
both paths are detected and processed to calculate any offset
between the two structures.
[0007] While the above-described improved method may be useful for
some applications, it requires beams to be scanned across periodic
structures such as overlay targets. It is desirable to develop an
improved system with better performance and simplified scanning
characteristics.
SUMMARY OF THE INVENTION
[0008] This invention is based on the observation that by utilizing
optical phase detection, high sensitivity for detecting
misalignment of periodic structures can be achieved. Thus, two
periodic structures such as overlay targets are placed side-by-side
so that they are periodic substantially along the same direction,
where portions of both structures are illuminated by coherent
radiation. The size(s) of the beam(s) illuminating portions of the
structures are large enough to generate diffraction signals by the
structures. These diffraction signals are caused to interfere
leading to the detection of optical phase which is a measurement of
the misalignment between the structures. The misalignment may then
be used to control lithographic instruments such as a lithographic
stepper or to determine whether or not the patterns of the
structures are correctly placed and will yield functional
devices.
[0009] When the paths of radiation traveling between the radiation
source, the structures and detectors are close together, the phase
sensitive detection is less sensitive to environmental factors such
as vibration and thermal drifts. Since the system employs larger
spot illumination, the optics of the system are less sensitive to
focus accuracy. The system is compact and readily integratable with
process equipment. Due to the enhanced sensitivity compared to
conventional systems, the system is able to detect misalignment of
periodic structures that are low contrast.
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] FIG. 1 is a top view of two overlay targets placed next to
each other that are illuminated by two corresponding radiation
beams to illustrate the invention.
[0011] FIG. 2 is a perspective view of two overlay targets placed
side-by-side and a Wollaston prism useful for illustrating the
invention.
[0012] FIGS. 3A and 3B are side views from different angles of an
overlay target metrology system illustrating an embodiment of the
invention. FIG. 3C is a perspective view of the system of FIGS. 3A
and 3B.
[0013] FIG. 4 is a flow diagram illustrating cross-sectional light
distributions at various positions in the paths of radiation beams
in the system of FIGS. 3A and 3B.
[0014] FIG. 5 is a schematic view of the Wollaston prism of FIGS.
3A and 3B illustrating its function in dividing a beam of radiation
into two beams.
[0015] FIG. 6 is a graphical plot of output signals from the
detectors of FIGS. 3A and 3B to illustrate a method for obtaining
offset or misalignment information between the overlay targets from
the detector outputs.
[0016] FIG. 7A is a schematic view of a conventional Wollaston
prism dividing an input polarized beam into two polarized beams
useful for illustrating the invention.
[0017] FIG. 7B is a schematic view of a modified Wollaston prism
dividing an input polarized beam into two, useful for illustrating
the invention.
[0018] FIG. 8A is a partially schematic and partially
cross-sectional view of a dual heterodyne differential phase
metrology system for measuring misalignment of two adjacent overlay
targets to illustrate a preferred embodiment of the invention.
[0019] FIG. 8B is a partially schematic and partially
cross-sectional view of a dual heterodyne differential phase
metrology system for measuring misalignment of two adjacent overlay
targets to illustrate an embodiment of the invention similar to
that of FIG. 8A in many respects.
[0020] FIG. 9A is a cross-sectional view of a metrology system for
measuring misalignment of overlay targets to illustrate another
embodiment of the invention.
[0021] FIG. 9B is a graphical plot of the output signals from the
metrology system of FIG. 9A.
[0022] FIG. 9C is a perspective view of the system of FIGS. 9A and
9B.
[0023] FIG. 10 is a top view of two overlay targets placed adjacent
to each other illuminated by a single beam of radiation to
illustrate yet another embodiment of the invention.
[0024] FIGS. 11A and 11B are two side views from different angles
of a metrology system where a single beam of radiation is employed
to illuminate two overlay targets for detecting misalignment
illustrating yet another embodiment of the invention.
[0025] FIG. 12 is a flow diagram illustrating the cross-sectional
light distributions at various positions of radiation beams in the
system of FIGS. 11A and 11B.
[0026] FIG. 13 is a flow diagram illustrating the cross-sectional
light distributions at various positions of radiation beams in the
system of FIGS. 3A and 3B where the two overlay targets are
periodic in two orthogonal directions.
[0027] FIG. 14 is a block diagram illustrating a lithographical
instrument in combination with a metrology system for measuring
misalignment of overlay targets to illustrate still another
embodiment of the invention.
[0028] For simplicity of description, identical components are
labeled by the same numerals in this application.
DETAILED DESCRIPTION OF THE EMBODIMENTS
[0029] FIG. 1 is a top view of two periodic structures 22, 24 such
as overlay targets placed side-by-side, where both structures are
illuminated by a beam of radiation having a round cross-section as
shown in FIG. 1. Preferably both structures 22, 24 have the same
period .LAMBDA., and are preferably aligned so that they are
periodic substantially along the same line of direction X in the XY
coordinate system. The two illuminated spots 26, 28 are preferably
aligned with the centers of the spots substantially aligned along
the Y axis. This can be achieved, for example, by aligning the two
spots with two markers on one of the overlay target structures to
correctly position the two spots. The size of spot 26 along the X
axis is such that its dimension along the X axis is at least equal
to or greater than the period .LAMBDA. (more typically equal to or
greater than several periods .LAMBDA.'s) in order for a diffracted
signal to be generated by illuminating structure 22 at spot 26.
Preferably, spot 26 would span a number of grating lines 22a of
structure 22. The same can be said of spot 28 with respect to the
period .LAMBDA. and other features of structure 24.
[0030] A metrology system 30 for generating the two illuminated
spots 26, 28 on the two corresponding structures is illustrated in
FIG. 2. As shown in FIG. 2, a beam of radiation (not shown) may be
split by means of a Wollaston prism 32 to illuminate the two
structures 22 and 24. As shown in FIG. 2, the Wollaston prism is
located above the two structures at a predetermined z height above
the structures.
[0031] FIGS. 3A and 3B are side views of the metrology system 30 of
FIG. 2 in directions along the arrows 3A, 3B in FIG. 2,
respectively. FIG. 3C is a perspective view of system 30. As shown
in these figures, a beam of radiation 34 is split by prism 32 into
two beams 34a, 34b in the Y direction. These two beams are
collected by lens 36 towards two overlay target structures 22, 24
which are placed side-by-side above a substrate 38 such as a
silicon wafer. Beam 34a illuminates spot 26 on structure 22 and
beam 34b illuminates spot 28 of structure 24. Beam 34a is
diffracted by structure 22. The zeroth-order diffraction retraces
the original path of beam 34a and through prism 32. The first order
diffraction 40a- and 40a+ from structure 22 of beam 34a are
collected by lens 36 towards prism 32 and reflected by mirrors 44
and 46 towards two detectors 52, 54 as shown in FIG. 3B. Similarly,
the zeroth-order diffracted signal from structure 24 of beam 34b
retraces the original path of beam 34b. The first order diffracted
signals 42b+ and 42b- are collected by lens 36 through prism 32 to
the two detectors 52, 54. The relative positions of the various
beams of radiation in FIGS. 3A and 3B are shown more clearly in the
flow diagram of FIG. 4. The outputs of detectors 52, 54 are
S.sub.-1 and S.sub.+1, respectively, as shown in FIG. 3B. As shown
in FIG. 4, the input beam 34 is applied to Wollaston prism 32 which
splits or divides the beam into two beams 34a, 34b when it reaches
the objective 36. After passing through the objective, the two
beams 34a, 34b illuminate, respectively, spots 26, 28 of structures
22, 24. To simplify the figures, the zeroth-order diffracted signal
from the two structures have been omitted from FIG. 4. The positive
and negative diffracted signals 40a+ and 40a- are collected by lens
36. Similarly, the positive and negative first diffraction signal
42b+ and 42b- are collected by lens 36 towards prism 32. Prism 32
combines beams 40a+ and 42b+ from the two spots 26, 28 into one
beam towards detector 54. The input beam 34 is coherent so that
beams 34a, 34b are coherent after passing prism 32. Therefore, the
positive first order diffraction 40a+ and 42b+ are also coherent.
They would, therefore, interfere when combined by prism 32 and at
detector 54. Similarly, the negative first order diffracted signals
40a- and 42b- are combined by prism 32 and interfere at the prism
and at the detector 52. The phase difference between the outputs of
the two detectors 52, 54 is determined. This phase difference
indicates the phase difference between beams 40a+, 42b+ and that
between the pair of beams 40a- and 42b-, and provides information
concerning misalignment between the two structures 22, 24. The two
output signals S.sub.+1 and S.sub.-1 are shown in Equations 1 and 2
below:
S.sub.+1.varies.{.eta..sub.+1.sup.a+.eta..sub.+1.sup.b+2{square
root}{square root over (.eta..sub.+1.sup.a.eta..sub.+1.sup.b)} cos
(.phi..sub.+1+.phi..sub.w+.phi..sub.z+.phi..sub.x)} (1)
S.sub.-1.varies.{.eta..sub.-1.sup.a+.eta..sub.-1.sup.b+2{square
root}{square root over (.eta..sub.-1.sup.a.eta..sub.-1.sup.b)} cos
(.phi..sub.-1+.phi..sub.w+.phi..sub.z-.phi..sub.x)} (2)
[0032] In the equations (1) and (2) above, .eta..sub.+1.sup.a is
the diffraction efficiency of the diffracted signal 40a+ from
interaction between structure 22 and beam 34a at spot 26,
.eta..sub.+1.sup.b the diffraction efficiency of the +1 diffracted
signal 42b+ from interaction between structure 24 and beam 34b at
spot 28, .eta..sub.-1.sup.a the diffraction efficiency of the
diffracted signal 40a- from interaction between structure 22 and
beam 34a at spot 26, and .eta..sub.-1.sup.b the diffraction
efficiency of the -1 diffracted signal 42b- from interaction
between structure 24 and beam 34b at spot 28. The phase terms
.phi..sub.+1 and .phi..sub.-1 are residual phase differences due to
the grating property difference and are defined by the equations
below:
.phi..sub.+1=.phi..sub.+1.sup.a-.phi..sub.+1.sup.b;
.phi..sub.-1=.phi..sub.-1.sup.a-.phi..sub.-1.sup.b
[0033] where .phi..sub.+1.sup.a and .phi..sub.-1.sup.a are the
phase terms in the positive and negative diffracted signals, where
such phase terms depend on the material and other properties of
structure 22; and .phi..sub.+1.sup.b and .phi..sub.-1.sup.b are the
phase terms in the positive and negative diffracted signals, where
such phase terms depend on the material and other properties of
structure 24. The two phase differences .phi..sub.+1 and
.phi..sub.-1 are identical for +1 and -1 orders if the grating
profile is symmetric. The term .phi..sub.x is the phase difference
caused by any misalignment .DELTA.x between the two periodic
structures 22, 24 as set forth in equation (3) below. The amount of
this phase shift is readily determined from the shifting theorem of
Fourier transform. The grating with period .LAMBDA. in frequency
domain has a spatial frequency at 1/.LAMBDA.. A phase shift
.DELTA.x in space translates into
2.pi.(1/.LAMBDA.).DELTA..sub..chi. phase shift in frequency domain,
directly predicted by shifting theorem. The term .phi..sub.z is the
phase difference between the two detector outputs caused by the
height difference .DELTA.z between the two structures 22, 24 as
shown in equation (4) below. The term .phi..sub.w is the phase
difference between the two detector outputs caused by phase shift
induced by the Wollaston prism 32 as shown in equation (5) below. 1
x = 2 x ( 3 ) z = 2 n z ( 4 ) w = 2 ( n o - n e ) h ( 5 )
[0034] where .LAMBDA. is the grating period of structures 22, 24; n
the average index of refraction of material 39; n.sub.o, n.sub.e
the indices of refraction for the ordinary and extraordinary rays
of the prism 32; .lambda. the wavelength of beam 34, and .DELTA.h
is defined below.
[0035] FIG. 5 is a schematic view of the Wollaston prism 32 of
FIGS. 3A, 3B illustrating its function in dividing a beam of
radiation into two beams and the phase shift introduced by the
prism. As shown in FIG. 5, beam 34 is split by prism 32 into two
beams 34a, 34b with an angle .theta. in between them, where .theta.
is given by equation (6) below. The phase difference caused by
prism 32 between the two detector outputs S.sub.+1 and S.sub.-1 is
given by equation (7) below, where n.sub.o and n.sub.e are the
refractive indices of prism 32 in the ordinary and in the
extraordinary directions of the prism. The term .DELTA.h in
equation (5) above is the optical path length difference between h1
(the optical path length of beam 34 in prism 32 before reaching the
optical interface 32a)) and h2 (the optical path length of beams
34a and 34b through the prism after the interface 32a)). When prism
32 is moved along the y direction relative to beam 34, beam 34 will
reach the interface 32a) of prism 32 at a different position,
thereby changing the optical path lengths h1, h2 and also changing
the quantity .DELTA.h. .DELTA.h is proportional to the distance y
traveled by prism 32 as set forth below:
.DELTA.h=2y.alpha.
[0036] where .alpha. is the angle the optical interface 32a) makes
with the sides of the prism through which the beams 34, 34a, 34b
pass as shown in FIG. 5, so that equation (5) becomes equation (7)
below.
.theta.=2.alpha.(n.sub.o-n.sub.e)
[0037] 2 w = 2 2 ( n o - n e ) y = 2 y ( 7 )
[0038] where the origin of the Y axis is at the position where
h.sub.1=h.sub.2. The separation .delta.y between two spots on
target is given by: 3 y = 2 f tan ( 2 ) ( 8 )
[0039] where f is the objective focal length. For small angle
.theta., an approximation of this formula becomes:
.delta.y=f.theta.
[0040] Thus, by moving prism 32 along the y axis and thereby
changing the value of y, the phase term .phi..sub.w changes as a
function of y, which, in turn, causes the two detector outputs
S.sub.+1 and S.sub.-1 to also change as a function of the
displacement y of prism 32 along the y axis, as illustrated in FIG.
6. As noted above, if the grating profiles of structures 22, 24 are
symmetric, .phi..sub.+1 and .phi..sub.-1 are identical. The phase
term .phi..sub.z is the same for both detector outputs. Therefore,
the phase difference between the two detector outputs S.sub.+1 and
S.sub.-1 in FIG. 6 is caused only by the overlay target error in
equation (3). In other words, the phase difference .DELTA..phi.
between the two detector outputs is given by 2.phi..sub.x as
indicated in equation (9) below:
.DELTA..phi.=2.phi..sub.x (9)
[0041] wherein the phase difference between detector signals
S.sub.+1 and S.sub.-1: twice overlay error. Therefore, by measuring
the phase difference between the two detector outputs at different
displacement values of y as indicated in FIG. 6, the misalignment
.DELTA.x between the two structures 22, 24 can be determined from
equations (3) and (9) and the grating period of structures 22, 24.
This may be performed by a processing device 50 in FIG. 3B, where
the device may simply be a microprocessor, or programmable logic or
any other suitable processing device which can compute the
misalignment .DELTA.x from the two outputs S.sub.+1 and
S.sub.-1.
[0042] As a practical matter, normally Wollaston prisms are
designed such that the optical axis 32b (shown in a YZ plane as
shown in FIG. 7A) is parallel to the length L in one half and the
optical axis 32c (shown in an XZ plane as shown in FIG. 7A) is
perpendicular to the length in the other half of the prism, where
the two halves are separated by optical interface 32a, so that the
two beams 34a, 34b will be either parallel polarized or
orthogonally polarized with respect to the grating fingers of
structures 22, 24. The two different polarizations of the two beams
have different diffraction efficiencies. This can cause an
imbalance in amplitude between the two outputs S.sub.-1 and
S.sub.+1 . This tends to limit the sensitivity of system 30 for a
given bandwidth. It is possible to overcome this problem by
rotating the prism by 45.degree., so that the polarization
direction 34c is at 45.degree. to the optical axis 32b, and so that
the polarizations of the two beams 34a, 34b will be at +45 and
-45.degree. respectively with respect to the grating fingers of the
two structures as shown in FIG. 7A. However, that rotates the spot
on the sample, and results in a signal even in the absence of any
misalignment. The offset can be cancelled through calibration;
however, this requires a target with larger footprint.
[0043] Another solution is to use a modified Wollaston prism 32',
in which the optical axis 32b' (shown in a YZ plane as shown in
FIG. 7B) in the first half of the prism is at +45 .degree. with
respect to the length L of the prism 32' and the optic axis 32c'
(also shown in a YZ plane as shown in FIG. 7B) is at -45.degree. to
the length in the other half of prism 32' on the other side of
interface 32a' as shown in FIG. 7B. If the incoming beam 34 is
vertically polarized (as indicated by line 34c' in FIG. 7B), it
will be split by the prism into two substantially equal amplitude
beams 34a', 34b', each at 45.degree. polarization with respect to
the fingers of the gratings 22, 24 and giving substantially the
same reflectance for substantially identical gratings. At the same
time, the center line of the two spots will then be parallel to the
fingers of the gratings as desired. Any residual difference in
diffraction efficiencies of the two gratings can still be balanced
by tuning the incoming polarization, using, for example, a
half-wave plate.
[0044] Instead of having to move the Wollaston prism as described
above, an alternative differential heterodyne system 100 requiring
no moving parts is possible; this is illustrated in FIG. 8A as
another embodiment. As shown in FIG. 8A, a coherent beam of
radiation 102 is split into two beams 102a and 102b by polarizing
beamsplitter 104. Half wave plate 101 is employed to control the
intensity ratio between E.sub.1 and E.sub.2 to compensate for the
diffraction efficiency differences between the two overlay targets
22, 24. Beam 102a is frequency shifted upwards by modulator 106 by
frequency .omega..sub.1 to become beam E.sub.1 and beam 102b is
frequency shifted upwards by a modulator 108 by frequency
.omega..sub.2 to become beam E.sub.2. The two modulated beams are
reflected by mirrors 110 and passed through respective wave plates
112a, 112b so that E.sub.1 and E.sub.2 can pass through and be
reflected respectively by the polarizing beam splitter 114. The two
beams pass through beamsplitter 116 to a birefringent element such
as a Wollaston prism 32 which separates the two beams E.sub.1 and
E.sub.2 on account of their different polarizations. Beam E.sub.1
is collected by lens 36 towards structure 24 and beam E.sub.2 is
collected by the lens towards structure 22.
[0045] The + 1 diffraction order E.sub.1,+1 of beam E.sub.1 from
structure 24 is focused by lens 36 towards prism 32. The +1
diffraction order E.sub.2,+1 of beam E.sub.2 from structure 22 is
also focused by lens 36 and combined with the +1 diffraction order
E.sub.1,+1 by prism 32 into a single beam which is reflected by
beamsplitter 116. The polarizations of the fields E.sub.1,+1,
E.sub.2,+1 are orthogonal to each other. Analyzer 118 is placed at
substantially 45.degree. to the polarization directions of
E.sub.1,+1, E.sub.2,+1 so that the components passed by the
analyzer have the same polarization and will interfere at detector
54, after reflection by mirror 134. Similarly, the -1 diffraction
order signals E.sub.1,-1, E.sub.2,-1 of the respective beams
E.sub.1, E.sub.2 by structures 24, 22, respectively, are focused by
lens 36 towards the same spot in prism 32 which combines the two
beams into a single beam E.sub.1,-1+E.sub.2,-1. Again the
components of E.sub.1,-1, E.sub.2,-1 passed by the analyzer 118
have the same polarization and will interfere at detector 52, after
reflection by mirror 132. The outputs of the detectors 52, 54 are
then provided to a phase detector 140 which detects a beat
frequency .omega..sub.b which is equal to the difference between
.omega..sub.1 and .omega..sub.2 in order to determine the
misalignment or overlay target error .DELTA.x. The equations of the
various optical signals E.sub.1, E.sub.2, E.sub.1,+1, E.sub.2,+1,
E.sub.1,-1, E.sub.2,-1 are set forth below. In the equations below,
.vertline.E.sub.1.vertline., .vertline.E.sub.2.vertline. are the
amplitudes of the two beams E.sub.1 and E.sub.2 and .omega..sub.0
is the frequency of input beam 102. The terms .theta..sub.n1 and
.theta..sub.n2 indicate the initial phase of input beam 102 and
also the optical path length difference of the two beams between
beamsplitter 104 and polarizing beamsplitter 114. The terms
.phi..sub.+1, .phi..sub.-1, .phi..sub.x, .phi..sub.z, and
.phi..sub.w have the same meanings as those described above in the
previous embodiment where the Wollaston prism is moved, and are set
forth in equations (3)-(5) above.
E.sub.1=.vertline.E.sub.1.vertline.exp{j(.omega..sub.0t+.omega..sub.1t+.th-
eta..sub.n1)} (10)
E.sub.2=E.sub.2.vertline.exp{j(.omega..sub.0t+.omega..sub.2t+.theta..sub.n-
2)} (11)
E.sub.1,+1=.vertline.E.sub.1.vertline.{square root}{square root
over
(.eta..sub.+1.sup.b)}exp{j(.omega..sub.0t+.omega..sub.1t+.theta..sub.n1+.-
phi..sub.+1.sup.b+.phi..sub.x)} (12)
E.sub.1,-1=.vertline.E.sub.1.vertline.{square root}{square root
over
(.eta..sub.-1.sup.b)}exp{j(.omega..sub.0t+.omega..sub.1t+.theta..sub.n1+.-
phi..sub.-1.sup.b-.phi..sub.x)} (13)
E.sub.2,+1=.vertline.E.sub.2.vertline.{square root}{square root
over
(.eta..sub.+1.sup.a)}exp{j(.omega..sub.0t+.omega..sub.2t+.theta..sub.n2+.-
phi..sub.+1.sup.a+.phi..sub.z+.phi..sub.w)} (14)
E.sub.2,-1=.vertline.E.sub.2.vertline.{square root}{square root
over
(.eta..sub.-1.sup.a)}exp{j(.omega..sub.0t+.omega..sub.2t+.theta..sub.n2+.-
phi..sub.-1.sup.a+.phi..sub.z+.phi..sub.w)} (15)
[0046] The quantities {square root}{square root over
(.eta..sub.+1.sup.b)}, {square root}{square root over
(.eta..sub.+1.sup.a)}, {square root}{square root over
(.eta..sub.-1.sup.b)}, {square root}{square root over
(.eta..sub.-1.sup.a)} are the diffraction efficiencies of the two
structures for the +1 and -1 diffraction in the equations above.
The output of detector 54 is given by equation (16) below and the
output of detector 52 is given by equation 17 below.
S.sub.+1.varies.E.sub.1,+1+E.sub.2,+1.sup.2=.vertline.E.sub.1.vertline..su-
p.2.eta..sub.+1.sup.b+.vertline.E.sub.2.vertline..sup.2+72.sub.-1.sup.a+2.-
vertline.E.sub.1.vertline..vertline.E.sub.2.vertline.{square root
over (.eta..sub.+1.sup.a.eta..sub.+1.sup.b)} cos
[+107.sub.bt+(.theta..sub.n2--
.theta..sub.n1)+(.PHI..sub.+1.sup.a.PHI..sub.+1.sup.b)++101.sub.z+.PHI..su-
b.w-.PHI..sub.x]
S.sub.-1.varies.E.sub.1,-1+E.sub.2,-1.sup.2=.vertline.E.sub.1.vertline..su-
p.2.eta..sub.-1.sup.b+.vertline.E.sub.2.vertline..sup.2+72.sub.-1.sup.a+2.-
vertline.E.sub.1.vertline..vertline.E.sub.2.vertline.{square root
over (.eta..sub.-1.sup.a.eta..sub.-1.sup.b)} cos
{+107.sub.bt+(.theta..sub.n2--
.theta..sub.n1)+(.PHI..sub.-1.sup.a.PHI..sub.-1.sup.b)++101.sub.z+.PHI..su-
b.w-.PHI..sub.x}
[0047] S.sub.+1 and S.sub.-1 are sinusoidal functions at the beat
frequency or difference frequency
.omega..sub.b=.omega..sub.1-.omega..sub- .2 in the time domain.
Again, where the two structures 22, 24 are symmetric, the following
phase relation would hold:
.phi..sub.+1.sup.a=.phi..sub.-1.sup.a
.phi..sub.+1.sup.b=.phi..sub.-1.sup.b (18)
[0048] Thus, from equations (16)-(18) above, it is noted that the
phase difference between S.sub.+1 and S.sub.-1, is
2.phi..sub.x.
[0049] Therefore, by detecting the phase difference between the two
detector outputs, the overlay target error .DELTA.x between the two
structures 22, 24 may be obtained from equations (3) and (9)
above.
[0050] Advantageously, the detector 140 may be set to detect at the
beat or difference frequency .omega..sub.1-.omega..sub.2 to improve
the signal-to-noise ratio of .phi..sub.x.
[0051] From a comparison of FIG. 8A to FIGS. 3A, 3B, it will be
evident that the two diffracted and combined beams
E.sub.1,+1+E.sub.2,+1 and E.sub.1,-1+E.sub.2,-1 l as well as their
corresponding detectors 54, 52 are not in the plane of the paper;
instead, one combined beam and its corresponding detector are above
the plane of the paper and the other combined beam and its
corresponding detector are in the plane below the plane of the
paper.
[0052] FIG. 8B is a partially schematic and partially
cross-sectional view of a dual heterodyne differential phase
metrology system for measuring misalignment of two adjacent overlay
targets to illustrate another embodiment of the invention analogous
to that in FIG. 8A. As shown in FIG. 8B, metrology system 100' is
similar to system 100 on FIG. 8A, except that a non-polarizing
beamsplitter 114' is used instead of the polarizing beamsplitter
114 of FIG. 8A. Waveplate 101 is again used for controlling the
intensity in ratio between the two beams E.sub.1 and E.sub.2.
[0053] The typical heterodyne phase different .phi..sub.x detected
is of the order of 10.sup.-4 radians where the spacing .LAMBDA. of
structures 22, 24 is of the order of 2 microns in equation (3)
above. Therefore, with such or similar grating period of structures
22, 24, it is possible to detect misalignment .DELTA.x which can be
as small as 30 picometers.
[0054] FIGS. 9A and 9C illustrate another embodiment of the
invention without any moving parts. FIG. 9B is a graphical plot of
the output signals from the metrology system of FIGS. 9A, 9C. As
shown in FIGS. 9A, 9C, a portion of an input coherent beam of
radiation 102 at frequency .omega..sub.0 is reflected by mirror 202
and passes through an acoustooptic deflector (AOD) 204. A portion
212 of the input beam at frequency .omega..sub.0 passes through the
AOD without being deflected or changed in frequency and is
collected by lens 36 towards structure 24. Another portion 214 of
the input beam 102 is upshifted by frequency .omega. and is
deflected by the AOD 204 to provide an upshifted and deflected beam
at frequency .omega..sub.0+.omega. which is also collected by lens
36 towards structure 22.
[0055] Beam 214 is diffracted by structure 22 and the +1 and -1
orders of the diffraction are beams 214+ and 214-, respectively.
The two diffracted beams 214+ and 214- are both at frequency to
.omega..sub.0+.omega.. The two diffracted beams 214+ and 214- are
collected by lens 36 towards AOD 204 which passes without
deflection or change in frequency a portion thereof towards mirror
208. The undeflected beam 212 at frequency .omega..sub.0 is
diffracted by structure 24 into two diffracted beams 212+ and 212-,
both at frequency .omega..sub.0. These two beams are collected by
lens 36 towards the AOD 204 which deflects and downshifts in
frequency by .omega. a portion of the two beams towards mirror 208.
The AOD 204 combines the deflected portion of beam 212+ and the
undeflected portion of beam 214+ to form the +1 order mixed output
signal S.sub.+1' so that the two portions would interfere and yield
an output at beat frequency equal to 2.omega.. Similarly, AOD 204
would combine the deflected portion of beam 212- and the
undeflected portion of beam 214- as the -1 order mixed signal
output S.sub.-1' at beat frequency 2.omega..
[0056] As in the previous embodiment shown in FIGS. 8A, 8B, any
misalignment between the two structures 22, 24 in the X direction
would introduce a phase difference .phi..sub.x between the +1 order
(S.sub.+1 ) and the -1 order (S.sub.-) of mixed outputs in FIG. 9C.
Thus, equations 16-18 may then be used to derive the overlay target
area .DELTA..sub.x, where the beat frequency .omega..sub.b in
equations 16 and 17 is equal to 2.omega..
[0057] FIG. 10 is a schematic view of the two structures 22, 24
illuminated by a single beam to illustrate another embodiment of
the invention. Thus, instead of splitting single beam into two
beams, for illuminating the two respective structures, a single
beam may be used. FIGS. 11A, 11B are the front and side views of a
system 300 employing a single beam for illuminating the two
structures to illustrate yet another embodiment of the invention.
As shown in FIGS. 10, 11A, 11B, this embodiment is analogous to
that shown in FIGS. 2, 3A and 3B, except that a single beam 302 is
employed that illuminates both structures 22, 24. As shown in FIGS.
10 and 11A, the single beam 302 illuminates a portion of structure
22 and a portion of structure 24 at their edges next to each other.
The +1 and -1 diffraction signals from the two structures are
collected by lens 36 and combined and detected by two detectors
(not shown) to provide the two outputs S.sub.+1 and S.sub.-1 as in
the embodiment of FIGS. 2, 3A and 3B. A flow diagram illustrating
the relative positions of the beams are shown in FIG. 12. As shown
in FIG. 12 the +1 diffraction beams from structures 22, 24 are
combined into a single beam 340 and the -1 diffraction beams from
the two structures are combined into a single beam 342. The two
beams are detected by respective detectors 52, 54 to provide the
output S.sub.+1 and S.sub.-1 as before.
[0058] One consideration in the embodiment of FIGS. 10, 11A, 11B
and 12 is that the alignment of the illuminating beam 302 is
preferably controlled relative to the two structures to ensure that
the relative intensity of the diffraction beams from both
structures are held stable and substantially equal. This can be
accomplished by precision beam positioning relative to the
patterning on the wafer which can be augmented using an imaging
system such as an optical microscope to visualize the beam location
and make fine adjustments to the location of the beam 302. Other
methods include measuring the relative intensity of the diffraction
components from the two structures in a feedback system to move the
beam 302 to even and consistent alignment between the two
structures. For such alignment, the wafer may be moved by means of
cross-roller translation stages driven by micrometers.
Alternatively, beam 302 may be moved by means of adjustable beam
steering mirrors or galvanometer mirrors and the like. All such
different alignment mechanisms are believed to be known to those
skilled in the art.
[0059] In the embodiments described above, misalignment is detected
along one direction, namely, the x direction. In order to discover
misalignment along an orthogonal direction such as the y direction,
typically a second optical detection system is employed. Yet
another embodiment envisions a two-dimensional alignment system as
illustrated in FIG. 13. For this purpose a pair of two-dimensional
gratings C and D is created and positioned on their respective
processing layers in the same manner as the previously described
structures 22, 24. In the case of a dual beam measurement
configuration described above in reference to FIGS. 2, 3A, 3B and
4, the two illumination beam on the two gratings would each give
rise to four principal diffraction components, two in each axis of
the gratings C and D.
[0060] Aside from the use of two-dimensional gratings C and D
instead of one-dimensional gratings 22, 24, and the need to detect
four outputs instead of two, the apparatus for the embodiment of
FIG. 13 is the same as the previous embodiment illustrated in
reference to FIGS. 2, 3A, 3B and 4. The two quartets of diffraction
orders from gratings C and D are brought into alignment by the
Wollaston prism 32 and four separate photodetectors 352, 354, 356
and 358 are used to separately measure the two sets of interference
signals corresponding to the positive and negative diffraction
orders in each of the two axes. Conveniently, the detector used in
this configuration may be a quadrant photodiode. Interpretation of
the phase information proceeds as before and alignment information
in two axes may be simultaneously derived. The misalignment along
the y direction is determined by calibration based on two targets
on the same layer to offset the DC phase shift induced by the
Wollaston prism for the diffraction in the y direction.
[0061] FIG. 14 is a block diagram illustrating a lithographic
instrument 400 in combination with a metrology system 100 for
measuring misalignment of overlay targets to illustrate still
another embodiment of the invention. As shown in FIG. 14, the
misalignment between two overlay targets measured by the metrology
system 100 of FIG. 8A may be supplied to lithographic instrument
400 for controlling a lithographic process, such as one performed
by a lithographic stepper. Thus, the metrology system 100 may be
integrated with instrument 400 to become an integrated system for
the convenient operation of misalignment detection and lithographic
process control. Alternatively misalignment information may be used
to determine probable device yield, and to make decision about
future processing or rework of the devices. The process for
performing such yield analysis and management is known to those
skilled in the art so that no details need to be provided herein.
This can also be done by supplying the misalignment of overlay
targets information to instrument 400, where the instrument is
equipped with yield analysis and management capability residing in
a microprocessor in the instrument 400, in addition to lithographic
control. Alternatively, such capability may reside in processing
device or processor 50.
[0062] While the invention has been described above by reference to
various embodiments, it will be understood that changes and
modifications may be made without departing from the scope of the
invention, which is to be defined only by the appended claims and
their equivalents.
* * * * *