U.S. patent application number 16/672480 was filed with the patent office on 2020-05-07 for topographic phase control for overlay measurement.
The applicant listed for this patent is KLA-Tencor Corporation. Invention is credited to Vladimir Levinski, Amnon Manassen, Yuri Paskover, Yoni Shalibo.
Application Number | 20200142322 16/672480 |
Document ID | / |
Family ID | 57320785 |
Filed Date | 2020-05-07 |
![](/patent/app/20200142322/US20200142322A1-20200507-D00000.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00001.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00002.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00003.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00004.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00005.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00006.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00007.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00008.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00009.png)
![](/patent/app/20200142322/US20200142322A1-20200507-D00010.png)
View All Diagrams
United States Patent
Application |
20200142322 |
Kind Code |
A1 |
Levinski; Vladimir ; et
al. |
May 7, 2020 |
Topographic Phase Control For Overlay Measurement
Abstract
Metrology tools and methods are provided, which estimate the
effect of topographic phases corresponding to different diffraction
orders, which result from light scattering on periodic targets, and
adjust the measurement conditions to improve measurement accuracy.
In imaging, overlay error magnification may be reduced by choosing
appropriate measurement conditions based on analysis of contrast
function behavior, changing illumination conditions (reducing
spectrum width and illumination NA), using polarizing targets
and/or optical systems, using multiple defocusing positions etc.
On-the-fly calibration of measurement results may be carried out in
imaging or scatterometry using additional measurements or
additional target cells.
Inventors: |
Levinski; Vladimir; (Migdal
HaEmek, IL) ; Paskover; Yuri; (Binyamina, IL)
; Manassen; Amnon; (Haifa, IL) ; Shalibo;
Yoni; (Binyamina, IL) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
KLA-Tencor Corporation |
Milpitas |
CA |
US |
|
|
Family ID: |
57320785 |
Appl. No.: |
16/672480 |
Filed: |
November 3, 2019 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
15114175 |
Jul 26, 2016 |
10520832 |
|
|
PCT/US16/33353 |
May 19, 2016 |
|
|
|
16672480 |
|
|
|
|
62163783 |
May 19, 2015 |
|
|
|
62222724 |
Sep 23, 2015 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G02B 7/38 20130101; G03F
7/70633 20130101; H04N 5/23212 20130101; G06T 7/80 20170101; G02B
27/32 20130101 |
International
Class: |
G03F 7/20 20060101
G03F007/20; H04N 5/232 20060101 H04N005/232; G02B 27/32 20060101
G02B027/32; G02B 7/38 20060101 G02B007/38; G06T 7/80 20060101
G06T007/80 |
Claims
1. An optical system in an imaging metrology tool, the optical
system comprising an adjustable reference path having a reference
signal integrated in a collection path thereof, wherein the
reference path is configured to provide an adjustable phase of the
reference signal and the optical system is configured to adjust the
phase of the reference signal to modify a topographic phase of an
imaging metrology target to be an integer multiple of .pi.
2. The optical system of claim 1, wherein the adjustable reference
path includes a Linnik interferometer with a reference objective
identical to an objective of the imaging metrology optical system
and an adjustable mirror.
3. The optical system of claim 2, further comprising an
illumination path and a beam splitter, wherein the illumination
path and the collection path are associated with the objective via
the beam splitter.
4. The optical system of claim 1, further comprising a neutral
density filter and/or mirror configured to control amplitude of the
reference signal.
5. An optical system in an imaging overlay metrology tool, the
optical system comprising an illumination path, a main objective,
and a collection path, wherein the optical system further comprises
a reference path having a reference signal with controllable
amplitude and phase which are configured to minimize a difference
of topographic phases between zeroth and first order diffraction
signals from at least two target layers.
6. The optical system of claim 5, wherein the reference path is
incorporated in the optical system in a Linnik interferometer
configuration.
7. The optical system of claim 5, wherein the minimization is
carried out according to Equation 12.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application is a divisional of U.S. patent application
No. 15/114,175 filed on Jul. 26, 2016, which is a U.S. national
stage application of PCT/US16/33353 filed May 19, 2016, which
claims the benefit of U.S. Provisional Patent Application No.
62/163,783 filed on May 19, 2015 and of U.S. Provisional Patent
Application No. 62/222,724 filed on Sep. 23, 2015, which are
incorporated herein by reference in their entirety.
BACKGROUND OF THE INVENTION
1. TECHNICAL FIELD
[0002] The present invention relates to the field of metrology, and
more particularly, to overlay metrology.
2. DISCUSSION OF RELATED ART
[0003] Current methods for optical overlay measurement rely on two
main technologies: Imaging and Scatterometry. In imaging, the
position of periodic targets is measured in the field of view of
the optical system and the overlay (OVL) is deduced from positions
of targets printed in different layers. Scatterometry utilizes
interference between electromagnetic (EM) waves scattered by
periodic overlay marks (targets with periodic structures) printed
at different layers to deduce the relative displacement of the
layers. In both cases a control on amplitudes and phases of the
diffraction orders of the scattered EM waves may provide a crucial
effect on accuracy and precision of overlay measurement.
SUMMARY OF THE INVENTION
[0004] The following is a simplified summary providing an initial
understanding of the invention. The summary does not necessarily
identify key elements nor limits the scope of the invention, but
merely serves as an introduction to the following description.
[0005] One aspect of the present invention provides metrology tools
and methods, which estimate the effect of topographic phases
corresponding to different diffraction orders, which result from
light scattering on periodic targets, and adjust the measurement
conditions to improve measurement accuracy. In imaging, overlay
error magnification may be reduced by choosing appropriate
measurement conditions based on analysis of contrast function
behavior, changing illumination conditions (reducing spectrum width
and illumination NA), using polarizing targets and/or optical
systems, using multiple defocusing positions etc. On-the-fly
calibration of measurement results may be carried out in imaging or
scatterometry using additional measurements or additional target
cells.
[0006] These, additional, and/or other aspects and/or advantages of
the present invention are set forth in the detailed description
which follows; possibly inferable from the detailed description;
and/or learnable by practice of the present invention.
BRIEF DESCRIPTION OF THE DRAWINGS
[0007] For a better understanding of embodiments of the invention
and to show how the same may be carried into effect, reference will
now be made, purely by way of example, to the accompanying drawings
in which like numerals designate corresponding elements or sections
throughout.
[0008] In the accompanying drawings:
[0009] FIG. 1 is a high level schematic illustration of diffraction
orders in typical imaging based overlay (IBO) metrology, according
to some embodiments of the invention.
[0010] FIG. 2A is a high level schematic illustration of an
approximation of an asymmetric grating with a small side-wall angle
on the right side by a structure that is a sum of symmetric
rectangular areas with different centers, according to some
embodiments of the invention.
[0011] FIGS. 2B and 2C schematically illustrate exemplary
simulation results for the model in FIG. 2A, relating the overlay
error with the defocusing, according to some embodiments of the
invention.
[0012] FIG. 2D schematically illustrates exemplary simulation
results for using a small illumination numerical aperture, relating
the overlay error to the contrast function, according to some
embodiments of the invention.
[0013] FIGS. 3A-3C are high level schematic illustrations of
corresponding optical systems according to some embodiments of the
invention.
[0014] FIGS. 4A-4C are high level schematic illustrations of
optical systems for simultaneous measurement of multiple focus
positions, according to some embodiments of the invention.
[0015] FIG. 5A is a high level schematic illustration of
polarization control targets, according to some embodiments of the
invention.
[0016] FIG. 5B is a high level schematic illustration of an optical
system, according to some embodiments of the invention.
[0017] FIG. 6 is a high level schematic illustration of an optical
system, according to some embodiments of the invention.
[0018] FIG. 7 is a high level schematic illustration of contrast
functions as criteria for topographic phase control, according to
some embodiments of the invention.
[0019] FIG. 8 is a high level schematic illustration of diffraction
orders in typical diffraction based overlay (DBO) metrology,
according to some embodiments of the invention.
[0020] FIG. 9 is a high level schematic illustration of a SCOL
target with auxiliary cells, measured by a metrology tool,
according to some embodiments of the invention.
[0021] FIG. 10 is a high level flowchart illustrating a method,
according to some embodiments of the invention.
DETAILED DESCRIPTION OF THE INVENTION
[0022] In the following description, various aspects of the present
invention are described. For purposes of explanation, specific
configurations and details are set forth in order to provide a
thorough understanding of the present invention. However, it will
also be apparent to one skilled in the art that the present
invention may be practiced without the specific details presented
herein. Furthermore, well known features may have been omitted or
simplified in order not to obscure the present invention. With
specific reference to the drawings, it is stressed that the
particulars shown are by way of example and for purposes of
illustrative discussion of the present invention only, and are
presented in the cause of providing what is believed to be the most
useful and readily understood description of the principles and
conceptual aspects of the invention. In this regard, no attempt is
made to show structural details of the invention in more detail
than is necessary for a fundamental understanding of the invention,
the description taken with the drawings making apparent to those
skilled in the art how the several forms of the invention may be
embodied in practice.
[0023] Before at least one embodiment of the invention is explained
in detail, it is to be understood that the invention is not limited
in its application to the details of construction and the
arrangement of the components set forth in the following
description or illustrated in the drawings. The invention is
applicable to other embodiments that may be practiced or carried
out in various ways as well as to combinations of the disclosed
embodiments. Also, it is to be understood that the phraseology and
terminology employed herein is for the purpose of description and
should not be regarded as limiting.
[0024] Unless specifically stated otherwise, as apparent from the
following discussions, it is appreciated that throughout the
specification discussions utilizing terms such as "processing",
"computing", "calculating", "determining", "enhancing" or the like,
refer to the action and/or processes of a computer or computing
system, or similar electronic computing device, that manipulates
and/or transforms data represented as physical, such as electronic,
quantities within the computing system's registers and/or memories
into other data similarly represented as physical quantities within
the computing system's memories, registers or other such
information storage, transmission or display devices.
[0025] Metrology tools and methods are provided, which estimate the
effect of topographic phases corresponding to different diffraction
orders, which result from light scattering on periodic targets, and
adjust the measurement conditions to improve measurement accuracy.
In imaging, overlay error magnification may be reduced by choosing
appropriate measurement conditions based on analysis of contrast
function behavior, changing illumination conditions (reducing
spectrum width and illumination NA), using polarizing targets
and/or optical systems, using multiple defocusing positions etc.
On-the-fly calibration of measurement results may be carried out in
imaging or scatterometry using additional measurements or
additional target cells.
[0026] Embodiments of the present invention provide efficient and
economical methods and mechanisms for carrying out imaging and/or
scatterometry metrology measurements with better accuracy.
Metrology overlay (OVL) measurement is performed on specially
designed "proxy" metrology targets having typical scales (pitches)
larger than hundreds nanometers. Device design rule pitches are
unresolved by imaging and scatterometry overlay optical tools and
the gap between device pitch (<90nm) and metrology target pitch
increases with time. As the lithography processing steps are
optimized to device scales, the metrology targets are not fully
process compatible which results in all type of target asymmetries
appearing in OVL targets. In most cases the geometrical asymmetry,
like the asymmetry in side wall angles (SWA) of the target edges,
is not large (about 1 nm) and leads to some ambiguity in the
definition of OVL within allowed tolerances. However, both Imaging
and Scatterometry OVL approaches under unsuccessful measurement
conditions may magnify the effect of target asymmetry by order of
magnitude which leads to significant errors in OVL measurements.
Approaches which were considered for improving the accuracy of OVL
measurement comprise (i) seeking a recipe optimization for optimal
measurement conditions without any sufficient measurement tool
modifications; (ii) providing a tool modification which allows OVL
measurement under conditions excluding any magnification of the
target asymmetry effect; (iii) using a two-beam imaging scheme
which solves the problem of target asymmetry magnification but
requires using a blocker in the collection pupil plane, and (iv)
carrying out measurements under special illumination conditions,
discussed below.
[0027] Major parameter controlling both sensitivity and accuracy of
optical Overlay metrologies for both Diffraction Based Overlay
(DBO) and Imaging Based Overlay (IBO) is the phase difference
between EM fields interfering to produce measured signal. FIGS. 8
and 1 are high level schematic illustrations of diffraction orders
in typical DBO and IBO metrology, respectively, according to some
embodiments of the invention. The topographic phase in DBO (FIG. 8,
discussed below) is defined as the mean phase difference between
electromagnetic (EM) fields diffracted by upper and lower gratings
(91B, U and 91A, L respectively) of a SCOL target 90A, to the same
diffraction orders (-1, +1). In IBO case (FIG. 1), the topographic
phase controlling the measurement quality is the mean phase between
zeroth order and the symmetric (e.g., .+-.1.sup.st) diffraction
orders (DO) diffracted from an imaging target 90.
[0028] Both technologies suffer of similar inaccuracy mechanism as
sensitivity to asymmetries of the targets. Such asymmetries, mostly
stemming in incompatibility of large pitch overlay targets to
processes optimized for production of smaller pitches devices, are
manifested as imbalance of both phases and amplitudes of
diffraction orders. The former cannot be distinguished from grating
displacement (overlay), but its effect is limited by simple
geometrical ambiguity. The effect of amplitude imbalance, though,
can be grossly magnified, and is controlled solely by the
interference of the fields constituting the signal. The mechanism
of target asymmetry amplification is described in detail in WIPO
Patent Publication No. PCT/U.S. Ser. No. 15/62523, incorporated
herein by reference in its entirety, and below. In both
technologies the effect of target asymmetry increases when
topographic phase behavior causes a significant signal contrast
reduction for imaging OVL or differential signal reduction for
scatterometry OVL. Accordingly, the control of topographic phase
behavior may play a crucial role in improvement of accuracy of OVL
measurement. In the following, several possibilities are disclosed
for topographic phase control including modifications to
measurement hardware and specialized target design, as well as
various approaches for establishing of the best measurement
conditions required for improvement of OVL measurement
accuracy.
[0029] For example, in the IBO case (FIG. 1), an estimation of the
effect of the phase difference between zero and first diffraction
orders on the accuracy of OVL measurement is provided for an
imaging tool with small NA (numerical aperture, NA<0.2)
illumination conditions (Illumination ray denoted by I). The
positions of diffraction orders in the pupil is shown in FIG. 1,
where .theta..sub.0 denoting the illumination angle and
.theta..sub.1 denoting the angle of the first diffraction order
provided by scattering on a periodic structure (imaging target 90),
having a period P, both with respect to a normal to the target
plane. .theta..sub.1 and .theta..sub.0 are related in Equations 1,
with .lamda. denoting the illumination wavelength.
sin .theta. 1 = - sin .theta. 0 + .lamda. P sin .theta. 1 .apprxeq.
- .theta. 0 + .lamda. P ; Equations 1 cos .theta. 1 = 1 - ( -
.theta. 0 + .lamda. P ) 2 .apprxeq. 1 - ( .lamda. P ) 2 + .theta. 0
.lamda. P 1 - ( .lamda. P ) 2 ##EQU00001##
[0030] Denoting the difference of topographic phases between the
first and the zeroth diffraction orders as .PSI., Equations 2
define the corresponding defocus .DELTA.F needed to compensate the
topographic phase .PSI., and the corresponding topographic phase
spread in the pupil, estimated for the case that for .lamda./P=1/2
and .PSI..about..pi./2 (worst case).
2 .pi. .lamda. .DELTA. F ( 1 - 1 - ( .lamda. P ) 2 ) = .PSI.
.DELTA. F = .lamda..PSI. 2 .pi. ( 1 - 1 - ( .lamda. P ) 2 )
Equations 2 2 .pi. .lamda. .DELTA. F .theta. 0 .lamda. P 1 - (
.lamda. P ) 2 = .theta. 0 .lamda. P .PSI. ( 1 - 1 - ( .lamda. P ) 2
) 1 - ( .lamda. P ) 2 .apprxeq. 2 .theta. 0 .PSI. ( 1 - 3 2 ) 3
.about. 4 .pi. NA ##EQU00002##
[0031] Even for small illumination NA.about.0.2, a large
topographic phase spread of .about..pi. means that when the central
part of the illumination points is in the best contrast position,
the peripheral illumination points are around the zero contrast
position which leads to large magnification of the effect of target
asymmetry on accuracy of OVL measurement and to the lack of
possibility to control the topographic phases of scattered light
may course inaccurate OVL measurement. This effect is drastically
increased with larger illumination NA.
[0032] Advantageously, the disclosed systems and methods overcome
the main disadvantage of the standard imaging tool and
scatterometry tools, namely the uncontrollable magnification of
target asymmetry effect which is present when inappropriate
measurement conditions are used.
[0033] In the following, the target asymmetry contributions to the
accuracy budget are analyzed in detail. The expression for the
amplitude of the electric field in the image plane can be written
as in Equation 4, with f,g being the pupil coordinates, which
relate to the actual dimensional pupil coordinates .zeta.,.eta.
as
f = .xi. R NA .lamda. , g = .eta. R NA .lamda. , ##EQU00003##
where R is the lens radius, E(f,g) is the amplitude of the electric
field in the pupil plane and
e.sub.2.pi.i(1-cos.theta.).DELTA.z/.lamda. describes the effect of
defocusing .DELTA.z on the amplitude of the electric field in the
pupil plane.
E ( x , y ) = .intg. .intg. f 2 + g 2 .ltoreq. NA .lamda. E ( f , g
) e 2 .pi. i ( 1 - cos .theta. ) .DELTA. z / .lamda. e - 2 .pi. i (
fx + gy ) dfdg Equation 4 ##EQU00004##
[0034] In the most simple, non-limiting, optical configuration with
only the .+-.1 and 0 diffraction orders being captured by lens,
Equation 4 can be simplified into Equation 5, with P denoting the
target pitch in the X direction, with a.sub.0, a.sub.1 and a.sub.-1
being complex amplitudes of diffraction orders (depending on
process variations and target asymmetries), GP denoting the grating
position and .DELTA.F denoting the defocus.
E ( x ) = a 0 e i 2 .pi. .DELTA. F .lamda. [ 1 - cos ( .theta. 0 )
] + a 1 e i 2 .pi. P ( x - GP ) + i 2 .pi..DELTA. F .lamda. [ 1 -
cos ( .theta. 1 ) ] + a - 1 e - i 2 .pi. P ( x - GP ) + i 2 .pi.
.DELTA. F .lamda. [ 1 - cos ( .theta. - 1 ) ] Equation 5
##EQU00005##
[0035] FIG. 2A is a high level schematic illustration of an
approximation of an asymmetric grating 92 with a small side-wall
angle on the right side by a structure 94 that is a sum of
symmetric rectangular areas 94A with different centers, according
to some embodiments of the invention. The scattering off asymmetric
grating 92 is approximated as the sum of scatterings from areas 94A
in structure 94.
[0036] The amplitude of each diffraction order is a sum of plane
waves corresponding to scattering from different rectangular areas
94A. For example, Equations 6 express the amplitudes A.sub.(1),
A.sub.(-1) for the .+-.1 first diffraction orders, with
A.sub.0.sup.(1) denoting the amplitude of the first diffraction
order without target asymmetry, A.sub.n.sup.(0) , denoting the
amplitudes from areas 94A and
.delta..sub.Re.sup.(1)+i.delta..sub.Im.sup.(1) denoting the effect
of target asymmetry on the amplitude of the first diffraction
order. .DELTA.n denotes the extension of the n.sup.th area relative
to the nominal area length corresponding to the grating without
asymmetry and .PSI..sub.n denotes the topographic phase of the
n.sup.th area.
A ( 1 ) = n A n ( 1 ) e i .psi. n + 2 .pi. i x + .DELTA. n / 2 P
.apprxeq. n A n ( 1 ) e i .psi. n + 2 .pi. i x P ( 1 + .pi. i
.DELTA. n P ) = A 0 ( 1 ) e i .psi. ( 1 ) + 2 .pi. i x P ( 1 +
.delta. Re ( 1 ) + i .delta. Im ( 1 ) ) Equations 6 A ( - 1 ) = A 0
( - 1 ) e i .psi. ( - 1 ) - 2 .pi. i x P ( 1 - .delta. Re ( - 1 ) -
i .delta. Im ( - 1 ) ) ##EQU00006##
[0037] Under normal illumination condition, Equation 7 expresses
simplifications of Equations 5 and 6, with
sin .theta. 1 = .lamda. P ##EQU00007##
and
.DELTA..psi.=.psi..sup.(1)-.psi..sup.(0)-2.pi.(1-cos.theta..sub.1).DE-
LTA.z/.lamda. and assuming A.sup.(1)=A.sup.(-1),
.psi..sup.(1)=.psi..sup.(-1),
.delta..sub.Im.sup.(1)=.delta..sub.Im.sup.(-1) and the
corresponding field intensity, approximated to the leading order
(neglecting for simplicity the square of first diffraction orders
amplitudes):
Equations 7 ##EQU00008## E ( x ) .about. A ( 0 ) e i .psi. ( 0 ) +
A 0 ( 1 ) e i .psi. ( 1 ) - 2 .pi. i ( 1 - cos .theta. 1 ) .DELTA.
z / .lamda. [ ( 1 + .delta. Re + i .delta. Im ) e 2 .pi. i x P + (
1 - .delta. Re - i .delta. Im ) e - 2 .pi. i x P ] == A ( 0 ) e i
.psi. ( 0 ) + 2 A 0 ( 1 ) e i .psi. ( 1 ) - 2 .pi. i ( 1 - cos
.theta. 1 ) .DELTA. z / .lamda. [ cos ( 2 .pi. x P ) + i ( .delta.
Re + i .delta. Im ) sin ( 2 .pi. x P ) ] .about. .about. A ( 0 ) +
2 A 0 ( 1 ) e i .DELTA. .psi. [ cos ( 2 .pi. x P - .delta. Im ) + i
.delta. Re sin ( 2 .pi. x P ) ] I ( x ) .apprxeq. Constant + 2 A (
0 ) A 0 ( 1 ) cos ( .psi. ( 1 ) - .psi. ( 0 ) - 2 .pi. ( 1 - cos
.theta. 1 ) .DELTA. z / .lamda. ) . cos [ 2 .pi. P x - .delta. Im -
.delta. Re tan ( .psi. ( 1 ) - .psi. ( 0 ) - 2 .pi. ( 1 - cos
.theta. 1 ) .DELTA. z / .lamda. ) ] ##EQU00008.2##
[0038] In terms of Equation 7, due to the target asymmetry there is
a natural ambiguity in the position of the target center, in the
order of magnitude of
.THETA. ( P 2 .pi. { .delta. Re , .delta. Im } ) ##EQU00009##
However, wrong tool measurement conditions may magnify this target
geometrical ambiguity by a factor of
tan(.psi..sup.(1)-.psi..sup.(0)-2.pi.(1-cos.theta..sub.1).DELTA.z/.lamda.-
). This factor is almost zero in the best contrast position but
goes to infinity in the zero contrast position. The value of this
factor can be controlled by a correct choice of the measurement
focus position, however, the following issues are encountered: (i)
Each illumination pupil position and each wavelength provide their
own topographic phase and, correspondingly, their own best contrast
focus position. Since the field image is a sum of images
corresponding to different illumination points and wavelengths it
may collect images for which the target asymmetry effect is
strongly magnified. This first issue can hardly be solved by
changing the measurement focus position since for relatively large
illumination NA the spread of best contrast positions corresponding
to different illumination angles can be as large as micron. Since
the distance between the measured focus position and the best
contrast positions of a part of illumination angles can be as large
as half a micron, there is target accuracy magnification in any
chosen measurement focus position. (ii) The best contrast position
varies with process variations. In case the focus acquisition
procedure provides a measurement focus position which is not
strongly correlated with the best contrast position, it becomes an
additional factor that deteriorates OVL measurement accuracy.
[0039] These issues are exemplified in simulations, assuming
SWA=88.degree. (corresponding to <.+-.1 nm OVL ambiguity for a
layer height <100 nm), resulting in the relations shown in FIG.
2B, in which for certain measurement conditions (e.g., .lamda.) and
measurement focus position, the imaging tool provided OVL error
within the range of .+-.5 nm, i.e., the imaging tool enhanced the
OVL error originating from the SWA by a factor of 5, under the
specific simulated conditions. FIG. 2C illustrates another
simulation example, showing even larger magnification factors.
[0040] Proposed solutions comprise any of the following approaches:
(i) an appropriate choice of measurement conditions for which the
best contrast focus position coincides with grating position (ii)
Reduction of the spectral range and the illumination NA, (iii)
Grabbing several images in different focus positions on each site
and finding the best focus position, (iv) Using large illumination
wavelengths, and (v) simultaneous grabbing of several images in
different focus positions. These approaches are discussed below in
detail.
[0041] Radically reducing the spectral range (e.g., below 10 nm)
and the illumination NA (e.g., below NA.about.0.1) (approach (ii))
reduces the spread of best contrast positions to 200-300 nm. FIG.
2D illustrates simulation results using a small illumination NA of
0.1 in which points 105 are identified as the best contrast
positions with low OVL error magnification and points 95 are
identified as zero contrast positions. As illustrated in FIG. 2D,
the OVL error is small and changes slowly with focus changes around
the best contrast position (105) while the OVL error varies much
more strongly with focus changes around the zero contrast position
(95). Positions 105 and 95 represent different types of zero OVL
error measurement points--in zero contrast position 95 the zero OVL
error results from the fact that amplitude of the first harmonic is
exactly zero and the OVL is measured with the second harmonic,
which is usually much smaller than the amplitude of the first
harmonic. The measurement at zero contrast position 95 is discussed
below and requires hardware (HW) modifications of the metrology
tool (e.g., a zero order blocker in the collection path, as
illustrated in FIG. 3A).
[0042] FIG. 3A is a high level schematic illustration of a
theoretical model and an optical system 110, according to some
embodiments of the invention. Illumination 81 (in illumination path
71) enters system 110 and directed via optics 82 (e.g., a beam
splitter) and objective 83 onto target 90, from which diffraction
signals are collected and directed via beam splitter 82 and optics
84 (e.g., a tube lens) (in collection path 79, also termed
detection path) to a detector 80 (e.g., a CCD--charge coupled
device). A spatial filter 115 may be introduced to block the zeroth
order diffraction signal, ideally at the Fourier plane (pupil plane
120). In a non-limiting manner, only the first diffraction orders
are illustrated.
[0043] System 110 may be configured according to the following
guidelines to reduce or cancel accuracy error magnification: a
low-NA source such as a laser may be used to provide relaxed
requirements for light uniformity; only first order diffraction may
be passed to provide clean two-beam interference and large
depth-of-focus; both process and resist signals (i.e., diffraction
signals from the different target layers) may be passed through the
same part of pupil plane 120 to cancel aberrations; and focusing
may be carried out on the fly, noting that good image contrast
allows measurement without implementing adaptive noise reduction
algorithms (ANRA) to achieve a short MAM (move-acquire-measure)
time, e.g., under 200 msec.
[0044] In particular, the following disclosed algorithmic approach
overcomes difficulties related to the introduction of opaque zero
order blocker 115 and involved in the requirement for good
separation of the diffraction orders in the collection pupil, to
avoid either leakage of the zeroth order into the image, or sharp
truncation of higher DOs. The inventors note that this requirement
limits the minimal target size to be above 10-15 um to provide
small diffraction tails. On the other hand, the very limited size
of zero order blocker 115 requires very small illumination NA
causing a light budget problem. In particular, the reduction of the
illumination NA beyond a particular threshold value (van
Cittert-Zernike theorem) gives rise to spatially extended coherence
effects (ringing) distorting the image. However, the inventors have
found out that the following algorithm, while maintaining the
advantages inherent in the optical scheme illustrated in FIG. 3A,
overcomes the above-listed limitations by not using zero order
blocker 115, possibly requiring minor hardware modifications,
thereby achieving a superior accuracy in imaging OVL measurement,
improves tool performance and reduces effect of process variations.
In particular, the inventors have found out that selecting only
even harmonics for OVL measurement and/or using through-focus
averaging to reduce the contribution of even diffraction orders to
the even signal harmonics--provide these advantages, as explained
below.
[0045] FIGS. 3B and 3C are high level schematic illustrations of
metrology tool 85 and optical system 110, respectively, according
to some embodiments of the invention. Metrology tool 85 is
illustrated schematically as comprising optical system 110 with
illumination path 71 and collection path 79 with corresponding
illumination and collection numerical apertures (NA.sub.IL and
NA.sub.C, respectively), detector 80 and metrology module(s) 87
associated with processor(s) 88. Optical system 110 is illustrated
schematically as in FIG. 3A, but without zero order blocker 115,
and configured for carrying out high topography stack measurements
wherein collection and illumination numerical apertures thereof are
selected to have a sum smaller than 2.lamda./P, and/or optical
system 110 is configured to integrate multiple images captured
thereby over multiple focal positions thereof, to average out
non-symmetric contributions--as explained below. Under such
configurations, metrology tool 85 and optical system 110 mimic the
two-beam configuration disclosed herein above and below, yet avoid
zero order blocking by including only the second harmonic with
specific .lamda./P ratios as explained below.
[0046] The main requirements for the disclosed optical scheme are
(i) adjustable illumination spectral range and collection NA
(NA.sub.C) to make sure that for any chosen target pitch objective
lens 83 collects only the zeroth and .+-.1 (first) diffraction
orders; and (ii) a relatively small illumination NA (to provide
large DOF (depth of focus). Specifically, NA.sub.C and NA.sub.IL
may be selected to satisfy the condition
2.lamda./P>NA.sub.C+NA.sub.IL with .lamda. denoting the
illumination wavelength and P denoting the grating pitch (of target
90). Under such conditions, the second harmonic of the measured
signal is formed as a result of interference between .+-.1
diffraction orders only, and in this sense it is fully equivalent
to the signal measured with zero order blocker 115 (apart from the
precision issues which should be solved by other hardware
means).
[0047] Concerning the DOF, after scattering on target 90, the
oblique plain waves are the .+-.1 diffraction orders propagating at
angles defined by sin(.theta..sub.1,
-1)=-sin(.theta..sub.0).+-..lamda./P, with .theta..sub.0 denoting
the illumination angle. In the case of zero illumination NA
(NA.sub.IL=0) and exact fulfillment of a normal illumination
condition (.theta..sub.0.ident.0), it follows that
cos(.theta..sub.1).ident.cos(.theta..sub.-1) and the relative phase
between the two plane waves doesn't change with focus, i.e.,
corresponding to infinite DOF. Due to the finite size of the
illumination ring, the normal illumination condition can only be
satisfied approximately. However, as in this case the DOF is
determined by the illumination NA (NA.sub.IL) rather than by the
collection NA (NA.sub.C) as in a general case, it can be shown that
the value of DOF for small illumination NA can be approximated
as
DOF .apprxeq. P 1 - ( .lamda. / P ) 2 2.5 NA IL . ##EQU00010##
For example, for P=1800 nm, .lamda./P.about.0.5 and NA.sub.IL=0.2
yields a DOF>3 .mu.m, which allows measurement of high
topography stacks with a single grab.
[0048] Alternatively or complementarily, multiple images may be
integrated over multiple focal positions to average out the
non-symmetric contributions. Deep stack single grab measurements
may be implemented with large DOF of the interference pattern
between .+-.1st orders (or any other symmetric pair of orders). The
averaging utilizes the contrast reversal of interference between
any non-symmetric DOs pair as the object (target 90) moves through
focus, while interferences between symmetric orders do not change
the contrast sign. Integration may be implemented by software
and/or by hardware, allowing focus measurement during the exposure.
In such way, single grab deep stack measurement may be conducted
disregarding the wavelength to pitch ratio and collection NA.
[0049] Concerning the accuracy, as shown above and below, the
mechanism of target asymmetry amplification (the main source of OVL
measurement inaccuracy) is connected to the value of topographic
phase difference between the diffraction orders forming the image.
The most accurate measurement is done when the phase difference
between diffraction orders corresponding to normal illumination
condition is almost zero. This condition is automatically satisfied
for image formed by interference between .+-.1 diffraction orders,
providing high accuracy.
[0050] Returning to the five approaches presented above, grabbing a
few images in different focus positions (e.g., three images as a
non-limiting example) on each site (approach (iii)) may be used to
find the best contrast position on the fly (see last point in
approach (ii)) using, for example, a parabolic approximation of
contrast values calculated for each grabbed image. Since the
inaccuracy magnification factor changes its sign around the best
contrast position (at points 105, 95 in FIG. 2D), an accurate OVL
measurement can be obtained combining OVL values calculated for
images with focus positions on different sides with respect to the
best contrast position using appropriate weights. This approach
provides a solution for site to site process variations.
[0051] Using large illumination wavelengths (approach (iv)) allows
extension of the spectral range and the illumination NA since the
rate of topographic phase change with wavelength and illumination
angle is drastically reduced. This approach requires large pitches,
e.g. around 2000 nm.
[0052] Simultaneous grabbing of several images in different focus
positions (approach (v)) may be used to overcome the difficulty
arising from the fact that since the OVL is measured between two
layers having different best contrast positions, the most accurate
measurement can be achieved if the center of symmetry position for
each layer is measured in its own best contrast position. In this
case the synthetic kernel (or synthetic OVL value as a result of
interpolation to the focus position corresponding to the best focus
position for this layer) is a different combination of signals
corresponding to each one of the grabbed images. Accordingly, the
resulted OVL includes the effect of the stage motion during the
grabbing of images which may affect the precision of OVL
measurement. In order to eliminate this effect from OVL measurement
approach the imaging tool optical configuration may be changed to
allow simultaneous grabbing of several images in different focus
positions. The details of one possible implementation of approach
(v) are described below.
[0053] Grabbing of the targets printed at each one of the layers of
interest in its own best contrast focus (approach (v))
advantageously provides process robust measurements of multilayered
targets, which ensures, under conditions of small enough
illumination NA and sufficiently narrow illumination bandwidth, the
cancelation of topographic phases that are responsible for
sensitivity of the measurements to process variations.
[0054] FIGS. 4A-4C are high level schematic illustrations of
optical systems 110 for simultaneous measurement of multiple focus
positions, according to some embodiments of the invention. FIG. 4A
is a high level schematic illustration of an imaging metrology tool
85 having optical system 110 and a calibration module 112
associated therewith, and possibly operated by one or more
processor(s) (see FIG. 9).
[0055] Calibration module 112 may be configured to derive a
dependency of an overlay error magnification on a level of
defocusing and optical system 110 may be configured to operate at a
narrow spectral range, .DELTA..lamda..ltoreq.10 nm, at a narrow
illumination numerical aperture, NA.ltoreq.0.1, and at a focus
position that corresponds to zero error overlay magnification
according to the derived dependency (see approach (ii) above).
[0056] Optical system 110 may be, alternatively or complementarily,
configured to grab a plurality of metrology target images at a
corresponding plurality of focus positions, and calibration module
112 may be configured to estimate an inaccuracy magnification
factor of the grabbed images and determine a best contrast position
by identifying a sign change of the inaccuracy magnification factor
with respect to the focus positions, with imaging metrology tool 85
being re-configured to be operated at the determined best contrast
position (see approach (iii) above). Calibration module 112 may be
configured to operate on-the-fly during regular metrology tool
operation.
[0057] Optical systems 110 may be configured to enable simultaneous
measurement of target 90 in multiple de-foci (focus locations),
while calibration module 112 in metrology tool 85 (FIG. 4A) may be
configured to provide a particular per-layer grab centering without
risk of compromised precision (see approach (iv) above).
[0058] Optical system 110 may have a first detection focus location
131C of a collection path 79 (schematically illustrated by
objective 83 and optics 84) in metrology tool 85 and comprise at
least two beam splitting elements 132, 134 along collection path 79
which are positioned to provide at least two corresponding
additional focus locations 131A, 131B having different collection
path lengths than first detection focus location 131C.
[0059] For example, optical system 110 may comprise (FIG. 4B) an
optical assembly 130 comprising beam splitters 132, 134 (e.g., BS
30/70 and BS 50/50 respectively) followed by a mirror 136
configured to provide three respective focal locations (foci)
131A-C approximately at detection plane 89. The exemplary
configuration of optical assembly 130 provides three images having
equal power and corresponding to the three different focus
locations, which may be detected by same detector 80 without any
mechanical drift. Optical system 110 may thus be designed to have a
static multi-grab architecture. Quantification of the parameters in
optical system 110, defining .DELTA.Z.sub.o as the longitudinal
displacement of the object, .DELTA.Z.sub.i as the longitudinal
displacement of the image, M as the magnification of the imaging
system and n.sub.o and n.sub.i as the refractive indexes
correspondingly in the object and image media, is provided by
Equations 8 and exemplified for use typical dimensions of metrology
targets 90, namely lateral dimensions of L=30 .mu.m and
magnification of 125, without overlap between images, to provide
the minimal .DELTA.Z.sub.o that can be measured in optical system
110 illustrated in FIG. 4B.
.DELTA. Z i .DELTA. Z o = n i n o M 2 ; L M = .DELTA. Z o = 240 nm
Equations 8 ##EQU00011##
[0060] In another example, optical system 110 may comprise (FIG.
4C) an optical assembly 130 comprising a reticle 140 at a field
stop (e.g., a plane 89A equivalent to detector plane 89), optics
133 and beam splitters 132, 134 configured to provide three
respective focal locations (foci) 131A-C, e.g., detected by three
corresponding detectors 80A-C. Detectors 80A-C may be separate and
capture images at different axial displacements or may be at least
partially unified and enable simultaneous imaging of targets from
different wafer layers (double or more grab).
[0061] Reticle 140 is being illuminated mainly by the specular
reflection of illumination light 81 from the wafer (target 90), so
that the illumination NA of reticle 140 (NA.sub.reticle) can be
estimated in terms of the illumination NA (NA.sub.ill), and the
range of foci (.DELTA.Z.sub.o) coverable with optical system 110
illustrated in FIG. 4C, as expressed in Equations 9, using the
known estimation of the depth of field
.DELTA.Z=.lamda./NA.sub.reticle.sup.2.
NA reticle = NA ill M ; .DELTA. Z o = .DELTA. Z i M 2 = .lamda. NA
ill 2 Equations 9 ##EQU00012##
[0062] For example, using exemplary data .lamda.=700 nm and
NA.sub.ill=0.2 yields de-focus range of .DELTA.Z.sub.0=17.5
.mu.which is satisfactory for covering focal differences between
targets of several layers. Optical systems 110 such as illustrated
in FIG. 4C overcome the requirements for the mechanical stability
of lateral positions of the detector that may be e.g., 100 nm, to
maintain tolerable precision (1 nm) while using separate images to
detect displacement of each one of the separate target layers.
Reticle 140 in the field stop serves as a mutual reference center
for determination of displacements of each one of the layers in the
images.
[0063] Advantageously, proposed configurations of multi-grab
through-focus measurements architecture overcome disadvantages in
prior art autofocus techniques such as (i) measurement of all sites
across wafer at a constant offset from the interferometric focus
and (ii) sequential multi-grab acquisition of multiple images
through focus, followed by separate determination of current and
process layers' targets positions, for later calculation of overlay
between those. Both methodologies suffer significant disadvantages.
Method (i) relies on the uniformity of all process parameters
across the wafer and has been shown to miss the best contrast focus
positions by hundreds of nanometers, which, in turn leads to
multiple nanometers of inaccuracy in overlay determination. Method
(ii) suffers significant precision challenges due to involuntary
drift of stage between image acquisitions, reaching several
nanometers.
[0064] The new optical configuration allow achieving a superior
accuracy in imaging OVL measurement, improves tool performance and
reduces effect of process variations. In particular, the imaging
configuration has a small illumination NA and a narrow spectral
range, uses a new on-the-fly OVL measurement algorithm based on a
few simultaneous images grabbed in different focus positions,
and/or introduces changes in the optical configuration allowing
on-the-fly alignment of images corresponding to different focal
positions. The invention may be implemented in any existing
metrology platform for use in OVL control.
[0065] Returning to FIG. 1, additional ways are disclosed for
gaining topographic phase control in imaging OVL: (i) polarizing
targets, (ii) wavelength control and (iii) interferential
control.
[0066] FIG. 5A is a high level schematic illustration of
polarization control targets 150, according to some embodiments of
the invention. Certain embodiments comprise imaging metrology
target 150 comprising at least one periodic structure having
elements 151 at a pitch p (p.sub.x) along a measurement direction
(X), wherein elements 151 are segmented at an unresolved pitch
(p.sub.y) along a direction (Y) perpendicular to the measurement
direction (X). The unresolved pitch (p.sub.y) is selected to
provide a topographic phase of target 150 which is an integer
multiple of .pi., as explained below.
[0067] Polarization control targets 150 (i) may be designed to
provide a different responses to horizontal and vertical
polarizations of the electromagnetic field (denoted as X and Y with
respect to the target periodicity direction), for example using a
small segmentation pitch Py which is sub-resolved by the
measurement tool.
[0068] For example, in case the axes of the plane of incidence
coincides with the main axes of target 150 the effective
permittivity of the lines (or trenches) 151 segmented with
unresolved pitch P.sub.y may be described using the effective
medium approximation as being effectively equivalent to an
anisotropic film with the directional permittivity vector expressed
in Equations 10, with .epsilon..sub.1 being the permittivity of one
material, .epsilon..sub.2 being the permittivity of another
material, .eta. being the duty cycle of the segmentation in the y
direction and .epsilon..sub.x.noteq..epsilon..sub.y.
x = z = 1 + .eta. 2 1 + .eta. ; y = 1 2 ( 1 + .eta. ) 2 + .eta. 1
Equations 10 ##EQU00013##
[0069] Any of a plurality of possible target designs providing
different responses to vertical and horizontal polarizations of the
electric field may be selected. The actual segmentation pitch may
be selected based on simulations and/or on measurements of
topographic phases on test wafers targets with different
parameters. FIG. 5B is a high level schematic illustration of an
optical system 160, according to some embodiments of the invention.
In optical system 160 of a metrology tool, the direction of the
linear polarization may be used to change the phase between
diffraction orders of the light scattered from target 150. Optical
system 160 may also comprise a polarizer 161 in illumination path
71 (e.g., providing circular light polarization, e.g., using a wave
plate with different angle and retardation parameter) and an
analyzer 169 in the collection path 79 to provide more effective
control of the topographic phases of scattered light by changing
the angle of analyzer 169 in the case of targets responding
differently to different polarization conditions of illumination
light 81.
[0070] Alternatively or complementarily, (ii) the illumination
wavelength may be modified to control the topographic phase. The
inventors have found out that changing the wavelength in ca. 50 nm
for the most stacks provides contrast reversal which is equivalent
to change of .pi. in the topographic phase. Accordingly, even small
shifts of illumination spectrum of ca. .+-.10 nm may be used to
provide significant changes in the topographic phase--making
spectral control an additional factor which may be used to improve
OVL measurement conditions. Moreover, since resist and process
layers can be measured with different wavelengths (double grab as
an example shown above), the topographic phase correction can be
carried out independently for resist and process layers.
[0071] FIG. 6 is a high level schematic illustration of an optical
system 170, according to some embodiments of the invention. Optical
system 170 may be used in imaging metrology tools for OVL
measurement. Optical system 170 comprises an adjustable reference
path 178 having a reference signal integrated in a collection path
79 of optical system 170. Reference path 178 is configured to
provide an adjustable phase of the reference signal optical system
170 is configured to adjust the phase of the reference signal to
modify a topographic phase of an imaging metrology target to be an
integer multiple of .pi.. Adjustable reference path 178 may be
integrated as a Linnik interferometer with a reference objective
174 identical to a main objective 83 of imaging metrology optical
system 170 and an adjustable mirror 175, as explained in detail
below.
[0072] Optical system 170 comprises illumination path 71, main
objective 83 and collection path 79, and further comprises
reference path 178 having a reference signal with controllable
amplitude and phase which are configured to minimize a difference
of topographic phases between zeroth and first order diffraction
signals from at least two target layers of target 90. Illumination
and collection paths 71, 79 may be associated with objective 83 via
beam splitter 82 and reference path 178 may be integrated in
optical system 170 via beam splitter 82. Reference path 178 may
comprise objective 174 identical to main objective 83 with mirror
175, as well as associated illumination source 171 and optics 172,
173 (e.g., focusing lens 172 and beam splitter 173), in exemplary
embodiments in a Linnik interferometer configuration. The amplitude
of the reference signal may by controlled by an attenuator 179,
e.g., by a neutral density (ND) filter and objective 174 and/or
mirror 175 may be moved to control the phase of the reference
signal. The resulting zero diffraction order field appears as a
coherent sum of the zero diffraction orders reflected from the
first and second layers of wafer (gratings of target 90) denoted
Ae.sup.i.alpha. and Be.sup.i.beta.; and the zero diffraction order
reflected from reference mirror 175, denoted Ce.sup.i.gamma..
Accordingly, the zero order fields can be described as expressed in
Equations 11.
A e i .alpha. + C e i .gamma. = A 2 + C 2 + 2 A C cos ( .alpha. -
.gamma. ) e i arctg [ A sin .alpha. + C sin .gamma. A cos .alpha. +
C cos .gamma. ] Equations 11 B e i .beta. + C e i .gamma. = B 2 + C
2 + 2 B C cos ( .beta. - .gamma. ) e i arctg [ B sin .beta. + C sin
.gamma. B cos .beta. + C cos .gamma. ] ##EQU00014##
[0073] Denoting the topographic phases of the first diffraction
orders of the first and second gratings as .phi..sub.1 and
.phi..sub.2, respectively, the amplitude and phase of the reference
signal may be configured to minimize a difference of topographic
phases between zeroth and first order diffraction signals from at
least two target layers of target 90. Resulting from Equations 11,
the best operating condition may be found by minimizing the
expression of Equation 12.
min { [ A sin .alpha. + C sin .gamma. A cos .alpha. + C cos .gamma.
- tg ( .PHI. 1 ) ] 2 + [ B sin .beta. + C sin .gamma. B cos .beta.
+ C cos .gamma. - tg ( .PHI. 2 ) ] 2 } Equation 12 ##EQU00015##
[0074] Reducing the difference of the topographic phases between
first and zero orders for both gratings simultaneously, the best
contrast positions of both layers are provided to be close to each
other, to allow measuring both layers in the same focus
position.
[0075] A complementary approach sets dark field imaging metrology
as departing point. While bright field imaging uses zeroth and
first diffraction order in collection path, dark field imaging
blocks the zeroth order and uses only higher diffraction orders,
typically the first diffraction orders for image formation,
achieving superior precision and accuracy of the overlay
measurement. The common limitation of bright and dark field imaging
is low diffraction efficiency, namely when amplitudes of the EM
waves diffracted by target 90 into the first diffraction orders is
very low. The bright field (BF) and dark field (DF) imaging
intensities are expressed in Equations 13, with I.sub.BF(x)
denoting the intensity observed at detector 80, a.sub.0, a.sub.1
and a.sub.-1 are respectively the amplitudes of the
0.sup.th+1.sup.st and -1.sup.st diffraction orders; .PSI. is the
mean phase at the pupil plane between the 0.sup.th and .+-.1.sup.st
diffraction orders, .delta.a.sub.1 is the difference in the
amplitudes of the positive and negative orders, and .delta..PHI. is
the phase difference.
I BF ( x ) = a 0 2 + a 1 2 + a - 1 2 + 4 a 0 a 1 _ cos [ .PSI. ]
cos [ 2 .pi. ( x - x 0 ) P + .delta. .phi. + .delta. a 1 tan .PSI.
] + a 1 a - 1 cos [ 4 .pi. ( x - x 0 ) P + 2 .delta..phi. ] I DF (
x ) = a 1 2 + a - 1 2 + 2 a 1 a - 1 cos [ 2 2 .pi. ( x - x 0 ) P +
2 .delta. .phi. ] Equations 13 ##EQU00016##
[0076] In bright field imaging, it was shown that the phase
disturbance introduces an error that is limited by the geometrical
ambiguity of the target itself, while the term
.delta.a.sub.1tan.PSI. might introduce errors exceeding several
nanometers, if measured in improper conditions
( .PSI. .fwdarw. .+-. .pi. 2 , ##EQU00017##
see derivations above). While dark field imaging resolves the
inaccuracy problem, it typically suffers significant light
starvation as well as the effect of stray light and ghosts of the
optical system, as its signal consists solely of high diffraction
orders.
[0077] In the case of low diffraction efficiency, the precision of
the grating position measurement in either type of the imaging can
be expressed as in Equations 14, with .alpha. and .beta. being the
noise properties of the light source and the detector,
respectively, and A.sub.0 and A.sub.signal are the amplitudes of
the zeroth modulation frequency and pitch harmonics, respectively.
The general expression is approximated for bright field imaging,
assuming that the intensity at the zeroth order dominates over the
detector noise, and for dark field imaging.
.DELTA. x .varies. PixelSize P .alpha. A 0 + .beta. A Signal
Equation 14 .DELTA. x BF .varies. PixelSize P .alpha. a 0 2 +
.beta. a 0 a 1 _ .apprxeq. PixelSize P .alpha. a 1 _ .DELTA. x DF
.varies. PixelSize P .alpha. a 1 _ 2 + .beta. a 1 _ 2 .apprxeq.
PixelSize P .beta. a 1 _ 2 ##EQU00018##
[0078] Therefore, once the diffraction efficiency of the target
falls to the level in which the signal is dominated by the detector
noise, bright-field measurement is becoming advantageous over
dark-field imaging. Nevertheless, total signal is limited by the
dynamic range of the detector, such that non-saturation of the
camera requires the fulfillment of the condition expressed in
Equation 15, with .GAMMA. denoting the saturation level of the
camera (detector).
|a.sub.0|.sup.2+4a.sub.0a.sub.1<.GAMMA. Equation 15
[0079] Equations 16 express the resulting limitations for the
amplitude of the zeroth order EM field pathing the collection
pupil, in terms of amplitude and intensity.
a 0 2 < a 1 2 [ .GAMMA. a 1 2 + 4 - 2 ] 2 ; I 0 I 1 < (
.GAMMA. I 1 + 4 - 2 ) 2 Equation 16 ##EQU00019##
[0080] This derivation suggests several implementations: (i)
Referring to FIG. 3A, spatial filter 115 may be implemented as a
leaky blocker, for example to control the amplitude of the zeroth
order using an adjustable ND filter as spatial filter 115 while
using narrow illumination NA, to ensure separation between zeroth
and higher diffraction orders. As measurement accuracy require
particular phase relations between zeroth and first diffraction
orders phase control may be implemented by zonal phase plates at
collection pupil area 120. (ii) Referring to FIG. 3A, spatial
filter 115 may be implemented as an adaptive optical element
configured to provide simultaneous phase and amplitude control,
e.g., a DLP (digital light processing device, such as an array of
individually actuated micro-mirrors, or deformable mirror membrane)
providing the phase and amplitude control by surface adjustments
(e.g., mirror angles and device topography). (iii) Referring to
FIG. 6, attenuator 179 may be variable, and implemented according
to the following derivation.
[0081] Equations 17 expresses the zeroth order signal as a sum of
the signals reflected from target 90 and from mirror 175, with
a.sub.w denoting the amplitude of the zeroth order reflected from
the wafer (target 90), and a.sub.r denoting the amplitude of the EM
field reflected from reference mirror 175; and with .PSI. denoting
the topographic phase of the zeroth order with respect to first
diffraction orders, and .PHI..sub.r denoting the phase of the
reference EM field at the pupil plane. The effective zeroth order
signal is expressed using a.sub.0' and .PSI..sub.0'.
E 0 = a w e i .PSI. + a r e i .PHI. r = a 0 ' e i .PSI. ' ;
Equations 17 a 0 ' = a w 2 + a r 2 + 2 a w a r cos ( .PSI. - .PHI.
) ; .PSI. ' = tan - 1 a w sin .PSI. + a r sin .PHI. a w cos .PSI. +
a r cos .PHI. ##EQU00020##
[0082] As a result, using attenuator 179 to control the amplitude
a.sub.r and phase .PHI..sub.r of reference arm 178 provides any
arbitrary zeroth order field to be collected at detector 80 and
according to the principles disclosed above, provides thereby
superior accuracy and precision as well as improved contrast and
signal intensity in imaging-based overlay (IBO) measurements.
[0083] FIG. 7 is a high level schematic illustration of contrast
functions as criteria for topographic phase control, according to
some embodiments of the invention. Contrast functions which are
defined as the grating contrast as a function of the lens focus
position, are presented as practical success criterion for
topographic phase control in imaging. As shown above (see e.g.,
Equations 2), any deviation of the focus lens position from the
grating position leads to an increase of the phase spread between
different illumination angles, resulting in contrast reduction (the
larger the deviation the greater the contrast reduction). The best
measurement condition in imaging corresponds to a phase difference
between zero and first diffraction orders which is integer multiple
of .pi.. In this case the grating focus position coincides with
best contrast position of the measured signal and there is no
spread of topographic phase (induced by defocus) between different
pupil illumination points. Under these conditions, the contrast
function (marked in FIG. 7 by ".pi.") is a symmetric function with
one well distinguished peak. In the opposite, worst measurement
condition, the phase difference between the zeroth and first
diffraction orders is .pi./2 and the equivalent contrast is
achieved in the contrast reversal position corresponding to the
phase difference of .pi./2. The corresponding behavior of the
contrast function has two equal peaks (marked in FIG. 7 by
".pi./2"). Any intermediate condition provides an asymmetric
contrast function (marked in FIG. 7 by "intermediate"). The
contrast function may this be used to optimize the imaging
measurement conditions.
[0084] Metrology tool 85 comprising optical system 110 and
calibration module 112 (see FIG. 4A) may be configured to derive,
through optical system 110, a contrast function of an imaging
target, and adjust measurement conditions of optical system 110 to
modify the derived contrast function to have a single peak (as in
FIG. 7). Metrology tool 85 may be configured to carry out imaging
metrology measurements at the adjusted measurement conditions.
[0085] In scatterometry overlay (SCOL) metrology, similar
considerations are applicable as presented above for imaging
metrology. In particular, similar target design considerations in
conjunction with selections of wavelength and polarization are
likewise applicable. Moreover, as SCOL signal is measured in the
pupil coordinates (angle of incidence multiplexing, at pupil plane
120), post-measurement selection of proper AOI (pupil pixels) is
possible to reach the most accurate measurement conditions.
[0086] For best sensitivity to overlay and accuracy (robustness to
asymmetry and pad-to-pad variations), the preferable phase between
fields is
.+-. .pi. 2 , ##EQU00021##
as shown in the following derivation. FIG. 8 is a high level
schematic illustration of a scatterometry grating-over-grating
target 90, according to some embodiments of the invention. SCOL
target 90 comprises at least two cells 90A ("+cell"), 90B ("-cell")
with opposite offsets +f.sub.0 and -f.sub.0, respectively, of top
grating 91B with respect to bottom grating 91A (grating pitch is
marked by .DELTA.). A cell model 90C is presented for the
diffracted electromagnetic (EM) fields, diffracted by upper and
lower gratings (91B, U and 91A, L respectively) of a SCOL target
cell, at the same diffraction orders (-1, +1). The topographic
phase in DBO (diffraction-based overlay metrology) is defined as
the mean phase difference between electromagnetic (EM) fields
diffracted by the upper and lower gratings. The diffracted orders
are approximated in Equations 18. U.sub..+-.1.sup.+ and
U.sub..+-.1.sup.- denote the total fields scattered by upper
grating 91B of first (positive offset) and second (negative offset)
cells 90A, 90B, respectively. The sign in superscript indicate
cell's offset, the subscript indicates scattering order.
L.sub..+-.1.sup.+ and L.sub..+-.1.sup.- denote the fields
diffracted by lower (process) gratings 91A. u.sub..+-.1.sup..+-.
and l.sub..+-.1.sup..+-. denote amplitudes of the fields,
.psi..sub..+-.1.sup..+-. denotes the topographic phase of the field
diffracted by upper gratings 91B (cell and order as indicated by
corresponding super and sub scripts) and .PHI..sub..+-.1.sup..+-.
denotes the total phase (topographic+Optical Path Difference (OPD)
with respect to upper grating 91B) of the fields diffracted by
lower grating 91A. I.sub..+-.1.sup..+-. and I.sub..+-.1.sup..+-.
denote the corresponding signal intensities. The asterisk (*)
denotes complex conjugation as an operation and c.c. stands for
additional complex conjugated terms.
U .+-. 1 + .apprxeq. u .+-. 1 + e i ( .+-. 2 .pi. ( .DELTA. + f 0 )
P + .psi. .+-. 1 + ) ; L .+-. 1 + = l .+-. 1 + e i .phi. .+-. 1 + ;
Equations 18 U .+-. 1 - .apprxeq. u .+-. 1 - e i ( .+-. 2 .pi. (
.DELTA. - f 0 ) P + .psi. .+-. 1 - ) ; L .+-. 1 - = l .+-. 1 - e i
.phi. .+-. 1 - I .+-. 1 + = u .+-. 1 + 2 + l .+-. 1 + 2 + U .+-. 1
+ .times. L .+-. 1 + * + c . c . = u .+-. 1 + l .+-. 1 + cos ( .+-.
2 .pi. ( .DELTA. + f 0 ) P + .psi. .+-. 1 + - .phi. .+-. 1 + ) I
.+-. 1 - = u .+-. 1 - 2 + l .+-. 1 - 2 + U .+-. 1 - .times. L .+-.
1 - * + c . c . = u .+-. 1 - l .+-. 1 - cos ( .+-. 2 .pi. ( .DELTA.
- f 0 ) P + .psi. .+-. 1 - - .phi. .+-. 1 - ) ##EQU00022##
[0087] Equations 19 introduce and define four physical variables.
In the following derivation constant terms
|u.sub..+-.1.sup..+-.|.sup.2+|l.sub..+-.1.sup..+-.|.sup.2 are
omitted for simplicity, assuming sufficient similarity between
cells and symmetry of the gratings.
.sup..+-.--Mean interference term amplitude of corresponding
cell.
.+-. = u + 1 .+-. l + 1 .+-. + u - 1 .+-. l - 1 .+-. 2
##EQU00023##
.sup..+-.--Asymmetry between amplitudes light scattered to each one
of the orders.
.+-. = u + 1 .+-. l + 1 .+-. - u - 1 .+-. l - 1 .+-. 2
##EQU00024##
.alpha..sup..+-.--Mean Phase difference between EM waves arriving
to detector after scattering at upper and lower gratings.
.alpha. .+-. = ( .psi. + 1 .+-. - .phi. + 1 .+-. ) + ( .psi. - 1
.+-. - .phi. - 1 .+-. ) 2 ##EQU00025##
[0088] .beta..sup..+-.--Asymmetry in phase differences of EM waves
scattered by gratings into either +1.sup.st or -1.sup.st
orders.
.beta. .+-. = ( .psi. + 1 .+-. - .phi. + 1 .+-. ) - ( .psi. - 1
.+-. - .phi. - 1 .+-. ) 2 Equations 19 ##EQU00026##
[0089] Using .sup..+-., .sup..+-., .alpha..sup..+-.,
.beta..sup..+-., the differential signals for cells 90A, 90B are
expressed in Equations 20. The differential signals are expressed
in a general form, disregarding terms stemming from multiple
re-scattering and depending differently on relative displacements
between the gratings, as the relative intensity of these terms
depends on high powers of the diffraction efficiency (DE) of the
gratings which practically does not exceed several percent.
D .+-. = I + 1 .+-. - I - 1 .+-. = .+-. sin [ 2 .pi. ( .DELTA. .+-.
f 0 ) P + .beta. .+-. ] sin .alpha. .+-. + .+-. cos [ 2 .pi. (
.DELTA. .+-. f 0 ) P + .beta. .+-. ] cos .alpha. .+-. = D .+-. =
.+-. sin .alpha. .+-. { sin [ 2 .pi. ( .DELTA. .+-. f 0 ) P +
.beta. .+-. ] + .+-. .+-. tan .alpha. .+-. cos [ 2 .pi. ( .DELTA.
.+-. f 0 ) P + .beta. .+-. ] } ##EQU00027##
Alternative form 1:
D .+-. = ( .+-. sin .alpha. .+-. ) 2 + ( .+-. cos .alpha. .+-. ) 2
.times. sin { .+-. 2 .pi. f 0 P + 2 .pi. P [ .DELTA. + P 2 .pi. (
.beta. .+-. + tan - 1 [ .+-. .+-. tan .alpha. .+-. ] ) ] }
##EQU00028##
Alternative form 2:
D .+-. = .+-. sin .alpha. .+-. 1 + ( .+-. .+-. tan .alpha. .+-. ) 2
.times. sin { .+-. 2 .pi. f 0 P + 2 .pi. P [ .DELTA. + P 2 .pi. (
.beta. .+-. + tan - 1 [ .+-. .+-. tan .alpha. .+-. ] ) ] }
Equations 20 ##EQU00029##
[0090] Since the differential signal at each of the cells
(D.sup..+-.) serves as main observable in first order SCOL,
Equations 20 provide the basis for sensitivity and error
analysis.
[0091] The term
P 2 .pi. ( .beta. .+-. + tan - 1 [ .+-. .+-. tan .alpha. .+-. ] )
##EQU00030##
may be measured as an addition to the overlay due to target
asymmetry. In a way similar to situation in imaging (see Equation
7), phase asymmetry (.beta..sup..+-.) is manifested as a linear
addition to the proper overlay, while the effect of amplitude
asymmetry
( .+-. .+-. ) ##EQU00031##
is amplitied by a combination of topographic phases and OPD of the
orders scattered by upper and lower gratings
( 1 tan .alpha. ) . ##EQU00032##
In the worst possible case of .fwdarw.n.pi.(tan.alpha..fwdarw.0) ,
the error introduced by the amplitude asymmetry might reach
.+-. P 2 . ##EQU00033##
On the otner hand, in optimal situation of
.alpha. .fwdarw. .+-. .pi. 2 ; ##EQU00034##
the amplified error vanishes, as tan.alpha..fwdarw..infin.. It is
noted that the topographic phases yielding best and worst results
are opposite in scatterometry with respect to imaging, as in SCOL
differential signals (difference of the interference terms) are
measured whereas in imaging a sum of the interference terms is
measured.
[0092] Equations 21 estimate the magnitude d of the differential
signal and it's first and second derivate.
Equations 21 ##EQU00035## d = ( sin .alpha. ) 2 + ( cos .alpha. ) 2
= 1 2 ( 2 + 2 ) - 1 2 ( 2 - 2 ) cos 2 .alpha. ; ##EQU00035.2## d '
= d d .alpha. [ ( sin .alpha. ) 2 + ( cos .alpha. ) 2 ] = ( 2 - 2 )
sin 2 .alpha. = 0 .alpha. = n .pi. 2 ; ##EQU00035.3## d '' = d 2 d
.alpha. 2 [ ( sin .alpha. ) 2 + ( cos .alpha. ) 2 ] = 2 ( 2 - 2 )
cos 2 .alpha. ; ##EQU00035.4## assuming < for : .alpha. = 0 ,
.pi. ; d '' > 0 D is at minimum ; for : .alpha. = .+-. .pi. 2 ;
d '' < 0 D is at maximum . ##EQU00035.5##
[0093] Similarly to situation in imaging, the strongest signal
should be obtained at phase relations which are optimal for the
measurement. However, it is difficult to decouple the scattering
efficiency of the gratings from the relative phase between the EM
fields diffracted by upper and lower gratings 91B, 91A. Since
neither nor are known a priori, and since nothing is known about
a.alpha.; no clear behavior of a measurable quantity (e.g., the
magnitude of D.sup..+-. with .alpha.) can be formulated.
[0094] Currently SCOL algorithms calculate the signal at each pixel
using the expression
S = D + + D - D + - D - . ##EQU00036##
Two assumptions are customarily made, that the gratings are
symmetric and the illumination is nearly normal and that target
cells 90A, 90B are identical except for the offset. The model
implications of these assumptions are expressed in Equations
22.
u + 1 .+-. = u - 1 .+-. l + 1 .+-. = l - 1 .+-. .+-. = 0 Equations
22 .psi. + 1 .+-. = .psi. - 1 .+-. .phi. + 1 .+-. = .phi. - 1 .+-.
.beta. .+-. = 0 D .+-. = I + 1 .+-. - I - 1 .+-. = .+-. sin [ 2
.pi. ( .DELTA. .+-. f 0 ) P ] sin .alpha. .+-. + = - .ident. ;
.alpha. + = .alpha. - .ident. .alpha. D .+-. = I + 1 .+-. - I - 1
.+-. = sin [ 2 .pi. ( .DELTA. .+-. f 0 ) P ] sin .alpha. S = D + +
D - D + - D - = tan ( 2 .pi..DELTA. P ) cot ( 2 .pi. f 0 P )
##EQU00037##
[0095] However, in case the gratings have the same shape but their
optical thickness differs, Equations 23 express the model
implications.
Equation 23 ##EQU00038## u + 1 .+-. = u - 1 .+-. l + 1 .+-. = l - 1
.+-. .+-. = 0 ; + = - .ident. ; ##EQU00038.2## .psi. + 1 + = .psi.
- 1 + .noteq. .psi. + 1 - = .psi. - 1 - ; ##EQU00038.3## .phi. + 1
+ = .phi. - 1 + .noteq. .phi. + 1 - = .phi. - 1 - ; .beta. .+-. = 0
; .alpha. + .noteq. .alpha. - ; ##EQU00038.4## D .+-. = I + 1 .+-.
- I - 1 .+-. = .+-. sin [ 2 .pi. ( .DELTA. .+-. f 0 ) P ] sin
.alpha. .+-. ##EQU00038.5## S = D + + D - D + - D - = sin [ 2 .pi.
( .DELTA. + f 0 ) P ] sin .alpha. + + sin [ 2 .pi. ( .DELTA. - f 0
) P ] sin .alpha. - sin [ 2 .pi. ( .DELTA. + f 0 ) P ] sin .alpha.
+ - sin [ 2 .pi. ( .DELTA. - f 0 ) P ] sin .alpha. - ##EQU00038.6##
sin [ 2 .pi. ( .DELTA. + f 0 ) P ] = sin [ 2 .pi. P .DELTA. ] cos [
2 .pi. P f 0 ] .+-. cos [ 2 .pi. P .DELTA. ] sin [ 2 .pi. P f 0 ] S
= D + + D - D + - D - = sin [ 2 .pi. P .DELTA. ] cos [ 2 .pi. P f 0
] { sin .alpha. + + sin .alpha. - } + cos [ 2 .pi. P .DELTA. ] sin
[ 2 .pi. P f 0 ] { sin .alpha. + - sin .alpha. - } sin [ 2 .pi. P
.DELTA. ] cos [ 2 .pi. P f 0 ] { sin .alpha. + - sin .alpha. - } +
cos [ 2 .pi. P .DELTA. ] sin [ 2 .pi. P f 0 ] { sin .alpha. + + sin
.alpha. - } ##EQU00038.7##
[0096] Using the notation
.alpha. = .alpha. + + .alpha. - 2 ; .gamma. = .alpha. + - .alpha. -
2 .alpha. .+-. = .alpha. .+-. .gamma. ; ##EQU00039##
and remembering the physical meaning of the parameters: .DELTA.
being the overlay between two gratings, f.sub.0 being the
intentional offset between the two cells, .alpha. being the phase
(including OPD) between orders scattered by upper and lower
gratings to the same diffraction order, and y.gamma. being mainly
the OPD difference between cells (due to the additional
assumptions), the signal may be further derived as expressed in
Equation 24, resulting in the following simplified expression
S = D + + D - D + - D - = sin [ 2 .pi. P .DELTA. ] cos [ 2 .pi. P f
0 ] sin .alpha.cos .gamma. + cos [ 2 .pi. P .DELTA. ] sin [ 2 .pi.
P f 0 ] cos .alpha. sin .gamma. sin [ 2 .pi. P .DELTA. ] cos [ 2
.pi. P f 0 ] cos .alpha.sin .gamma. + cos [ 2 .pi. P .DELTA. ] sin
[ 2 .pi. P f 0 ] sin .alpha. cos .gamma. = tan [ 2 .pi. P .DELTA. ]
tan .alpha. + tan [ 2 .pi. P f 0 ] tan .gamma. tan [ 2 .pi. P
.DELTA. ] tan .gamma. + tan [ 2 .pi. P f 0 ] tan .alpha. Equation
24 ##EQU00040##
[0097] The implication of assuming different optical thickness
between the gratings results in tan.gamma..noteq.0 , returning
otherwise to Equations 22. While in case |tan .alpha.|>>|tan
.gamma.| a similar approximation holds
( e . g . , when tan .alpha. .fwdarw. .+-. .infin. as .alpha.
.fwdarw. .+-. .pi. 2 ) , ##EQU00041##
such approximation for nearly any
.gamma. .noteq. .+-. .pi. 2 ##EQU00042##
clearly results in neglecting the wrong terms from the standard
model. Moreover, if tan.alpha..fwdarw.0 (termed resonance
conditions below), the measured signal for any .gamma., with
tan.gamma..noteq.0, behaves exactly as S.sup.-1 of the expected
signal S.
[0098] The consequence is that reducing optical thickness
(pad-to-pad) sensitivity and grating asymmetry amplification
effects (see discussion above for Equations 20) may be achieved by
finding conditions for which
.alpha. .apprxeq. .+-. .pi. 2 . ##EQU00043##
However, the aspired conditions of
.alpha. = .+-. .pi. 2 ##EQU00044##
show no special features, neither in pupil image nor in the
differential signal, in contrast to unwanted conditions of
.alpha.=.+-.n.pi. which yield constructive or destructive
interference between orders from the upper and lower gratings, and
are often indicated by clear fringes (either bright or dark) at
pupil images and drastically reduced sensitivity to overlay (dark
fringe in differential signal).
[0099] FIG. 9 is a high level schematic illustration of a SCOL
target 180 with auxiliary cells 185, measured by metrology tool 85,
according to some embodiments of the invention. Metrology tool 85
comprises optical system 110 as well as metrology module(s) 87
associated with one or more processor(s) 88--applicable to any of
the disclosed embodiments, possibly with the disclosed
configurations of optical system(s) 110 and metrology module(s) 87.
Metrology target 90 may comprise at least two cells, each having at
least two target layers with periodic structures having a pitch p
and shifted with respect to each other by an opposite offsets in
the at least two cells. At least two auxiliary, specially designed
measurement cells 185 may be introduced in target 180 in addition
to SCOL target cells 90 to determine the topographic phase on a
pixel-wise basis. For example, two auxiliary cells 185A, 185B
having intentional offsets of a quarter pitch
.+-. P 4 , ##EQU00045##
respectively may be introduced, so that auxiliary cells have the
periodic structures of target 90 shifted by .+-.p/4 with respect to
each other, at the respective auxiliary cells. Pupil images of
auxiliary cells 185 may be used to perform the calculation of W
expressed in Equation 25, with a representing the topographic
phase,
I + 1 1 4 and I - 1 1 4 ##EQU00046##
denoting the intensities of the pupil images measured for the first
and minus first orders (as indicated by subscripts) of auxiliary
cell 185A with offset of
P 4 ##EQU00047##
in the positive overlay direction.
I + 1 - 1 4 and I - 1 - 1 4 , ##EQU00048##
denote the respective intensities for auxiliary cell 185B with
offset
P 4 ##EQU00049##
in the negative overlay affection, I.sub.+1.sup.0 and
I.sub.-1.sup.0 denote the respective intensities for standard SCOL
cell(s) 90 with no special displacement (i.e., with standard
designed offsets with standard .+-.f.sub.0).
W = ( I + 1 1 4 + I - 1 1 4 ) - ( I + 1 - 1 4 + I - 1 - 1 4 ) ( I +
1 0 - I - 1 0 ) = cos .alpha. sin .alpha. .apprxeq. cot .alpha.
Equation 25 ##EQU00050##
[0100] It is noted that
.alpha. = .+-. .pi. 2 ##EQU00051##
in pixels tor wnicn W=0. Therefore, selecting pixels or regions of
detector 80 (corresponding to illumination angle .theta.) and/or
wavelengths .lamda. corresponding to small values of W (e.g.,
W.apprxeq.0) provide significant improvement of accuracy by
suppressing the target's sensitivity to layers' thickness
variations between adjacent cells, as well as the effect of grating
asymmetries.
[0101] FIG. 10 is a high level flowchart illustrating a method 200,
according to some embodiments of the invention. The method stages
may be carried out with respect to the systems and tools described
above, which may optionally be configured to implement method 200.
Method 200 may be at least partially implemented by at least one
computer processor, e.g., in a metrology module. Certain
embodiments comprise computer program products comprising a
computer readable storage medium having computer readable program
embodied therewith and configured to carry out of the relevant
stages of method 200. Certain embodiments comprise target design
files of respective targets designed by embodiments of method
200.
[0102] These method stages are described in more detail with
respect to the systems and tools described above and optionally
configured to implement method 200. Method stages of different
aspects of the invention may be combined according to specified
requirements.
[0103] Method 200 may comprise deriving, in an optical system of an
imaging metrology tool, a dependency of an overlay error
magnification on a level of defocusing (stage 210), and operating
the optical system at a narrow spectral range,
.DELTA..lamda..ltoreq.10 nm, at a narrow illumination numerical
aperture, NA.ltoreq.0.1, and at a focus position that corresponds
to zero overlay error magnification according to the derived
dependency (stage 220).
[0104] Method 200 may comprise grabbing a plurality of metrology
target images at a corresponding plurality of focus positions
(stage 230), estimating an inaccuracy magnification factor of the
grabbed images (stage 250), determining a best contrast position by
identifying a sign change of the inaccuracy magnification factor
with respect to the focus positions (stage 260), and operating the
metrology tool at the determined best contrast position (stage
270). Grabbing 230 may be is carried out simultaneously (stage 240)
and simultaneous grabbing 240 may be carried out by positioning at
least two beam splitting elements along a collection path of the
metrology tool (stage 242) to provide the plurality of focus
positions having different collection path lengths (stage 244). In
certain embodiments, method 200 further comprises using a reticle
at a field plane of the metrology tool, configured as a reference
for target images at the focus locations (stage 246).
[0105] Method 200 may comprise deriving a dependency of a
topographic phase of an imaging metrology target on a measurement
wavelength (stage 280), adjusting the measurement wavelength to
make the topographic phase an integer multiple of .pi. (stage 290),
e.g., in a range of .+-.10 nm, and carrying out imaging metrology
measurements of the imaging metrology target at the adjusted
measurement wavelength (stage 300).
[0106] Method 200 may comprise integrating, in a collection path of
an imaging metrology optical system, an adjustable reference path
comprising a reference signal (stage 320), and adjusting a phase of
the reference signal to modify a topographic phase of an imaging
metrology target to be an integer multiple of .pi. (330). The
adjustable reference path may be configured in the optical system
as a Linnik interferometer with a reference objective identical to
an objective of the imaging metrology optical system and an
adjustable mirror (stage 322). Method 200 may comprise minimizing a
difference of topographic phases between zeroth and first order
diffraction signals from at least two target layers (stage 332),
e.g., according to Equation 12.
[0107] Method 200 may comprise adding to a scatterometry target at
least two auxiliary cells having periodic structures with a same
pitch p as periodic structures in the scatterometry target, shifted
by .+-.p/4 with respect to each other (stage 340), and measuring a
topographic phase of the scatterometry target by measuring
diffraction signals from the at least two auxiliary cells according
to Equation 25 (stage 350).
[0108] Method 200 may comprise deriving a contrast function of an
imaging target (stage 360), adjusting measurement conditions to
modify the derived contrast function to have a single peak (stage
370), and carrying out imaging metrology measurement at the
adjusted measurement conditions (stage 375).
[0109] Method 200 may comprise carrying out high topography stack
measurements (stage 380) by selecting the collection and
illumination numerical apertures (NA's) to have a sum smaller than
2.lamda./P (stage 382), possibly minimizing the illumination NA to
increase the depth of field (DOF) (stage 384); and/or integrating
multiple images over multiple focal positions to average out
non-symmetric contributions (stage 386).
[0110] Any of the data processing stages of method 200 may be
carried out by at least one computer processor (stage 390), such as
processor(s) 88.
[0111] Aspects of the present invention are described above with
reference to flowchart illustrations and/or portion diagrams of
methods, apparatus (systems) and computer program products
according to embodiments of the invention. It will be understood
that each portion of the flowchart illustrations and/or portion
diagrams, and combinations of portions in the flowchart
illustrations and/or portion diagrams, can be implemented by
computer program instructions. These computer program instructions
may be provided to a processor of a general purpose computer,
special purpose computer, or other programmable data processing
apparatus to produce a machine, such that the instructions, which
execute via the processor of the computer or other programmable
data processing apparatus, create means for implementing the
functions/acts specified in the flowchart and/or portion diagram or
portions thereof.
[0112] These computer program instructions may also be stored in a
computer readable medium that can direct a computer, other
programmable data processing apparatus, or other devices to
function in a particular manner, such that the instructions stored
in the computer readable medium produce an article of manufacture
including instructions which implement the function/act specified
in the flowchart and/or portion diagram or portions thereof.
[0113] The computer program instructions may also be loaded onto a
computer, other programmable data processing apparatus, or other
devices to cause a series of operational steps to be performed on
the computer, other programmable apparatus or other devices to
produce a computer implemented process such that the instructions
which execute on the computer or other programmable apparatus
provide processes for implementing the functions/acts specified in
the flowchart and/or portion diagram or portions thereof.
[0114] The aforementioned flowchart and diagrams illustrate the
architecture, functionality, and operation of possible
implementations of systems, methods and computer program products
according to various embodiments of the present invention. In this
regard, each portion in the flowchart or portion diagrams may
represent a module, segment, or portion of code, which comprises
one or more executable instructions for implementing the specified
logical function(s). It should also be noted that, in some
alternative implementations, the functions noted in the portion may
occur out of the order noted in the figures. For example, two
portions shown in succession may, in fact, be executed
substantially concurrently, or the portions may sometimes be
executed in the reverse order, depending upon the functionality
involved. It will also be noted that each portion of the portion
diagrams and/or flowchart illustration, and combinations of
portions in the portion diagrams and/or flowchart illustration, can
be implemented by special purpose hardware-based systems that
perform the specified functions or acts, or combinations of special
purpose hardware and computer instructions.
[0115] In the above description, an embodiment is an example or
implementation of the invention. The various appearances of "one
embodiment", "an embodiment", "certain embodiments" or "some
embodiments" do not necessarily all refer to the same embodiments.
Although various features of the invention may be described in the
context of a single embodiment, the features may also be provided
separately or in any suitable combination. Conversely, although the
invention may be described herein in the context of separate
embodiments for clarity, the invention may also be implemented in a
single embodiment. Certain embodiments of the invention may include
features from different embodiments disclosed above, and certain
embodiments may incorporate elements from other embodiments
disclosed above. The disclosure of elements of the invention in the
context of a specific embodiment is not to be taken as limiting
their use in the specific embodiment alone. Furthermore, it is to
be understood that the invention can be carried out or practiced in
various ways and that the invention can be implemented in certain
embodiments other than the ones outlined in the description
above.
[0116] The invention is not limited to those diagrams or to the
corresponding descriptions. For example, flow need not move through
each illustrated box or state, or in exactly the same order as
illustrated and described. Meanings of technical and scientific
terms used herein are to be commonly understood as by one of
ordinary skill in the art to which the invention belongs, unless
otherwise defined. While the invention has been described with
respect to a limited number of embodiments, these should not be
construed as limitations on the scope of the invention, but rather
as exemplifications of some of the preferred embodiments. Other
possible variations, modifications, and applications are also
within the scope of the invention. Accordingly, the scope of the
invention should not be limited by what has thus far been
described, but by the appended claims and their legal
equivalents.
* * * * *