U.S. patent application number 16/655434 was filed with the patent office on 2020-02-13 for method of measuring a structure, inspection apparatus, lithographic system and device manufacturing method.
This patent application is currently assigned to ASML Netherlands B.V.. The applicant listed for this patent is ASML Netherlands B.V.. Invention is credited to Murat Bozkurt, Grzegorz Grzela, Mohammadreza Hajiahmadi, Martin Jacobus Johan Jak, Lukasz Jerzy Macht, Maurits Van Der Schaar, Patrick Warnaar.
Application Number | 20200050114 16/655434 |
Document ID | / |
Family ID | 62044686 |
Filed Date | 2020-02-13 |
![](/patent/app/20200050114/US20200050114A1-20200213-D00000.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00001.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00002.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00003.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00004.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00005.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00006.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00007.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00008.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00009.png)
![](/patent/app/20200050114/US20200050114A1-20200213-D00010.png)
View All Diagrams
United States Patent
Application |
20200050114 |
Kind Code |
A1 |
Bozkurt; Murat ; et
al. |
February 13, 2020 |
Method of Measuring a Structure, Inspection Apparatus, Lithographic
System and Device Manufacturing Method
Abstract
An overlay metrology target (600, 900, 1000) contains a
plurality of overlay gratings (932-935) formed by lithography.
First diffraction signals (740(1)) are obtained from the target,
and first asymmetry values (As) for the target structures are
derived. Second diffraction signals (740(2)) are obtained from the
target, and second asymmetry values (As') are derived. The first
and second diffraction signals are obtained using different capture
conditions and/or different designs of target structures and/or
bias values. The first asymmetry signals and the second asymmetry
signals are used to solve equations and obtain a measurement of
overlay error. The calculation of overlay error makes no assumption
whether asymmetry in a given target structure results from overlay
in the first direction, in a second direction or in both
directions. With a suitable bias scheme the method allows overlay
and other asymmetry-related properties to be measured accurately,
even in the presence of two-dimensional overlay structure.
Inventors: |
Bozkurt; Murat; (Uden,
NL) ; Van Der Schaar; Maurits; (Eindhoven, NL)
; Warnaar; Patrick; (Tilburg, NL) ; Jak; Martin
Jacobus Johan; ('s-Hertogenbosch, NL) ; Hajiahmadi;
Mohammadreza; (Rotterdam, NL) ; Grzela; Grzegorz;
(Eindhoven, NL) ; Macht; Lukasz Jerzy; (Eindhoven,
NL) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
ASML Netherlands B.V. |
Veldhoven |
|
NL |
|
|
Assignee: |
ASML Netherlands B.V.
Veldhoven
NL
|
Family ID: |
62044686 |
Appl. No.: |
16/655434 |
Filed: |
October 17, 2019 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
15967861 |
May 1, 2018 |
10481506 |
|
|
16655434 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G03F 7/70616 20130101;
G01N 21/4788 20130101; G03F 7/70633 20130101; G01N 21/956 20130101;
G01N 21/9501 20130101; G03F 7/7085 20130101 |
International
Class: |
G03F 7/20 20060101
G03F007/20; G01N 21/95 20060101 G01N021/95; G01N 21/956 20060101
G01N021/956; G01N 21/47 20060101 G01N021/47 |
Foreign Application Data
Date |
Code |
Application Number |
May 8, 2017 |
EP |
17169918 |
Nov 21, 2017 |
EP |
17202806 |
Claims
1. A metrology target for use in overlay metrology, the metrology
target comprising: a plurality of target structures, each target
structure having first features periodic in both a first direction
and a second direction and second features periodic in both the
first direction and the second direction, wherein the first and
second directions are non-parallel, wherein different ones of the
target structures have different programmed offsets in placement of
the second features relative to the first features in both the
first direction and the second direction, and wherein the target
structures are arranged into the metrology target such that any
target structure bordering two neighboring target structures has a
programmed offset intermediate between the programmed offsets of
those two neighboring target structures.
2. The metrology target of claim 1, wherein a bias vector
representing the programmed offset rotates less than ninety degrees
between neighboring target structures.
3. The metrology target of claim 2, wherein a bias vector
representing the programmed offset rotates 45 degrees between
neighboring target structures.
4. The metrology target of claim 1, wherein five or more of the
target structures are arranged in a closed ring. The metrology
target of claim 1, wherein eight or more of the target structures
are arranged in the closed ring.
6. The metrology target of claim 1, wherein five or more of the
target structures are arranged in a line.
7. The metrology target of claim 1, wherein seven or more of the
target structures are arranged in a line.
8. A set of patterning devices for use in a lithographic process,
the patterning devices including at least a first patterning device
configured to define the first features of a metrology target of
claim 1, and a second patterning device configured to define the
second features of the metrology target.
9. The set of patterning devices of claim 8, wherein the first
features of the target structures are formed in a first continuous
array and the second features of the first subset of target
structures are formed in a second continuous array of features, the
different target structures being defined by variation of the
positional offsets over one or other of the continuous arrays.
10. The metrology target of claim 2, wherein five or more of the
target structures are arranged in a closed ring.
11. The metrology target of claim 3, wherein five or more of the
target structures are arranged in a closed ring.
12. The metrology target of claim 2, wherein eight or more of the
target structures are arranged in the closed ring.
13. The metrology target of claim 3, wherein eight or more of the
target structures are arranged in the closed ring.
14. The metrology target of claim 4, wherein eight or more of the
target structures are arranged in the closed ring.
15. The metrology target of claim 2, wherein five or more of the
target structures are arranged in a line.
16. The metrology target of claim 3, wherein five or more of the
target structures are arranged in a line.
17. The metrology target of claim 4, wherein five or more of the
target structures are arranged in a line.
18. The metrology target of claim 5, wherein five or more of the
target structures are arranged in a line.
19. The metrology target of claim 6, wherein seven or more of the
target structures are arranged in a line.
20. A set of patterning devices for use in a lithographic process,
the patterning devices including at least a first patterning device
configured to define the first features of a metrology target of
claim 8, and a second patterning device configured for to define
the second features of the metrology target.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of U.S. patent
application Ser. No. 15/967,861, filed May 1, 2018, which claims
priority of European Application No. 17202806, which was filed on
Nov. 21, 2017, and European Application No. 17169918, which was
filed on May 8, 2017, and are incorporated herein in their entirety
by reference.
BACKGROUND
Field of the Invention
[0002] The present invention relates to methods and apparatus for
metrology usable, for example, in the manufacture of devices by
lithographic techniques, and to methods of manufacturing devices
using lithographic techniques.
Background Art
[0003] A lithographic apparatus is a machine that applies a desired
pattern onto a substrate, usually onto a target portion of the
substrate. A lithographic apparatus can be used, for example, in
the manufacture of integrated circuits (ICs). In that instance, a
patterning device, which is alternatively referred to as a mask or
a reticle, may be used to generate a circuit pattern to be formed
on an individual layer of the IC. This pattern can be transferred
onto a target portion (e.g. including part of a die, one die, or
several dies) on a substrate (e.g., a silicon wafer). Transfer of
the pattern is typically via imaging onto a layer of
radiation-sensitive material (resist) provided on the substrate. In
general, a single substrate will contain a network of adjacent
target portions (known as fields) that are successively
patterned.
[0004] In lithographic processes, it is desirable frequently to
make measurements of the structures created, e.g. for process
control and verification. Various tools for making such
measurements are known, including scanning electron microscopes,
which are often used to measure critical dimension (CD), and
specialized tools to measure overlay, the accuracy of alignment of
two layers in a device. Recently, various forms of scatterometers
have been developed for use in the lithographic field. These
devices direct a beam of radiation onto a target and measure one or
more properties of the scattered radiation--e.g. intensity at a
single angle of reflection as a function of wavelength; intensity
at one or more wavelengths as a function of reflected angle; or
polarization as a function of reflected angle--to obtain a
diffraction "spectrum" from which a property of interest of the
target can be determined.
[0005] Examples of known scatterometers include angle-resolved
scatterometers of the type described in US2006033921A1 and
US2010201963A1. The targets used by such scatterometers are
relatively large gratings, e.g. 40 .mu.m by 40 .mu.m, and the
measurement beam generates a spot that is smaller than the grating
(i.e., the grating is underfilled). In addition to measurement of
feature shapes by reconstruction, diffraction based overlay can be
measured using such apparatus, as described in published patent
application US2006066855A1. Diffraction-based overlay metrology
using dark-field imaging of the diffraction orders enables
measurement of overlay and other parameters on smaller targets.
These targets can be smaller than the illumination spot and may be
surrounded by product structures on a substrate. The intensities
from the environment product structures can efficiently be
separated from the intensities from the overlay target with the
dark-field detection in the image-plane.
[0006] Examples of dark field imaging metrology can be found in
patent applications US20100328655A1 and US2011069292A1 which
documents are hereby incorporated by reference in their entirety.
Further developments of the technique have been described in
published patent publications US20110027704A, US20110043791A,
US2011102753A1, US20120044470A, US20120123581A, US20120242970A1,
US20130258310A, US20130271740A and WO2013178422A1. Typically, in
these methods, it is desired to measure asymmetry as a property of
the target. Targets can be designed so that measurement of
asymmetry can be used to obtain measurement of various performance
parameters such as overlay, focus or dose. Asymmetry of the target
is measured by detecting differences in intensity between opposite
portions of the diffraction spectrum using the scatterometer. For
example, the intensities of +1 and -1 diffraction orders may be
compared, to obtain a measure of asymmetry.
[0007] In order to reduce measurement time, known apparatuses for
dark-field metrology have apertures and detection systems
configured to detect simultaneously the radiation diffracted from
component gratings in both X and Y directions, and to detect these
different directions of diffraction independently. Thus, the need
for separate detection steps in X and Y orientation is avoided.
Examples of such techniques are included in the prior patent
publications mentioned above, and also for example in unpublished
patent application EP16157503.0.
[0008] There is often a desire for grating structures in metrology
targets of the type described to be segmented in a direction other
than their main direction of periodicity. Reasons for this
segmentation may be to induce asymmetry-related effects to allow
measurement of properties other than overlay by the same technique.
Other reasons for this segmentation may be to make the grating
structures more "product-like", so that they are printed with
patterning performance more like the product structures that are
primarily of interest. Grating structures may simply be completely
two-dimensional in layout, for example to resemble an array of
contact holes or pillars. Nevertheless, overlay or other parameters
of the performance of the patterning process are normally
controlled and measured separately in two or more directions,
typically the X and Y directions defined relative to the
substrate.
[0009] A particular problem arises when target structures have
segmentation or other two-dimensional character in both sets of
features (in both layers). Unfortunately, where the grating
structures in a metrology target are two-dimensionally structured,
either being fully two-dimensional gratings or having some kind of
segmentation in the orthogonal to their main direction of
periodicity, diffraction by a structure in the orthogonal direction
becomes mixed with diffraction in the main direction, and the
separate measurements become subject to noise or cross-talk.
Moreover, in such targets, overlay error in two different
directions will influence the diffraction signals captured by the
inspection apparatus. The known methods tend to assume that each
target structure has asymmetry only in a primary direction. When
this assumption is no longer valid, known techniques inevitably
become less accurate. To exacerbate this problem, in general it may
not even be known to the operator of the metrology apparatus,
whether metrology targets under investigation have two-dimensional
properties of the type described.
SUMMARY OF THE INVENTION
[0010] The present invention in a first aspect aims to allow
efficient measurement of a performance parameter such as overlay,
even when target structures may be two-dimensional in nature. The
present invention in another aspect aims to allow recognition of
two-dimensional character in metrology targets, without relying on
advance information.
[0011] The invention in a first aspect provides a method of
determining overlay performance of a lithographic process, the
method including the following steps:
[0012] (a) obtaining a plurality of target structures that have
been formed by the lithographic process, each target structure
comprising a set of first features arranged periodically in at
least a first direction and a set of second features arranged
periodically in at least the first direction and being subject to
overlay error in the placement of the second features relative to
the first features,
[0013] (b) using a detection system to capture first diffraction
signals comprising selected portions of radiation diffracted by at
least a subset of the target structures;
[0014] (c) using the detection system to capture second diffraction
signals comprising selected portions of radiation diffracted by at
least a subset of the overlay targets;
[0015] (d) processing asymmetry information derived from the first
diffraction signals and the second diffraction signals to calculate
at least a measurement of said overlay error in at least the first
direction,
[0016] wherein said target structures have been formed with
programmed offsets in the placement of the second features relative
to the first features in addition to said overlay error, the
programmed offsets within each subset differing in both the first
direction and in a second direction, the first and second
directions being non-parallel,
[0017] and wherein the calculation of overlay error in step (d)
combines said asymmetry information with knowledge of said
programmed offsets while making no assumption whether asymmetry in
a given target structure results from relative displacement of the
second features in the first direction, in the second direction or
both directions.
[0018] With a suitable bias scheme the method allows overlay and
other asymmetry-related properties to be measured accurately, even
in the presence of (potentially unknown) two-dimensional structure
and unknown overlay in two directions. Additional sets of
diffraction signals can be added, if desired, to enhance accuracy
further.
[0019] In a first embodiment, the first and second diffraction
signals are captured under different capture conditions. Capture
conditions may differ for example in one wavelength, polarization,
and/or angular distribution of radiation used for illumination
and/or detection of the target structures.
[0020] In a second embodiment, first diffraction signals comprise
radiation diffracted by a first subset of target structures and the
second diffraction signals comprise radiation diffracted by a
second subset of target structures, different from the first subset
of target structures. The target structures of said first subset
and the target structures of said second subset may for example
differ in one or more of pitch, feature size, relative placement,
and segmentation in the second direction.
[0021] In a third embodiment, the first and second subsets of
target structures of similar design are printed in one step, with
more than four different combinations of programmed offsets. Seven
or eight different programmed offsets may be included in a
composite metrology target.
[0022] The first, second and third embodiments can be combined, if
desired.
[0023] The invention further provides an inspection apparatus for
determining overlay performance of a lithographic process, the
inspection apparatus comprising:
[0024] a support for a substrate on which are provided a plurality
of target structures that have been formed by the lithographic
process, each target structure comprising a set of first features
arranged periodically in at least a first direction and a set of
second features arranged periodically in at least the first
direction and being subject to overlay error in the placement of
the second features relative to the first features,
[0025] an illumination system and a detection system which are
together operable to capture first diffraction signals comprising
selected portions of radiation diffracted by at least a subset of
the target structures and second diffraction signals comprising
selected portions of radiation diffracted by at least a subset of
the overlay targets;
[0026] a processor for processing asymmetry information derived
from the first diffraction signals and the second diffraction
signals to calculate at least a measurement of said overlay error
in at least the first direction,
[0027] wherein said processor is operable on the basis that said
target structures have been formed with programmed offsets in the
placement of the second features relative to the first features in
addition to said overlay error, the programmed offsets within each
subset differing in both the first direction and in a second
direction, the first and second directions being non-parallel,
[0028] and said processor is arranged to calculate overlay error by
combining said asymmetry information with knowledge of said
programmed offsets while making no assumption whether asymmetry in
a given target structure results from relative displacement of the
second features in the first direction, in the second direction or
both directions.
[0029] The inspection apparatus can be implemented applying optical
systems and techniques known from the prior art, or using new
apparatus. The inspection apparatus can be implemented for example
using the above-mentioned dark-field imaging techniques, thereby
obtaining the first and/or second diffraction signals for a
plurality of target structures in a single image.
[0030] The invention in another aspect provides a metrology target
for use in a method according to the first aspect of the invention
as set forth above, wherein said metrology target includes at least
four target structures, each target structure comprising first
features periodic in both a first direction and a second direction
and second features periodic in both the first direction and the
second direction, the first and second directions being
non-parallel, and wherein said target structures have programmed
offsets in placement of the second features relative to the first
features in both the first direction and the second direction, each
target structure within said at least four target structures having
a different combination of programmed offset in the first and
second directions.
[0031] The invention in a further, independent aspect provides a
metrology target for use in overlay metrology, said metrology
target including a plurality of target structures, each target
structure comprising first features periodic in both a first
direction and a second direction and second features periodic in
both the first direction and the second direction, the first and
second directions being non-parallel, and wherein different ones of
said target structures have different programmed offsets in
placement of the second features relative to the first features in
both the first direction and the second direction, and wherein said
target structures are arranged into said metrology target such that
any target structure bordering two neighboring target structures
has a programmed offset intermediate between the programmed offsets
of those two neighboring target structures.
[0032] The invention in a further aspect provides a set of
patterning devices for use in a lithographic process, the
patterning devices including at least a first patterning device
configured to define the first features of a metrology target
according to any aspect of the invention as set forth above, and a
second patterning device configured for to define the second
features of the metrology target.
[0033] The invention in another aspect provides a processing device
arranged to receive at least first and second diffraction signals
captured from a plurality of target structures and to derive a
measurement of overlay error in at least a first direction by
performing the step (d) in the method according to the first aspect
of the invention as set forth above.
[0034] The invention further provides one or more computer program
products comprising machine readable instructions for causing a
programmable processing device to implement one or more aspects of
the invention as set forth above. The machine readable instructions
may be embodied, for example, in a non-transitory storage
medium.
[0035] The machine readable instructions may be further arranged to
cause the programmable processing device to control automatically
the operation of an inspection apparatus to cause capture of the
first and second diffraction signals by steps (b) and (c) of the
method.
[0036] The invention further provides a lithographic system
including a lithographic apparatus and an inspection apparatus
according to the second aspect of the invention, as set forth
above.
[0037] The invention further provides a method of manufacturing
devices wherein a device pattern is applied to a series of
substrates using a lithographic process, the method including
measuring one or more performance parameters using a plurality of
target structures formed as part of or beside said device pattern
on at least one of said substrates using a method according to the
invention as set forth above, and controlling the lithographic
process for later substrates in accordance with the result of the
measuring.
[0038] Further features and advantages of the invention, as well as
the structure and operation of various embodiments of the
invention, are described in detail below with reference to the
accompanying drawings. It is noted that the invention is not
limited to the specific embodiments described herein. Such
embodiments are presented herein for illustrative purposes only.
Additional embodiments will be apparent to persons skilled in the
relevant art(s) based on the teachings contained herein.
BRIEF DESCRIPTION OF THE DRAWINGS/FIGURES
[0039] Embodiments of the invention will now be described, by way
of example, with reference to the accompanying drawings in
which:
[0040] FIG. 1 depicts a lithographic apparatus together with other
apparatuses forming a production facility for semiconductor
devices;
[0041] FIGS. 2(a)-2(b) illustrate schematically 2(a) an inspection
apparatus adapted to perform angle-resolved scatterometry and
dark-field imaging inspection methods in accordance with some
embodiments of the invention and 2(b) an enlarged detail of the
diffraction of incident radiation by a target grating in the
apparatus of FIG. 2(a);
[0042] FIGS. 3(a)-3(c) illustrate 3(a) a segmented illumination
profile, 3(b) the production of diffraction signals in different
directions under the segmented illumination profile and 3(c) the
layout of a prism device in a segmented detection system, all in
the operation of one embodiment of the inspection apparatus of FIG.
2;
[0043] FIGS. 4(a)-4(b) illustrate a composite metrology target
including a number of component gratings 4(a) in a case where each
component grating is periodic in only one direction and 4(b) in a
case where each component grating is or may be periodic in two
directions;
[0044] FIG. 5 illustrates a multiple image of the target of FIG. 4,
captured by the apparatus of FIG. 4 with spatial separation of
diffraction orders;
[0045] FIGS. 6(a)-6(c) illustrate an example target layout
according to a first embodiment of the present disclosure, in plan
view and with cross-sections along lines B and C;
[0046] FIG. 7 illustrates dark-field images of the target of FIG. 6
obtained using first and second measurement conditions in a method
according to the first embodiment of the present disclosure;
[0047] FIG. 8 is a flowchart of a method of measuring a property of
a target structure and a method of controlling a lithographic
process using the principles of the present disclosure; and
[0048] FIGS. 9(a)-9(b) illustrates 9(a) a target layout similar to
that of FIGS. 6 and 9(b) implementation of part of the method of
FIG. 8 using such a target in accordance with the first embodiment
of the present disclosure;
[0049] FIGS. 10(a)-10(b) illustrate 10(a) a target layout according
to a second embodiment of the present disclosure and 10(b)
implementation of part of the method of FIG. 8 using such a target
in accordance with the second embodiment of the present
disclosure;
[0050] FIG. 11 illustrates a dark-field image of the target of FIG.
6 obtained using first and second target types in a method
according to the second embodiment of the present disclosure;
[0051] FIGS. 12(a)-12(c) illustrate 12(a) a metrology target
according to a modified first embodiment of the present disclosure,
12(b) one set of features defined in the target layout and 12(c)
detail of a central portion of the target layout circled in
12(a);
[0052] FIG. 13 illustrates a variant of the target of FIG. 12,
including transition zones;
[0053] FIGS. 14(a)-14(b) illustrate 14(a) a metrology target
according to a third embodiment of the present disclosure, and
14(b) part of a multiple image of the target, captured by the
apparatus of FIG. 4 with spatial separation of diffraction orders,
with a schematic representation of signal processing to obtain
asymmetry signals from a plurality of target structures;
[0054] FIG. 15 illustrates implementation of part of the method of
FIG. 8 using such a target in accordance with the third embodiment
of the present disclosure;
[0055] FIGS. 16(a)-16(b) illustrate 16(a) an enlarged metrology
target according to a modified third embodiment of the present
disclosure, and 16(b) part of a multiple image of the target,
captured by the apparatus of FIG. 4 with spatial separation of
diffraction orders, with a schematic representation of signal
processing to obtain asymmetry signals from a plurality of target
structures; and
[0056] FIGS. 17(a)-17(b) illustrate alternative groupings 17(a) and
17(b) of target structures in the embodiment of FIG. 16.
DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS
[0057] Before describing embodiments of the invention in detail, it
is instructive to present an example environment in which
embodiments of the present invention may be implemented.
[0058] FIG. 1 at 100 shows a lithographic apparatus LA as part of
an industrial facility implementing a high-volume, lithographic
manufacturing process. In the present example, the manufacturing
process is adapted for the manufacture of semiconductor products
(integrated circuits) on substrates such as semiconductor wafers.
The skilled person will appreciate that a wide variety of products
can be manufactured by processing different types of substrates in
variants of this process. The production of semiconductor products
is used purely as an example which has great commercial
significance today.
[0059] Within the lithographic apparatus (or "litho tool" 100 for
short), a measurement station MEA is shown at 102 and an exposure
station EXP is shown at 104. A control unit LACU is shown at 106.
In this example, each substrate visits the measurement station and
the exposure station to have a pattern applied. In an optical
lithographic apparatus, for example, a projection system is used to
transfer a product pattern from a patterning device MA onto the
substrate using conditioned radiation and a projection system. This
is done by forming an image of the pattern in a layer of
radiation-sensitive resist material.
[0060] The term "projection system" used herein should be broadly
interpreted as encompassing any type of projection system,
including refractive, reflective, catadioptric, magnetic,
electromagnetic and electrostatic optical systems, or any
combination thereof, as appropriate for the exposure radiation
being used, or for other factors such as the use of an immersion
liquid or the use of a vacuum. The patterning MA device may be a
mask or reticle, which imparts a pattern to a radiation beam
transmitted or reflected by the patterning device. Well-known modes
of operation include a stepping mode and a scanning mode. As is
well known, the projection system may cooperate with support and
positioning systems for the substrate and the patterning device in
a variety of ways to apply a desired pattern to many target
portions across a substrate. Programmable patterning devices may be
used instead of reticles having a fixed pattern. The radiation for
example may include electromagnetic radiation in the deep
ultraviolet (DUV) or extreme ultraviolet (EUV) wavebands. The
present disclosure is also applicable to other types of
lithographic process, for example imprint lithography and direct
writing lithography, for example by electron beam.
[0061] The lithographic apparatus control unit LACU controls the
movements and measurements of various actuators and sensors,
causing the apparatus LA to receive substrates W and reticles MA
and to implement the patterning operations. LACU also includes
signal processing and data processing capacity to implement desired
calculations relevant to the operation of the apparatus. In
practice, control unit LACU will be realized as a system of many
sub-units, each handling the real-time data acquisition, processing
and control of a subsystem or component within the apparatus.
[0062] Before the pattern is applied to a substrate at the exposure
station EXP, the substrate is processed in at the measurement
station MEA so that various preparatory steps may be carried out.
The preparatory steps may include mapping the surface height of the
substrate using a level sensor and measuring the position of
alignment marks on the substrate using an alignment sensor. The
alignment marks are arranged nominally in a regular grid pattern.
However, due to inaccuracies in creating the marks and also due to
deformations of the substrate that occur throughout its processing,
the marks deviate from the ideal grid. Consequently, in addition to
measuring position and orientation of the substrate, the alignment
sensor in practice must measure in detail the positions of many
marks across the substrate area, if the apparatus is to print
product features at the correct locations with very high accuracy.
The apparatus may be of a so-called dual stage type which has two
substrate tables, each with a positioning system controlled by the
control unit LACU. While one substrate on one substrate table is
being exposed at the exposure station EXP, another substrate can be
loaded onto the other substrate table at the measurement station
MEA so that various preparatory steps may be carried out. The
measurement of alignment marks is therefore very time-consuming and
the provision of two substrate tables enables a substantial
increase in the throughput of the apparatus. If the position sensor
IF is not capable of measuring the position of the substrate table
while it is at the measurement station as well as at the exposure
station, a second position sensor may be provided to enable the
positions of the substrate table to be tracked at both stations.
Lithographic apparatus LA for example is of a so-called dual stage
type which has two substrate tables WTa and WTb and two
stations--an exposure station and a measurement station--between
which the substrate tables can be exchanged.
[0063] Within the production facility, apparatus 100 forms part of
a "litho cell" or "litho cluster" that contains also a coating
apparatus 108 for applying photosensitive resist and other coatings
to substrates W for patterning by the apparatus 100. At an output
side of apparatus 100, a baking apparatus 110 and developing
apparatus 112 are provided for developing the exposed pattern into
a physical resist pattern. Between all of these apparatuses,
substrate handling systems take care of supporting the substrates
and transferring them from one piece of apparatus to the next.
These apparatuses, which are often collectively referred to as the
"track", are under the control of a track control unit which is
itself controlled by a supervisory control system SCS, which also
controls the lithographic apparatus via lithographic apparatus
control unit LACU. Thus, the different apparatuses can be operated
to maximize throughput and processing efficiency. Supervisory
control system SCS receives recipe information R which provides in
great detail a definition of the steps to be performed to create
each patterned substrate.
[0064] Once the pattern has been applied and developed in the litho
cell, patterned substrates 120 are transferred to other processing
apparatuses such as are illustrated at 122, 124, 126. A wide range
of processing steps is implemented by various apparatuses in a
typical manufacturing facility. For the sake of example, apparatus
122 in this embodiment is an etching station, and apparatus 124
performs a post-etch annealing step. Further physical and/or
chemical processing steps are applied in further apparatuses, 126,
etc. Numerous types of operation can be required to make a real
device, such as deposition of material, modification of surface
material characteristics (oxidation, doping, ion implantation
etc.), chemical-mechanical polishing (CMP), and so forth. The
apparatus 126 may, in practice, represent a series of different
processing steps performed in one or more apparatuses.
[0065] As is well known, the manufacture of semiconductor devices
involves many repetitions of such processing, to build up device
structures with appropriate materials and patterns, layer-by-layer
on the substrate. Accordingly, substrates 130 arriving at the litho
cluster may be newly prepared substrates, or they may be substrates
that have been processed previously in this cluster or in another
apparatus entirely. Similarly, depending on the required
processing, substrates 132 on leaving apparatus 126 may be returned
for a subsequent patterning operation in the same litho cluster,
they may be destined for patterning operations in a different
cluster, or they may be finished products to be sent for dicing and
packaging.
[0066] Each layer of the product structure requires a different set
of process steps, and the apparatuses 126 used at each layer may be
completely different in type. Further, even where the processing
steps to be applied by the apparatus 126 are nominally the same, in
a large facility, there may be several supposedly identical
machines working in parallel to perform the step 126 on different
substrates. Small differences in set-up, or faults between these
machines can mean that they influence different substrates in
different ways. Even steps that are relatively common to each
layer, such as etching (apparatus 122) may be implemented by
several etching apparatuses that are nominally identical but
working in parallel to maximize throughput. In practice, moreover,
different layers require different etch processes, for example
chemical etches, plasma etches, according to the details of the
material to be etched, and special requirements such as, for
example, anisotropic etching.
[0067] The previous and/or subsequent processes may be performed in
other lithography apparatuses, as just mentioned, and may even be
performed in different types of lithography apparatus. For example,
some layers in the device manufacturing process which are very
demanding in parameters such as resolution and overlay may be
performed in a more advanced lithography tool than other layers
that are less demanding. Therefore some layers may be exposed in an
immersion type lithography tool, while others are exposed in a
`dry` tool. Some layers may be exposed in a tool working at DUV
wavelengths, while others are exposed using EUV wavelength
radiation.
[0068] In order that the substrates that are exposed by the
lithographic apparatus are exposed correctly and consistently, it
is desirable to inspect exposed substrates to measure properties
such as overlay errors between subsequent layers, line thicknesses,
critical dimensions (CD), etc. Accordingly a manufacturing facility
in which litho cell LC is located also includes metrology system
MET which receives some or all of the substrates W that have been
processed in the litho cell. Metrology results are provided
directly or indirectly to the supervisory control system (SCS) 138.
If errors are detected, adjustments may be made to exposures of
subsequent substrates, especially if the metrology can be done soon
and fast enough that other substrates of the same batch are still
to be exposed. Also, already exposed substrates may be stripped and
reworked to improve yield, or discarded, thereby avoiding
performing further processing on substrates that are known to be
faulty. In a case where only some target portions of a substrate
are faulty, further exposures can be performed only on those target
portions which are good.
[0069] Also shown in FIG. 1 is a metrology apparatus 140 which is
provided for making measurements of parameters of the products at
desired stages in the manufacturing process. A common example of a
metrology apparatus in a modern lithographic production facility is
a scatterometer, for example an angle-resolved scatterometer or a
spectroscopic scatterometer, and it may be applied to measure
properties of the developed substrates at 120 prior to etching in
the apparatus 122. Using metrology apparatus 140, it may be
determined, for example, that important performance parameters such
as overlay or critical dimension (CD) do not meet specified
accuracy requirements in the developed resist. Prior to the etching
step, the opportunity exists to strip the developed resist and
reprocess the substrates 120 through the litho cluster. As is also
well known, the metrology results 142 from the apparatus 140 can be
used to maintain accurate performance of the patterning operations
in the litho cluster, by supervisory control system SCS and/or
control unit LACU 106 making small adjustments over time, thereby
minimizing the risk of products being made out-of-specification,
and requiring re-work. Of course, metrology apparatus 140 and/or
other metrology apparatuses (not shown) can be applied to measure
properties of the processed substrates 132, 134, and incoming
substrates 130.
Example Inspection Apparatus
[0070] FIG. 2(a) shows schematically the key elements of an
inspection apparatus implementing so-called dark field imaging
metrology. The apparatus may be a stand-alone device or
incorporated in either the lithographic apparatus LA, e.g., at the
measurement station, or the lithographic cell LC. An optical axis,
which has several branches throughout the apparatus, is represented
by a dotted line 0. A target grating structure T and diffracted
rays are illustrated in more detail in FIG. 2(b).
[0071] As described in the prior applications cited in the
introduction, the dark-field-imaging apparatus of FIG. 2(a) may be
part of a multi-purpose angle-resolved scatterometer that may be
used instead of, or in addition to, a spectroscopic scatterometer.
In this type of inspection apparatus, radiation emitted by a
radiation source 11 is conditioned by an illumination system 12.
For example, illumination system 12 may include a collimating lens
system 12a, a color filter 12b, a polarizer 12c and an aperture
device 13. The conditioned radiation follows an illumination path
IP, in which it is reflected by partially reflecting surface 15 and
focused into a spot S on substrate W via an objective lens 16. A
metrology target T may be formed on substrate W. The objective lens
16 may be similar in form to a microscope objective lens, but has a
high numerical aperture (NA), preferably at least 0.9 and more
preferably at least 0.95. Immersion fluid can be used to obtain
numerical apertures over 1 if desired.
[0072] The objective lens 16 in this example serves also to collect
radiation that has been scattered by the target. Schematically, a
collection path CP is shown for this returning radiation. The
multi-purpose scatterometer may have two or more measurement
branches in the collection path. The illustrated example has a
pupil imaging branch comprising pupil imaging optical system 18 and
pupil image sensor 19. An imaging branch is also shown, which will
be described in more detail below. Additionally, further optical
systems and branches will be included in a practical apparatus, for
example to collect reference radiation for intensity normalization,
for coarse imaging of capture targets, for focusing and so forth.
Details of these can be found in the prior publications mentioned
above.
[0073] Where a metrology target T is provided on substrate W, this
may be a 1-D grating, which is printed such that, after
development, the bars are formed of solid resist lines. The target
may be a 2-D grating, which is printed such that after development,
the grating is formed of solid resist pillars or vias in the
resist. The bars, pillars or vias may alternatively be etched into
the substrate. Each of these gratings is an example of a target
structure whose properties may be investigated using the inspection
apparatus. In the case of gratings, the structure is periodic. In
the case of an overlay metrology target, the grating is printed on
top of or interleaved with another grating that has been formed by
a previous patterning step.
[0074] The various components of illumination system 12 can be
adjustable to implement different metrology `recipes` within the
same apparatus. In addition to selecting wavelength (color) and
polarization as characteristics of the illuminating radiation,
illumination system 12 can be adjusted to implement different
illumination profiles. The plane of aperture device 13 is conjugate
with a pupil plane of objective lens 16 and with the plane of the
pupil image detector 19. Therefore, an illumination profile defined
by aperture device 13 defines the angular distribution of light
incident on substrate W in spot S. To implement different
illumination profiles, an aperture device 13 can be provided in the
illumination path. The aperture device may comprise different
apertures 13a, 13b, 13c etc. mounted on a movable slide or wheel.
It may alternatively comprise a fixed or programmable spatial light
modulator (SLM). As a further alternative, optical fibers may be
disposed at different locations in the illumination pupil plane and
used selectively to deliver light or not deliver light at their
respective locations. These variants are all discussed and
exemplified in the documents cited above. The aperture device may
be of a reflective form, rather than transmissive. For example, a
reflective SLM might be used. Indeed, in an inspection apparatus
working in the UV or EUV waveband most or all of the optical
elements may be reflective.
[0075] Depending on the illumination mode, example rays 30a may be
provided so that the angle of incidence is as shown at `I` in FIG.
2(b). The path of the zero order ray reflected by target T is
labeled `0` (not to be confused with optical axis `O`). Similarly,
in the same illumination mode or in a second illumination mode,
rays 30b can be provided, in which case the angles of incidence and
reflection will be swapped compared with the first mode. In FIG.
2(a), the zero order rays of the first and second example
illumination modes are labeled 0a and 0b respectively.
[0076] As shown in more detail in FIG. 2(b), target grating T as an
example of a target structure is placed with substrate W normal to
the optical axis O of objective lens 16. In the case of an off-axis
illumination profile, a ray 30a of illumination I, impinging on
grating T from an angle off the axis O, gives rise to a zeroth
order ray (solid line 0) and two first order rays (dot-chain line
+1 and double dot-chain line -1). It should be remembered that with
an overfilled small target grating, these rays are just one of many
parallel rays covering the area of the substrate including
metrology target grating T and other features. Since the beam of
illuminating rays 30a has a finite width (necessary to admit a
useful quantity of light), the incident rays I will in fact occupy
a range of angles, and the diffracted rays 0 and +1/-1 will be
spread out somewhat. According to the point spread function of a
small target, the diffracted radiation of each order +1 and -1 will
be further spread over a range of angles, not a single ideal ray as
shown.
[0077] If the target has multiple periodic components, then each of
those will give rise to first and higher diffracted rays, which may
be in directions into or out of the page. The example of FIG. 2(b)
is merely describing a one-dimensional grating for simplicity.
[0078] In the branch of the collection path for dark-field imaging,
imaging optical system 20 forms an image T' of the target on the
substrate W on sensor 23 (e.g. a CCD or CMOS sensor). An aperture
stop 21 is provided in a plane in the imaging branch of the
collection path CP which is conjugate to a pupil plane of objective
lens 16. Aperture stop 21 may also be called a pupil stop. Aperture
stop 21 can take different forms, just as the illumination aperture
can take different forms. The aperture stop 21, in combination with
the effective aperture of lens 16, determines what portion of the
scattered radiation is used to produce the image on sensor 23.
Typically, aperture stop 21 functions to block the zeroth order
diffracted beam so that the image of the target formed on sensor 23
is formed only from the first order beam(s). In an example where
both first order beams were combined to form an image, this would
be the so-called dark field image, equivalent to dark-field
microscopy.
[0079] The images captured by sensor 23 are output to image
processor and controller PU, the function of which will depend on
the particular type of measurements being performed. For the
present purpose, measurements of asymmetry of the target structure
are performed. Asymmetry measurements can be combined with
knowledge of the target structures to obtain measurements of
performance parameters of lithographic process used to form them.
Performance parameters that can be measured in this way include for
example overlay, focus and dose. Special designs of targets are
provided to allow these measurements of different performance
parameters to be made through the same basic asymmetry measurement
method.
[0080] Processor and controller PU also generates control signals
such as .lamda. and AP, for controlling the illumination
characteristics (polarization, wavelength) and for selecting the
aperture using aperture device 13 or a programmable spatial light
modulator. Aperture stop 21 may also be controlled in the same way.
Each combination of these parameters of the illumination and the
detection is considered a "recipe" for the measurements to be
made.
[0081] Referring again to FIG. 2(b) and the illuminating rays 30a,
+1 order diffracted rays from the target grating will enter the
objective lens 16 and contribute to the image recorded at sensor
23. Rays 30b are incident at an angle opposite to rays 30a, and so
the -1 order diffracted rays enter the objective and contribute to
the image. Aperture stop 21 blocks the zeroth order radiation when
using off-axis illumination. As described in the prior
publications, illumination modes can be defined with off-axis
illumination in X and Y directions.
[0082] Apertures 13c, 13e and 13f in the aperture device 13 of FIG.
2(a) include off-axis illumination in both X and Y directions, and
are of particular interest for the present disclosure. Aperture 13c
creates what may be referred to as a segmented illumination
profile, and may for example be used in combination with a
segmented aperture defined for example by a segmented prism 22,
described below. Apertures 13e and 13f may for example be used in
combination with an on-axis aperture stop 21, in a manner described
in some the prior published patent applications, mentioned
above.
[0083] By comparing images of the target grating under these
different illumination modes, asymmetry measurements can be
obtained. Alternatively, asymmetry measurements could be obtained
by keeping the same illumination mode, but rotating the target.
While off-axis illumination is shown, on-axis illumination of the
targets may instead be used and a modified, off-axis aperture stop
21 could be used to pass substantially only one first order of
diffracted light to the sensor. In a further example, a segmented
prism 22 is used in combination with an on-axis illumination mode.
The segmented prism 22 can be regarded as a combination of
individual off-axis prisms, and can be implemented as a set of
prisms mounted together, if desired. These prisms define a
segmented aperture in which rays in each quadrant are deflected
slightly through an angle. This deflection in the pupil plane in
has the effect of spatially separating the +1 and -1 orders in each
direction in the image plane. In other words, the radiation of each
diffraction order and direction forms an image to different
locations on sensor 23 so that they can be detected and compared
without the need for two sequential image capture steps.
Effectively, separate images are formed at separated locations on
the image sensor 23. In FIG. 2(a) for example, an image T'(+1a),
made using +1 order diffraction from illuminating ray 30a, is
spatially separated from an image T'(-1b) made using -1 order
diffraction from illuminating ray 30b. This technique is disclosed
in the above-mentioned published patent application
US20110102753A1, the contents of which are hereby incorporated by
reference in its entirety. 2nd, 3rd and higher order beams (not
shown in FIG. 2) can be used in measurements, instead of, or in
addition to, the first order beams. As a further variation, the
off-axis illumination mode can be kept constant, while the target
itself is rotated 180 degrees beneath objective lens 16 to capture
images using the opposite diffraction orders.
[0084] Whichever of these techniques is used, the present
disclosure applies to methods in which radiation diffracted in two
directions, for example the orthogonal directions called X and Y,
is simultaneously captured.
[0085] While a conventional lens-based imaging system is
illustrated, the techniques disclosed herein can be applied equally
with plenoptic cameras, and also with so-called "lensless" or
"digital" imaging systems. There is therefore a large degree of
design choice, which parts of the processing system for the
diffracted radiation are implemented in the optical domain and
which are implemented in the electronic and software domains.
Image-Based Asymmetry Measurement
[0086] Referring to FIG. 3 (a), and viewing the pupil plane of the
illumination system P(IP) in the vicinity of aperture device 13,
aperture 13c has been selected to define a specific spatial profile
of illumination, illustrated at 902. In this desired spatial
profile of the illumination system, two diametrically opposite
quadrants, labeled a and b, are bright, while the other two
quadrants are dark (opaque). This spatial illumination profile,
when focused to form spot S on the target T, defines a
corresponding angular distribution of illumination, in which rays
from angles only in these two quadrants. This segmented type of
aperture is known in scatterometry apparatus, from the published
patent application US 2010/201963. The merits of this modified
illumination aperture will be described further below.
[0087] When rays from the bright segments of the illumination
profile 902 are diffracted by periodic features in a target
structure, they will be at angles corresponding to a shift in the
pupil plane. Arrows `x` in FIG. 3 (a) indicate the direction of
diffraction of illumination caused by structures periodic in the X
direction, while arrows `y` indicate the direction of diffraction
of illumination caused by structures periodic in the Y direction.
Arrows `0` indicate direct reflection, in other words zero order
diffraction. A feature of this segmented type of aperture is that,
with regard to lines of symmetry defined by expected directions of
diffraction (X and Y in this example), illuminated regions of the
illumination profile are symmetrically opposite dark regions.
Therefore there is the possibility to segregate the higher order
diffracted radiation, while collecting radiation directed in both
directions simultaneously.
[0088] FIG. 3 (b) illustrates a distribution of illumination in a
conjugate pupil plane P(CP) in the collection path of the
inspection apparatus. Assume firstly that the target T is a
one-dimensional diffraction grating, with a periodicity in the X
direction as a first direction. While the spatial profile 902 of
the illumination has bright quadrants labeled a and b, the
diffraction pattern resulting from diffraction by the lines of the
target grating is represented by the pattern at 904 in FIG. 3 (b).
In this pattern, in addition to zero order reflections labeled
a.sub.0 and b.sub.0 there are first order diffraction signals
visible, labeled a.sub.+x, b.sub.-x. Because other quadrants of the
illumination aperture are dark, and more generally because the
illumination pattern has 180.degree. rotational symmetry, the
diffraction orders a.sub.+x and b.sub.-x are "free", meaning that
they do not overlap with the zero order or higher order signals
from other parts of the illumination aperture (considering only the
X direction at this stage). This property of the segmented
illumination pattern can be exploited to obtain clear first order
signals from a diffraction grating (alignment mark) having a pitch
which is half the minimum pitch that could be imaged if a
conventional, circularly-symmetric illumination aperture were
used.
[0089] Now, assume that the target has periodic features in a
second direction, for example the Y direction which is orthogonal
to the first direction. These features in the second direction may
arise from segmentation in the nominally one-dimensional grating.
They may also arise from other one-dimensional gratings with Y
orientation, that may be present within the area of spot S and the
within the field of view of the inspection apparatus. They may also
arise from a mixture of these. Assume further that the features
periodic in the Y direction have the same period, and therefore the
same diffraction angle, as the features periodic in the X
direction. The result is diffraction signals a.sub.+y and b.sub.-y
that can be seen in the pupil 904 of the collection path. These
signals comprise first order diffraction signals in the Y
direction. For simplicity of illustration in the present drawings,
the diffraction signals in the Y direction and the X direction are
shown as free of one another. In practice, the X diffraction
signals and the Y diffraction may overlap in the pupil 904. The
reader skilled in the art will understand that this depends on the
pitches of the target in X and Y and the chosen wavelength.
[0090] Zero order signals a.sub.0 and b.sub.0 are also present in
the pupil of the collection system, as illustrated. Depending
whether these zero order signals are wanted or not, they may be
blocked by a segmented aperture stop 21, similar in form to
aperture 13d. For asymmetry-based measurements, it is generally the
higher order signals, for example the +1 and -1 order signals that
are of interest.
[0091] In the simple example illustrated, the Y direction
diffraction signals do not overlap the X direction diffraction
signals in the pupil of the collection path, but in other
situations they might overlap, depending on the pitch of the
grating and the wavelength of illumination. In any case, where
two-dimensional features of some kind are present, diffraction
signals from two directions can become mixed in the same quadrants
of the pupil in the collection path. In the case of segmented
gratings, the segmentation in one or both directions may be much
finer than the pitch of the grating in the other direction. Where
very fine segmentation is present, the higher order diffraction
signals may fall completely outside the aperture of the collection
path, but the present inventors have recognized that the
diffraction in the second direction may nevertheless cause a change
in the signals from the first direction, which do fall into the
quadrants at top left and bottom right in FIG. 3(b).
[0092] FIG. 3 (c) shows schematically the layout of the segmented
prism 22 in the imaging branch of the inspection apparatus of FIG.
2. The circular pupil P(CP) is represented by a dotted circle. In
each quadrant of the pupil, a differently angled prism is provided,
which deflects the radiation through a certain angle. This angular
deflection in the pupil plane translates into a spatial separation
of images in the plane of the detector 23, as illustrated already
above with reference to FIG. 2(a). The operation of the apparatus
in this type of configuration, and some practical benefits and
challenges, will now be described in further. The principles of the
present disclosure are applicable in other configurations,
however.
[0093] FIG. 4 depicts a composite metrology target formed on a
substrate W according to known practice. The composite target
comprises four target structures in the form of gratings 32 to 35
positioned closely together so that they will all be within the
measurement spot S formed by the illumination beam of the metrology
apparatus. A circle 31 indicates the extent of spot S on the
substrate W. The four target structures thus are all simultaneously
illuminated and simultaneously imaged on sensor 23. In an example
dedicated to overlay measurement, gratings 32 to 35 are themselves
overlay gratings formed by first features and second features that
are patterned in different lithographic steps. For ease of
description it will be assumed that the first features and second
features are formed in different layers of the semiconductor device
formed on substrate W, but they may alternatively be formed in one
layer, for example as part of a multiple patterning process.
Gratings 32 to 35 may be differently biased, meaning that they have
designed-in overlay offsets additional to any unknown overlay error
introduced by the patterning process. Knowledge of the biases
facilitates measurement of overlay between the layers in which the
different parts of the overlay gratings are formed. Gratings 32 to
35 may also differ in their orientation, as shown, so as to
diffract incoming radiation in X and Y directions.
[0094] In one known example, gratings 32 and 34 are X-direction
gratings with biases of +d, -d, respectively in the placement of
one grating relative to another. This means that grating 32 has its
overlying components arranged so that if they were both printed
exactly at their nominal locations one of the components would be
offset relative to the other by a distance d. Grating 34 has its
components arranged so that if perfectly printed there would be an
offset of d but in the opposite direction to the first grating and
so on. Gratings 33 and 35 are Y-direction gratings with offsets +d
and -d respectively. Separate images of these gratings can be
identified in the image captured by sensor 23. While four gratings
are illustrated, another embodiment might require a larger matrix
to obtain the desired accuracy.
[0095] FIG. 5 shows an example of an image that may be formed on
and detected by the sensor 23, using the target of FIG. 4 in the
apparatus of FIGS. 2-3, using the segmented illumination profile
and using the segmented prisms 22. Such a configuration provides
off-axis illumination in both X and Y orientations simultaneously,
and permits detection of diffraction orders in X and Y
simultaneously, from the quadrants at upper left and lower right of
the pupil 904 in FIG. 3(b).
[0096] The dark rectangle 40 represents the field of the image on
the sensor, within which the illuminated spot 31 on the substrate
is imaged into four corresponding circular areas, each using
radiation only from one quadrant of the pupil 904 in the collection
path CP. Four images of the target are labelled 502 to 508. Within
image 502 the image of the illuminated spot 31 using radiation of
the upper left quadrant of the pupil 904 is labelled 41. Within
this, rectangular areas 42-45 represent the images of the small
target gratings 32 to 35. If the gratings are located in product
areas, product features may also be visible in the periphery of
this image field. Image processor and controller PU processes these
images using pattern recognition to identify the separate images 42
to 45 of gratings 32 to 35. In this way, the images do not have to
be aligned very precisely at a specific location within the sensor
frame, which greatly improves throughput of the measuring apparatus
as a whole.
[0097] As mentioned and as illustrated in FIG. 5, because of the
action of the segmented prism 22 on the signals in the pupil 904 of
the collection path, and because of the segmented illumination
profile 902 and its orientation relative to the X and Y directions
of the target T, each of the four images 502-508 uses only certain
portions of the diffraction spectra of each target. Thus the images
504 and 508 at lower left and upper right respectively are formed
of the zero order radiation a.sub.0 and b.sub.0 respectively. The
image 502 is formed of higher order diffracted radiation,
specifically radiation diffracted in the negative X direction from
bright quadrant b and the positive Y direction from bright quadrant
a (diffraction signals a.sub.+y and b.sub.-y). Conversely, image
506 is formed of higher order diffracted radiation, specifically
radiation diffracted in the positive X direction from bright
quadrant b and the negative Y direction from bright quadrant a
(diffraction signals a.sub.-y and b.sub.+x).
[0098] From the target comprising only one-dimensional gratings,
there is no cross-talk between signals diffracted in the X
direction and signals diffracted in the Y direction. That is
because each component grating 31-35 diffracts radiation in only
one of the two directions, and the image of each grating is
spatially separated within the images 502-508 by the imaging action
of the optical system. Once the separate images of the gratings
have been identified, the intensities of those individual images
can be measured, e.g., by averaging or summing selected pixel
intensity values within the identified areas (ROIs). Intensities
and/or other properties of the images can be compared with one
another to obtain measurements of asymmetry for the four or more
gratings simultaneously. These results can be combined with
knowledge of the target structures and bias schemes, to measure
different parameters of the lithographic process. Overlay
performance is an important example of such a parameter, and is a
measure of the lateral alignment of two lithographic layers.
Overlay can be defined more specifically, for example, as the
lateral position difference between the center of the top of a
bottom grating and the center of the bottom of a corresponding
top-grating. To obtain measurements of other parameters of the
lithographic process, different target designs can be used. Again,
knowledge of the target designs and bias schemes can be combined
with asymmetry measurements to obtain measurements of the desired
performance parameter. Target designs are known, for example, for
obtaining measurements of dose or focus from asymmetry measurements
obtained in this way.
Problems with Two-Dimensional Targets
[0099] Referring now to FIG. 4 (b), as mentioned above, some
targets will scatter or diffract radiation in two directions within
the same part of the image. The target of FIG. 4 (b) has
two-dimensional structures in each of the four component gratings
432-435. The two dimensional structures may arise from segmentation
in a one-dimensional grating in one or more layers. The
two-dimensional structures may alternatively arise from gratings
representing arrays of contact holes or vias, for example, which
are fully 2-dimensional.
[0100] Although diffraction will therefore occur in both directions
X and Y, within each grating image 42-45, nevertheless the purpose
of the metrology target is to measure a parameter such as overlay
separately in one or both of the X and Y directions. The
contribution of diffraction from the other direction, in the same
part of the image, represents "contamination" or noise in the
wanted diffraction signals. In the overlay measurement we derive
X-overlay from the asymmetry (difference between +1st and -1st
order diffraction) in the X direction. Even at a simplistic level,
it can be appreciated that the added radiation from diffraction in
the Y direction leads to a worse signal to noise ratio. If the
segmentation is present in both layers (or has an asymmetric
shape), the added diffraction will not just add light, but also add
asymmetry. Moreover since overlay error may arise in both
directions, variations in asymmetry signals assumed to relate to
one direction may be sensitivity to overlay errors in the other
direction. This problem arises regardless whether the diffraction
signals in the second direction fall within the detection pupil
904. This will lead to measurement errors, on top of the signal to
noise degradation.
Overlay Targets for Two-Dimensional Overlay Measurement
[0101] FIG. 6 shows enlarged schematic views (a), (b) and (c) of a
metrology target 600 formed on a substrate and adapted for overlay
measurement in accordance with a first embodiment of the present
disclosure. The metrology target in this example comprises four
target structures 632, 633, 634, 635 which may have a size and
layout similar to the gratings 32-35 in the target of FIG. 4(a).
The view (a) is a plan view from above the substrate. The view (b)
is a cross-section along the line B in view (a) and the view (c) is
a cross-section along the line C. As can be seen, each target
structure 632-635 includes a set of first features 662 arranged
periodically in at least a first direction. The first direction is
the Y direction in this example, and each first feature 662
comprises a bar which is segmented in the second (X) direction. A
period Px of segmentation and a duty cycle of segmentation in the
second direction are different to a period Py and duty cycle in the
first direction, though they could be the same in another
example.
[0102] Each target structure 632-635 further includes a set of
second features 664 arranged periodically in at least the first
direction. The second features 664 in this example are also bars
segmented in the second direction, with the same period Py in the
first direction and the same period Px of segmentation in the
second direction. As shown in the cross-sectional views (b) and (c)
the first features in this example are formed in a first layer L1
of the target structure and the second features are formed in a
second layer L2. In other examples, formed for example by multiple
patterning processes, the first features and second features might
be formed in a single layer.
[0103] Overlay performance relates to the ability of a lithographic
manufacturing process to place second features precisely, relative
to the positions of existing first features. Suppose that the
target design is such that nominally each second feature is placed
exactly on top of a corresponding first feature. In the presence of
overlay error, the second features become displaced by an amount
OVx in the X direction and OVy in the Y direction, relative to
their corresponding first features. It is assumed that the overlay
error in both directions is constant over the small area of
metrology target 600, though it may vary between metrology targets
across a substrate and between substrates. The overlay error may
result from inaccurate placement of the second features themselves,
or it may result from distortion of the first features, caused for
example in the patterning step by which the first features were
formed, or in subsequent chemical and/or physical processing
steps.
[0104] As is known, target structures for overlay metrology can be
formed with programmed offsets (also known as "bias"), in addition
to the (unknown) overlay error. These bias values are programmed
into the target structures by appropriate design of the patterning
devices MA that are used to define the first features and second
features in the different layers L1 and L2 of the substrate. In the
known target of FIG. 4(a) each target structure has bias in only
one direction for measurement of overlay in that direction. It is
assumed that overlay error in the other direction does not
influence the measurement, but that turns out not to be the case.
The inventors have recognized that, even in cases where diffraction
orders in the second direction do not fall within the pupil of the
detection system, target structures that are two-dimensional in
both sets of features suffer from cross-talk between overlay in the
first direction and overlay in the second direction. The inventors
have further recognized that a bias scheme that includes
appropriate combinations of bias values in both the first direction
and the second direction can be used to obtain overlay measurements
in the first direction that are corrected for the effects of
periodic features and overlay variations in the second
direction.
[0105] In the example target 600, with respect to the first
direction (Y), positive bias values +dy are programmed into the
target structures 632 and 635, displacing the second features
upward in FIG. 6(a), while negative bias values -dy are programmed
into the target structures 633 and 634. With respect to the second
direction (X), positive bias values +dx are programmed into the
target structures 632 and 633, displacing the second features to
the right in FIG. 6(a), while negative bias values -dx are
programmed into the target structures 634 and 635. As illustrated
in the views (b) and (c), the actual placement of the second
features relative to the first features is a combination of the
programmed bias value in each direction and the unknown overlay
error in that direction.
[0106] Thus the metrology target illustrated in FIG. 6 includes
target structures with four different combinations of bias in the
two directions of periodicity. The target in this example clearly
has the Y direction as its primary direction, and segmentation in
the X direction will cause weaker diffraction. This target is
therefore designed primarily to measure overlay in the Y direction.
A similar metrology target can be provided, if desired, arranged so
that the primary direction of periodicity is the X direction,
allowing measurement of overlay more accurately in the X direction.
A target in which periodic effects are equally strong in both
directions could be used to measure overlay equally accurately in
both directions.
Mathematical Model
[0107] Now let us consider how the intensities of the individual
grating areas within the image are conventionally used to calculate
(one-dimensional) overlay error Ov from a pair of (one-dimensional)
biased gratings (FIG. 4(a)). In a simplified linear approximation,
the overlay OV is calculated by using the intensities from the
individual target structures (gratings 32-35):
Ov = d ( ( I + d + 1 - I + d - 1 ) + ( I - d + 1 - I - d - 1 ) ( I
+ d + 1 - I + d - 1 ) - ( I - d + 1 - I - d - 1 ) ) ( 1 )
##EQU00001##
where (I.sub.+d.sup.+1-I.sub.+d.sup.-1) represents the difference
of intensity between +1 and -1 order diffraction signals from a
target structure with bias value +d and
(I.sub.-d.sup.+1-I.sub.-d.sup.-1) represents the difference of
intensity between +1 and -1 order diffraction signals from a target
structure with bias -d.
[0108] Equation (1) can be re-written as:
Ov = d ( As + d + As - d A s + d - As - d ) ( 2 ) ##EQU00002##
where As.sub.+d is an asymmetry value derived from the diffraction
signals for the target structure with bias +d and As.sub.-d is an
asymmetry value derived from the diffraction signals for the target
structure with bias -d.
[0109] The above Equation (1) is derived from the assumption that
there is linear relationship between asymmetry As and overlay error
Ov:
As.sub..+-.d=I.sub..+-.d.sup.+1-I.sub..+-.d.sup.-1=K*Ov (3)
where K is a simple coefficient. In practice, an implementation may
use a different model of the relationship. For example a sinusoidal
model of the relationship is often used, in which case Equation (2)
becomes:
Ov = a tan ( As + d + As - d As + d - As - d tan ( d ) ) ( 2 ' )
##EQU00003##
[0110] In Equation (2'), the offset d is expressed as an angle,
relative to 2.pi. radians representing the period of the grating.
For the purposes of the present description, the simple, linear
model will be assumed. The skilled person can readily implement the
same principles using a sinusoidal model or other preferred model,
adapting the other Equation (3) as necessary.
[0111] In all the above equations, some scaling factors and
normalization factors are omitted for simplicity. For example, as
described in some of the prior published applications mentioned
above, it may be convenient to normalize the differences between
intensities using the average of those intensities as a
denominator. So, for example, where above is written:
As.sub..+-.d=I.sub..+-.d.sup.+1-I.sub..+-.d.sup.-1 (3)
the full expression might be:
As .+-. d = 2 ( I .+-. d + 1 - I .+-. d - 1 I .+-. d + 1 + I .+-. d
- 1 ) = K * Ov ( 3 ' ) ##EQU00004##
[0112] The shorter expression will be used for convenience in the
present disclosure, while the person skilled in the art can
incorporate normalization and other practical details with routine
skill and knowledge.
[0113] If each target structure were only one-dimensional, as in
FIG. 4(a), then a single captured image 40 as shown in FIG. 5 has
the complete information required to obtain independent
measurements of overlay Ov with respect to the X and Y directions.
In the case where a grating in the target has two-dimensional
structure, however, the diffraction signals for different
directions become mixed and inter-dependent as described above.
[0114] In terms of the mathematical model introduced above, the
presence of additional orthogonal diffraction orders adds
additional unknowns to the set of equations that must be solved to
calculate Ov. Maintaining for now the simplicity of a linear model,
the dependence of asymmetry value As on overly error incorporates
an additional term, illustrated in Equation (4) below.
As=K.sub.x*Ov.sub.x+K.sub.y*Ov.sub.y+K.sub.xy*Ov.sub.x*Ov.sub.y
(4)
[0115] Here we see that the asymmetry observed in the diffraction
signals from a given target structure results from the effects of
the overlay Ov.sub.x in the x direction and overlay Ov.sub.y in y
direction, and additionally a "cross-term" dependent on overlay in
both directions. Coefficients K.sub.x and K.sub.y express the
sensitivity of asymmetry to overlay in each respective direction. A
third coefficient K.sub.yx represents sensitivity to the cross-term
(assuming for this explanation that the additional term depends
also linearly on the product Ov.sub.x*Ov.sub.y). While these
coefficients are represented in a mathematical model, their values
are not known in advance, similar to the coefficient K in the
one-dimensional example. The coefficient K is calculated
(implicitly at least) when the Equation (2) or (2') is applied to
calculate a measurement of overlay.
[0116] Considering the example metrology target 600 of FIG. 6, the
number of target structures and combinations of programmed offsets
in both directions, gives rise to the following four equations,
representing asymmetry values for the four target structures 632,
633, 635 and 634 respectively:
As.sub.+dx+dy=K.sub.x*(Ov.sub.x+dx)+K.sub.y*(Ov.sub.y+dy)+K.sub.xy*(Ov.s-
ub.x+dx)*(Ov.sub.y+dy)
As.sub.+dx-dy=K.sub.x*(Ov.sub.x+dx)+K.sub.y*(Ov.sub.y-dy)+K.sub.xy*(Ov.s-
ub.x+dx)*(Ov.sub.y-dy)
As.sub.-dx+dy=K.sub.x*(Ov.sub.x-dx)+K.sub.y*(Ov.sub.y+dy)+K.sub.xy*(Ov.s-
ub.x-dx)*(Ov.sub.y+dy)
As.sub.-dx-dy=K.sub.x*(Ov.sub.x-dx)+K.sub.y*(Ov.sub.y-dy)+K.sub.xy*(Ov.s-
ub.x-dx)*(Ov.sub.y-dy) (5)
[0117] The asymmetry values themselves are obtainable from the
diffraction signals, extracted for example from an image of the
type shown in FIG. 5. However, this set of four equations has five
unknowns: Kx, Ky, Kxy, Ovx and Ovy, and so is not solvable with
standard techniques.
[0118] The inventors have recognized that, by obtaining a second
set of diffraction signals under different conditions, an
additional set of equations can be added. This second set of
diffraction signals results in the additional set of equations:
As.sub.+dx+dy'=K.sub.x'*(Ov.sub.x+dx)+K.sub.y'*(Ov.sub.y+dy)+K.sub.xy'*(-
Ov.sub.x+dx)*(Ov.sub.y+dy)
As.sub.+dx-dy'=K.sub.x'*(Ov.sub.x+dx)+K.sub.y'*(Ov.sub.y-dy)+K.sub.xy'*(-
Ov.sub.x+dx)*(Ov.sub.y-dy)
As.sub.-dx+dy'=K.sub.x'*(Ov.sub.x-dx)+K.sub.y'*(Ov.sub.y+dy)+K.sub.xy'*(-
Ov.sub.x-dx)*(Ov.sub.y+dy)
As.sub.-dx-dy'=K.sub.x'*(Ov.sub.x-dx)+K.sub.y'*(Ov.sub.y-dy)+K.sub.xy'*(-
Ov.sub.x-dx)*(Ov.sub.y-dy) (6)
where the prime symbol ' indicates (observed) asymmetry values and
(unknown) coefficients that are applicable to the second
diffraction signals. The overlay values in each direction are the
same for both sets of diffraction signals. Therefore the second
diffraction signals bring four additional equations but only three
additional unknowns. When combined with the previous set of
equations (which had 5 unknowns), they can be solved for Ovx and
Ovy.
[0119] FIG. 7 shows two images 740(1) and 740(2) obtained by two
image capture steps in a method according to a first embodiment of
the present disclosure. Each image captures diffraction signals
from the target illustrated in FIG. 6, but using different capture
conditions. Each image 740(1) and 740(2) is of the same form as
that shown in FIG. 5, with four spatially separated images
702(1/2)-708(1/2) of the target. As described already for FIG. 5,
each image 702(1/2) is formed of radiation diffracted by the target
in the negative X direction and the positive Y direction (labelled
-x/+y). Each image 706(1/2) is formed of radiation diffracted in
the positive X direction and the negative Y direction (+x/-y). A
spot indicates the region representing diffraction signals of the
individual target structure 632 in each case. The difference
between them is that images 702(1) and 706(1) are a record of first
diffraction signals captured under first capture conditions while
images 702(2) and 706(2) are a record of second diffraction signals
captured under second illumination conditions different from the
first illumination conditions.
[0120] The first and second capture conditions can differ in one or
more parameters chosen from a wide variety of operating parameters
of the inspection apparatus and its operation. For example the
difference may be in illumination conditions used for the capture
of diffraction signals, such that first illumination conditions and
second illumination conditions differ in one or more of radiation
wavelength, radiation polarization, and angular distribution of
illumination. The difference may be not in the illumination
conditions, or not only in the illumination conditions, but also
there may be difference in conditions on the detection side. For
example a wavelength filter, a difference in aperture and/or a
difference in polarization can all be applied at the detection
side, by suitable filters, for example. References to differences
in capture conditions should therefore be understood to include any
differences in the conditions, ranging from the source itself,
through the illumination path and the collection path, and through
to the detector and processing of signals.
[0121] Due to the different capture conditions used in capturing
images 740(1) and 740(2), asymmetry values calculated from their
diffraction signals will have different sensitivities to overlay in
the different directions. First asymmetry values calculated from
the first diffraction signals represented in image 740(1) can be
used as asymmetry values As input to the equations (5) above, while
second asymmetry values As', for the same target structures, can be
calculated from the second diffraction signals in image 740(2) and
used as input to the equations (6) above. With a total of 8
equations, the 8 unknowns can be calculated. These unknowns include
the overlay errors Ov.sub.x and Ov.sub.y in the two directions, so
that the desired overlay measurement can be obtained.
[0122] The obtained overlay measurement, for example OVy in case of
the target shown in FIG. 6, will be subject to reduced sensitivity
to variation of overlay in the second direction, even though the
target structures have strongly two-dimensional features. Note also
that the calculation, and the mathematical model underlying it,
makes no assumption that a particular target structure or
diffraction signal or asymmetry value is representing asymmetry and
overlay in a particular direction. The calculation is therefore
valid even when the effects of overlay in both directions are
completely mixed in the captured diffraction signals. With the
appropriate choice of bias scheme and solution of the sufficient
number of simultaneous equations, the overlay error specific to
each direction can be calculated to obtain a desired measurement.
The design of target structures can of course be optimized so that
a particular target gives a more reliable (accurate) measurement of
overlay in one direction than the other. The primary periodicity
will typically be the first direction in the language of the
introduction and claims, and could be the X direction, the Y
direction, or any arbitrary direction.
Application Example
[0123] FIG. 8 illustrates a method of measuring performance of a
lithographic process using the apparatus and methods outlined
above. In step S20, one or more substrates are processed to produce
a metrology target including a plurality of target structures. The
design of target can be for example the design shown in FIG. 6 and
described above. Other designs are of course possible, including
examples described below. Targets may be large target or small
target designs, depending whether the first measurement branch or
second measurement branch of the apparatus is to be used. Targets
may comprise a plurality of target structures in distinct areas.
For the purposes of the present description, it is assumed that
overlay is of interest as a performance parameter of the
lithographic manufacturing process.
[0124] At step S20 a substrate is loaded into an inspection
apparatus, such as the inspection apparatus of FIG. 2. The
substrate is one on which target structures (and optionally also
functional device structures) have been produced using the
lithographic manufacturing system of FIG. 1. For this purpose, a
set of patterning devices will be provided, to define features of
device structures and metrology targets through a series of
patterning operations, interleaved with chemical and physical
processing steps. One of these patterning devices will define,
directly or indirectly, the first features of a plurality of target
structures implementing the principles of the present disclosure.
Another patterning device will define, directly or indirectly, the
second features. The positions of the first and second features in
the patterning devices include the programmed offsets for a
two-dimensional bias scheme. If the lithographic tool used for some
or all of the patterning steps uses a programmable patterning
device, then the set of patterning devices may include one or more
sets of patterning data, rather than physical reticles.
[0125] In step S21 metrology recipes are defined, including a
recipe for a measurement of overlay using two or more sets of
diffraction data, such as the ones captured in the images described
above with reference to FIG. 7. All the usual parameters of such a
recipe are defined, including the wavelength polarization, angular
distribution and so forth of illuminating radiation.
[0126] In accordance with the principles of the present disclosure,
the recipe defines two (or more) different sets of parameters, from
which the first and second diffraction signals are obtained. In a
first example, the difference between the first and second
diffraction signals is the wavelength of the illuminating
radiation. In other embodiments, different polarizations may be
defined, or different angular distributions of illuminating
radiation (illumination profiles) may be defined. As mentioned
above, one also use different detection parameters, e.g. aperture
or wavelength or polarization filtering in the detection path. In
other embodiments, described below with reference to FIGS. 10 and
11, the recipes may specify different subsets of the target
structures to be used for obtaining the first and second
diffraction signals, under a single set of capture conditions.
[0127] In step S22, the inspection apparatus is operated to capture
two or more sets of diffraction signals from the plurality of
target structures. These may for be dark-field images (such as
images 740(1) and 740(2) in FIG. 7) using the specified capture
conditions/subsets.
[0128] As illustrated by the dotted box, third and further sets of
diffraction signals can be obtained using yet further different
capture conditions and/or target subsets. The mathematical
discussion above have shown that by switching to a different
capture condition, the variables Ov.sub.x and Ov.sub.y remain
constant while other new unknowns are introduced. This means that,
with every additional change of capture conditions, and thus the
introduction of a new set of equations, the number of unknowns gets
closer to the number of equations. In a case where, when using the
first and second diffraction signals together, there remain more
unknowns than equations the process can be repeated with third
diffraction signals, fourth diffraction signals up to any number.
With enough changes, the number of equations becomes equal to the
number of unknowns and thus the equations can be solved. Thus this
method can be applied to any mathematical model that has any number
of unknowns: enough equations can be generated as long as the
number of available wavelengths permits additional items.
[0129] A different model may imply a greater number of
coefficients, requiring additional diffraction signals to solve a
system having more than eight unknowns. Even in the case of a
linear model featuring coefficients K.sub.x, K.sub.y and K.sub.xy,
which is solved by the two sets of four asymmetry values,
additional accuracy in the measured overlay values can result from
capturing additional diffraction signals and solving a larger set
of equations for the parameters of interest. For example, using
three or four target structures to obtain a third set of
diffraction signals one can construct a system of equations in 11
unknowns: the 8 mentioned above plus three new K values. Provided
three, four or more new asymmetry values are obtained, with only
three new additional coefficients. Adding another capture with
third diffraction signals will therefore bring additional accuracy
to the measurements of the parameters of interest, such as
overlay.
[0130] It goes without saying, any of these captures may in
practice be performed multiple times, with the result being
averaged to reduce random noise. It will also be understood that
the captures may be repeated for multiple targets across the
substrate.
[0131] At step S23 asymmetry values As and As' are calculated from
the captured diffraction signals of the various target structures.
In the example using dark-field imaging and segmented illumination
and detection optics, these asymmetry values can be derived by
selecting and combining pixel intensities from different regions of
interest within one or more dark-field images. First asymmetry
values As can be calculated from the first diffraction signals
captured in image 740(1), while second asymmetry values As' can be
can derived from the second diffraction signals captured in image
740(2).
[0132] At step S24, once sufficient asymmetry values have been
obtained for the number of unknowns, the full set of equations can
be solved to calculate one or more parameters of interest relating
to the target structures and/or relating to the performance of the
lithographic process by which the target structures have been
formed. Parameters of interest include in particular the
directional overlay values Ov.sub.x and Ov.sub.y. Parameters of
interest may be simply whether the image of a target structure
contains a mixture of radiation diffracted in two directions or
not. The value of cross-coefficient K.sub.xy relative to K.sub.x
and/or K.sub.y can be used, for example, as an indicator of
significant two-dimensional character.
[0133] Note that any resulting set of equations in any of the
aforementioned methods can be solved by using numerical techniques,
and does not require an analytical solution. Solution for all the
variables may be left as merely an implicit step, while only the
parameters of interest (e.g. overlay Ov in one or both directions)
are calculated and output explicitly.
[0134] At step S25, the metrology recipe may be updated in response
to the obtained measurements and ancillary data. For example, the
metrology techniques for a new product or target layout may be
under development. Information about the two-dimensional
characteristics can be used to select a more appropriate
recipe.
[0135] In step S26, in a development and/or production phase of
operating the lithographic production facility of FIG. 1, recipes
for the lithographic process may be updated, for example to improve
overlay in future substrates. The ability to measure overlay more
accurately in one or both different directions allows more
effective corrections to be developed and applied. The techniques
disclosed herein are fully compatible with the efficient
measurement techniques using segmented illumination and segmented
detection systems, including when target structures have
significant two-dimensional structure. An inspection apparatus can
be used with a fixed, segmented detection system, while covering a
full range of targets, reducing cost and size of the apparatus.
[0136] The calculations to obtain measurements, and to control the
selection of wavelengths and other recipe parameters, can be
performed within the image processor and controller PU of the
inspection apparatus. In alternative embodiments, the calculations
of asymmetry and other parameters of interest can be performed
remotely from the inspection apparatus hardware and controller PU.
They may be performed for example in a processor within supervisory
control system SCS, or in any computer apparatus that is arranged
to receive the measurement data from the processor and controller
PU of the inspection apparatus. Control and processing of the
calibration measurements can be performed in a processor separate
from that which performs high-volume calculations using the
correction values obtained. All of these options are a matter of
choice for the implementer, and do not alter the principles applied
or the benefits obtained. Use of the term "processor" in the
description and claims should be understood also to encompass a
system of processors.
Additional Example of First Embodiment
[0137] FIG. 9 illustrates (a) a target 900 similar to the target
600 of FIG. 6 and (b) implementation of part of the method of FIG.
8 using such a target. The target 900 comprises four target
structures 932, 933, 934, 935. Each target structure in this
example has the X direction as its primary direction of
periodicity, and is intended for accurate measurement of overlay
Ov.sub.x in the X direction. Segmentation in the Y direction may be
present but is not visible on the scale of the drawing. Programmed
offsets in X and Y directions are included in the relative
placement of the first and second features in each target
structure. As shown by labels "+dx+dy" etc., these offsets
implement a bias scheme with four combinations of offset the same
as the one shown in FIG. 6.
[0138] As shown in FIG. 9 (b), in step S22 two sets of diffraction
signals are captured from the four target structures. When using
the inspection apparatus of FIG. 2, all four target structures
933-935 are illuminated simultaneously within the illumination spot
931. The inspection apparatus captures first and second diffraction
signals in two dark-field images 740(.lamda.1) and 740(.lamda.2).
The dark-field images in this example are examples of the images
740(1) an 740(2) shown in FIG. 7. The index labels .lamda.1 and
.lamda.2 indicate that the wavelength of radiation used to capture
the first and second diffraction signals is different. Predictably,
the result will be a difference in the angle of spread of the
diffraction orders from the gratings formed by the first and second
features. Importantly, however, the interaction of the radiation
with the stack of layers defining the target structures may be
different in a number of ways, not necessarily predictable or
known. Particular differences in interaction can result from the
three-dimensional nature of the target structure, in which the
thickness and material properties of the layers L1 and L2 and
intervening layers all influence propagation of the inspection
radiation. Each set of diffraction signals will be sensitive in
different ways to overlay and to process variations in the
different parameters. In terms of the mathematical model presented
above, the coefficients K' will be different when using the second
wavelength than coefficients K when using the first wavelength.
[0139] In step S23 first and second asymmetry values As and As' are
derived for each target structure. These are combined in step S24
to obtain a measurement of overlay in at least the first direction,
being the X direction in this example.
[0140] Depending on the construction of the inspection apparatus,
the first and second diffraction signals can be captured
sequentially or simultaneously. Selection of wavelengths can be
through color filter 12b, or by a tunable or switchable source 11.
Illumination with multiple wavelengths could be used, with
filtering at the detection side. The choice of wavelengths can be
made based on calculation and/or experiment with the designs of
target structure, with the aim of ensuring a significant difference
between the first coefficients K and the second coefficients K',
thereby to maximize the information content of the asymmetry values
when combined together. Other examples of the first embodiment can
be made by switching other parameters such as the polarization
(filter 12c) or angular distribution (aperture device 13) of the
illumination system. As mentioned, parameters can also be switched
in the detection system, in addition to or as an alternative to the
illumination system.
Second Embodiment
[0141] FIG. 10 illustrates (a) a different form of target and (b)
implementation of part of the method of FIG. 8 using such a target
in a second embodiment. The target 1000 comprises eight, rather
than four, target structures. The eight target structures are
divided into two distinct subsets, indicated by suffixes `a` and
`b`. As in FIG. 9, each target structure in this example has the X
direction as its primary direction of periodicity, and is intended
for accurate measurement of overlay Ov.sub.x in the X direction.
Segmentation in the Y direction may be present but is not visible
on the scale of the drawing. A first subset of target structures
comprises four target structures 1032a, 1033a, 1034a and 1035a. A
second subset of target structures comprises four target structures
1032b, 1033b, 1034b and 1035b. Within each subset, programmed
offsets in X and Y directions are included in the relative
placement of the first and second features in each target
structure. As shown by labels "+dx+dy" etc., these offsets
implement in each subset a bias scheme with four combinations of
offset the same as the one shown in FIG. 6.
[0142] As shown in FIG. 10 (b) in conjunction with FIG. 11, in step
S22 two sets of diffraction signals 740(a) and 740(b) are captured.
The capture conditions in this embodiment are assumed to be the
same for both the first diffraction signals and the second
diffraction signals. This may reduce measurement time. The
difference between the first and second diffraction signals is
achieved by a difference in design between the first and second
subsets of target structures. It will be seen that the eight target
structures are made smaller in the second direction, so that they
can all fit within the same illumination spot 1031 and field of
view of the apparatus. In this way, both sets of diffraction
signals can be captured from regions within a single dark-field
image 740(a/b). If preferred, the target structures could be kept
at the same size as the targets 600 and 900, but additional capture
steps would then be required to obtain a full set of diffraction
signals, and additional errors could be introduced through
inconsistency of the capture conditions.
[0143] The inspection apparatus in this example captures first and
second diffraction signals 740(a) and 740(b) and the index labels a
and b indicate that the target structure design used to capture the
first and second diffraction signals is different. Any kind of
difference that can be reliably produced in the lithographic
process may be considered. The target structures may have different
pitches and/or duty cycles in one or both of the directions. As
another simple difference, one subset of target structures may have
a "line on line" layout while the other subset has a "line on
trench" layout. In a line on line layout, the second features lie
directly on top of corresponding first features, as shown in the
FIG. 6 (b) cross-section. In the line on trench layout, the second
features lie over the spaces between the first features. In any
case, the interaction of the radiation with the stack of layers
defining the target structures may be different in a number of ways
between the two subsets of target structures, in ways which are not
necessarily predictable or known. Each set of diffraction signals
will be sensitive in different ways to overlay and to process
variations in the different parameters. In terms of the
mathematical model presented above, the coefficients K' will be
different for the second subset than coefficients K for the first
subset.
[0144] In step S23 a first asymmetry value As is derived for each
target structure 1032a-1035a within the first subset, and a second
asymmetry value As' are derived for each target structure within
the second subset. These four values As and four values As' are
combined in step S24 to obtain a measurement of overlay in at least
the first direction, being the X direction in this example.
[0145] It will be understood that third, fourth and further subsets
with further different designs can be included, if third, fourth
etc. sets of diffraction signals are required. Additionally, the
techniques of the first and second embodiments can be combined so
that, for example, two different capture conditions are used to
obtain diffraction signals from two different subsets of target
structures. Immediately this yields four set of diffraction
signals. By proper design of the different subsets and by proper
and choice of capture conditions, additional unknowns can be
solved. Alternatively, rather than solving a single large set of
equations, independent calculations of the overlay error can be
made using (say) first and second diffraction signals together and
third and fourth diffraction signals together. In this way, without
complicating the mathematical model and its solution, the same Ov
values can be measured multiple ways, and combined to increase the
overlay accuracy performance of the inspection apparatus.
[0146] Finally, as mentioned above, while the above techniques can
be used to measure a property of the target independently in two
directions, it may also be used as a simple check to see whether
significant two-dimensional structure in both sets of features is
present or not. If not, then a single set of diffraction signals
may be sufficient for measurement of further targets, saving time.
If significant two-dimensionality is present, indicated for example
by a significant value K, in one of both sets of diffraction
signals, then the techniques of the present disclosure can be
applied to obtain accurate measurements of overlay in one or both
directions.
Modified Embodiment
[0147] FIG. 12 illustrates (a) a target 1200 similar in function to
the target 1200 of FIG. 6, but with modifications that will now be
described. The target 1200 comprises four target structures 1232,
1233, 1234, 1235. Each target structure in this example has the
both the X direction and the Y direction as primary directions of
periodicity. As a simple example, each first feature 1262 may
comprise a square structure on the substrate with X-Y dimensions
200 by 200 nm, and the pitch Px and Py may be 800 nm in both
directions. Each of the first features and each of the second
features therefore has the same dimension in both the X direction
and the Y direction. In alternative implementations, the dimensions
and/or the pitch can be made different in the two directions. It
will be understood that the features and their spacing are
represented schematically, and are not shown to scale, nor in their
true numbers. As in the previous examples, each first feature
and/or each second feature may be sub-segmented into smaller
features, in one or two directions, this sub-segmentation not being
visible on the scale of the drawing. In the following examples, the
first and second features will be illustrated and described as if
they are unitary features, purely for simplicity. Programmed
offsets in X and Y directions are included in the relative
placement of the first features 1262 and the second features 1264
in each target structure. As shown by labels "+dx+dy" etc., these
offsets implement a bias scheme with four combinations of offset.
The combinations are the same as the ones shown in FIG. 6, but
arranged in a different relationship to one another. The bias
scheme of FIG. 6 is equally suitable, as will be understand from a
consideration of the principles explained below. Also, this target
is designed to have a high degree of symmetry, to reduce
sensitivity to aberrations of the optical system, and to make it
compatible with a existing metrology methods and apparatuses.
[0148] A main difference between the FIG. 12 target and the ones
described above is that the target structures are formed in one
continuous array. FIG. 12 (b) shows one layer of the target 1200,
comprising the first features 1262 only. This may be for example
the bottom layer of the overly metrology target 1200. As can be
seen, the first features 1262 are formed in a continuous periodic
array, with no gaps between distinct target structures. FIG. 12 (c)
is an enlarged view of a central portion of the target 1200,
corresponding to the dashed circle in FIGS. 12(a) and (b). Here it
can be seen that the four target structures 1232, 1233, 1234, 1235
are simply four regions of this larger, continuous array, in which
the programmed positional offset (overlay bias) is different from
one region to the next. The programmed positional offset in each
region can be represented conveniently by bias vectors, which are
shown in each region of FIG. 12 (c).
[0149] In the illustrated example, it is assumed that the first
features define the bottom layer of the overlay metrology target,
and the second features are in a top layer, applied subsequently.
The programmed offsets are in the placement of the second features
1264 in each region, while the array of first features is entirely
regular. This is only one possible example, however, and the
programmed offsets may be in the bottom layer, in the top layer or
in both. One benefit of providing the features of the bottom layer
in a continuous array is to reduce the process effects impacting
this structure. Furthermore, any process variations that are to be
corrected can be modelled over the complete target. In certain
processes and designs, however, there may be benefits to putting
the programmed offsets in the first layer, and a regular array in
the top layer.
[0150] There is in principle a risk that the diffraction signals
will include cross-talk due to edge effects between the different
target structures. The absence of gaps between the target
structures, and the uniformity of the array in both directions are
factors that may help to reduce edge effects. Moreover, the
selection of which target structures to place next to one another
in the metrology target can also be done so as to reduce noise due
to edge effects. As seen by the vectors in FIG. 12 (c), for
example, the arrangement is such that each target structure has two
immediate neighbors. The bias scheme is designed so that the
programmed offset of each target structure is intermediate between
the programmed offsets of its immediate neighbors. In terms of the
bias vectors, the bias vector rotates only 90 degrees from each
target structure to the next. The target structures may be
considered to be arranged in a ring, which can be traversed
clockwise or counter-clockwise, with the bias vector stepping
always by at most 90 degrees. The arrangement avoids having a
common border between, for example, two target structures with bias
vectors pointing in opposite directions. While edge effects are
inevitable when radiation is diffracted by a finite structure,
careful design can minimize them. By minimizing edge effects, the
signal noise can be reduced, or other constraints can be relaxed.
For example, the overall size of the metrology target may be able
to be reduced, and/or the positional accuracy of the target and the
regions of interest (ROI) can be relaxed.
[0151] FIG. 13 illustrates just the central detail from a metrology
target 1300 that is a further modification of the metrology target
of FIG. 12. In this example, narrow transition zones 1302 are
provided between the neighboring target structures, in which the
bias vector is intermediate between the bias vectors of the target
structures to either side. Such a transition zone, although it is
"wasted space" from one point of view, can help to improve accuracy
by reducing edge effects. In principle, the variation of bias could
be continuous from each feature to the next, without deviating from
the principles of the present disclosure. In such a case, however,
the signals may become too sensitive to errors in placement of the
regions of interest ROI.
Third Embodiment
[0152] FIG. 14 (a) shows a metrology target 1400 according to a
third embodiment of the present disclosure. In detail, the target
has the same basic structure as the targets 1200 and 1300 described
above. That is to say, a plurality of target structures are formed
as neighboring regions of a larger, continuous array. In this
target, however, the number of regions is greater than four, and
comprises eight outer regions arranged in a ring around a central
region. Numbering from the top left and proceeding
counter-clockwise, the eight outer regions form respective target
structures 1432-1 to 1432-8. As in the previous example, a
programmed positional offset (overlay bias) is different from one
region to the next. The programmed positional offset in this
example is represented conveniently by a bias vector, which is
shown in each region of FIG. 14 (a). The central region optionally
provides forms a central target structure 1432-0. As described
below, the central target structure may be used for different
purposes. For the present, it may be assumed to have a zero bias,
represented by a simple dot.
[0153] The form of the first features and second features may be
assumed to be the same as in the examples of FIGS. 12 and 13
(feature size 200.times.200 nm and pitch Px=Py=800 nm). If the
overall size of each continuous array is the same as in FIG. 12,
then of course each target structure will be smaller.
Alternatively, the overall size of the target can be increased to
achieve a desired size of individual target structure. The overall
size in one example is 16.times.16 .mu.m.
[0154] In the terminology of the introduction and claims, the eight
target structures 1432-1 to 1432-8 together provide both the first
subset of target structures and the second subset of target
structures in one metrology target. All eight target structures can
be imaged in a single capture step, if desired, and the diffraction
signals from all the target structures can be processed as one
large set. The division into subsets in such an embodiment is to
some extent arbitrary, since the only difference between them is
the overlay bias applied. For the purposes of the present
description, however, it is convenient to consider that the
odd-numbered target structures (1432-1, -3, -5, -7) form a first
subset and the even-numbered target structures (1432-2, -4, -6, -8)
form a second subset. It will be seen that the bias scheme of the
first subset of target structures is the same as the bias scheme of
the target 1200 in FIG. 12.
[0155] The second subset of target structures (1432-2, -4, -6. -8),
are positioned with each one in between two target structures of
the first subset. The bias scheme of the second subset is such that
each target structure has a programmed offset in the X and Y
directions that is intermediate between its neighbors on either
side. As a result, any target structure bordering two neighboring
target structures has a programmed offset intermediate between the
programmed offsets of those two neighboring target structures
(disregarding the central region which is not one of the target
structures for this purpose). In terms of bias vectors, it will be
seen that the bias vector now rotates less than 90 degrees between
neighboring target structures. More particularly, in this example,
the bias vector representing said programmed offset rotates 45
degrees between neighboring target structures.
[0156] The progressive change of the programmed offset from
neighbor to neighbor, together with the use of a single continuous
array of features has the benefit of reducing edge effects, in the
same manner as described already about for the examples of FIGS. 12
and 13. As already explained, this mitigates the loss of allows
signal quality that would normally be incurred by fitting a greater
number of targets to be fitted into a smaller total area.
Effectively, the detailed structure of target 1400 is the same as
shown in FIG. 13, but with the transition zones 1310 wide enough to
form target structures in their own right. Narrow transition zones
(not shown) could be provided between the regions in metrology
target 1400, the same manner as in FIG. 13. The transition zones
would have bias angles rotated less than 45 degrees from the target
structures on either side.
[0157] Referring now to FIG. 14 (b), this shows part of a multiple
image 1440 of the target, captured by the apparatus of FIG. 4 with
spatial separation of diffraction orders. To save space, only the
+1 and -1 diffraction order images are shown, and labeled 1440(+1)
and 1440(-1) respectively. The signals are processed in essentially
the same manner as in the first and second embodiments, except that
sufficient diffraction signals for an overlay measurement are
obtained from a single capture of the composite metrology target
1400. This does not exclude the possibility of increasing accuracy
by obtaining additional diffraction signals using different
radiation characteristics. Nor does it exclude the possibility of
providing additional targets with different design parameters. It
just means that sufficient signals to solve the required system of
equations can be obtained from the signal diffraction image
1440.
[0158] FIG. 14(b) is overlaid with a schematic representation of
signal processing to obtain asymmetry values As from the eight
target structures. A first asymmetry value As.sub.1 is obtained by
comparing the opposite order diffraction signals from regions of
interest (ROI) corresponding to the top right target structure
1432-7. As another example, a fourth asymmetry value As.sub.4 is
obtained by comparing the opposite order diffraction signals from
regions of interest (ROI) corresponding to the bottom left target
structure 1432-3. Numbers in circles identify the correspondence
between each target structure and a respective asymmetry value
As.sub.i, given a total of eight asymmetry values from the single
diffraction image.
[0159] Given these eight asymmetry values As.sub.1 to As.sub.8
obtained as shown in FIG. 14 (b), overlay can be determined by
solving a system of equations similar to those described above. For
the present example, the equations implement a sinusoidal model of
the relationship between asymmetry value and overlay, rather than
the simple linear approximation. The set of equations for the
target 1400 is thus:
AS 1 = K x sin ( 2 .pi. P ( OV x + d x ) ) + K y sin ( 2 .pi. P (
OV y + d y ) ) + K xy sin ( 2 .pi. P ( OV x + d x ) ) sin ( 2 .pi.
P ( OV y + d y ) ) ##EQU00005## AS 2 = K x sin ( 2 .pi. P ( OV x +
d x ) ) + K y sin ( 2 .pi. P ( OV y - d y ) ) + K xy sin ( 2 .pi. P
( OV x + d x ) ) sin ( 2 .pi. P ( OV y + d y ) ) ##EQU00005.2## AS
3 = K x sin ( 2 .pi. P ( OV x - d x ) ) + K y sin ( 2 .pi. P ( OV y
+ d y ) ) + K xy sin ( 2 .pi. P ( OV x + d x ) ) sin ( 2 .pi. P (
OV y + d y ) ) ##EQU00005.3## AS 4 = K x sin ( 2 .pi. P ( OV x - d
x ) ) + K y sin ( 2 .pi. P ( OV y - d y ) ) + K xy sin ( 2 .pi. P (
OV x - d x ) ) sin ( 2 .pi. P ( OV y - d y ) ) ##EQU00005.4## AS 5
= K x sin ( 2 .pi. P ( OV x + d x ) ) + K y sin ( 2 .pi. P ( OV y )
) + K xy sin ( 2 .pi. P ( OV x + d x ) ) sin ( 2 .pi. P ( OV y ) )
##EQU00005.5## AS 6 = K x sin ( 2 .pi. P ( OV x - d x ) ) + K y sin
( 2 .pi. P ( OV y ) ) + K xy sin ( 2 .pi. P ( OV x - d x ) ) sin (
2 .pi. P ( OV y ) ) ##EQU00005.6## AS 7 = K x sin ( 2 .pi. P ( OV x
) ) + K y sin ( 2 .pi. P ( OV y + d y ) ) + K xy sin ( 2 .pi. P (
OV x ) ) sin ( 2 .pi. P ( OV y + d y ) ) ##EQU00005.7## AS 8 = K x
sin ( 2 .pi. P ( OV x ) ) + K y sin ( 2 .pi. P ( OV y + d y ) ) + K
xy sin ( 2 .pi. P ( OV x ) ) sin ( 2 .pi. P ( OV y + d y ) )
##EQU00005.8##
[0160] If we assume that the ninth region provides a target
structure 1432-0 with zero bias, the equation for its asymmetry
value As.sub.0 would be a linear combination of the above eight
equations and does not add information. In matrix notation, the
matrix rank remains 8, with or without the `no-bias` case. This
raises the possibility to use the central region for other
purposes, as will be described below.
[0161] FIG. 15 illustrates implementation of part of the method of
FIG. 8 using such a target in accordance with the third embodiment
of the present disclosure. In step S23 asymmetry values As for both
first subset of the target structures and the second subset of the
target structures are obtained from the captured diffraction
signals. The illustration shows these being treated as a single set
of asymmetry values for further calculation. Rather than requiring
separate captures.
[0162] Because all the target structures have the same target
design parameters and the diffraction signals are all captured with
the same radiation characteristics, there is no difference in
coefficients K or period P, between the different target
structures. In other words, there are only five unknowns and, in
principle, only five of these equations should be sufficient to
solve all unknowns. However, it may be expected that increased
accuracy can be obtained by using six, seven or eight. If using
fewer than eight, it would be recommended to choose target
structures having a good balance of programmed offsets in both
directions. This is easier to achieve if one uses more than 5
values. In other embodiments, however, the number of target
structures over the first and second subsets may be fewer than
eight in any case.
[0163] Numerous variations of the metrology target 1400 can be
envisaged, without departing from the principles set forth above.
The regions defined by different programmed offsets do not all have
to be equal in size. Some could be larger, to give higher signal
quality, while others are smaller. For example, it could be decided
to provide a smaller area for target structures with mixed (X and
Y) biases, and larger areas to provide more emphasis on single
direction bias. (The central region can be smaller, too. The
regions do not need to be square. The examples shown have bias
vectors of different lengths, depending whether the target
structure has a mixed (X and Y) positional offset, or a single
direction offset (X only or Y only). As another modification, the
magnitude of the offset in each direction could be different
between different target structures for example to make the bias
vectors more equal in length. For example, to have a uniform length
of bias vector of 20 nm, the bias in target structure 1432-8 may be
for example (0, 20), while the bias vector in target structure
1432-7 would be (20h 2, 20h 2), or approximately (14, 14). As
another variation, the single bias target structures could be
arranged at the corners of the target, with the mixed bias regions
in between.
[0164] Additionally, as mentioned, the central regions which are
not used for overlay measurement target structures can be used for
additional purposes. As one example of this, instead of a zero-bias
overlay target, the second features (top-grating) can be omitted
entirely. This enables a measurement of the first layer by itself,
so that asymmetry value AS.sub.0 provides a measurement of
bottom-grating asymmetry as a useful performance parameter. In
alternative embodiments, the bottom-grating could be left-out, or
both gratings.
Modified Third Embodiment
[0165] FIG. 16 (a) illustrates an enlarged metrology target
according to a modified third embodiment of the present disclosure.
FIG. 16 (b) shows part of a multiple image of the target, captured
by the apparatus of FIG. 4. In detail, the target has the same
basic structure as the targets 1200, 1300 and 1400, described
above. That is to say, a plurality of target structures 1632-1 etc,
are formed as neighboring regions of a larger, continuous array. In
this target, however, the number of regions is 49, arranged in a
square array of 7.times.7 regions. The programmed positional offset
is again represented conveniently by a bias vector, which is shown
in each region. The bias scheme is again such that any target
structure bordering two neighboring target structures has a
programmed offset intermediate between the programmed offsets of
those two neighboring target structures (disregarding the regions
with zero bias, which are not one of the target structures for this
purpose). In terms of bias vectors, it will be seen that the bias
vector again rotates less than 90 degrees between neighboring
target structures. Again, in this example, the bias vector
representing said programmed offset rotates 45 degrees between
neighboring target structures. Also, this target is designed to
have 180-degree rotational symmetry, to reduce sensitivity to
aberrations of the optical system, and for compatibility with other
metrology methods.
[0166] The form of the first features and second features may be
assumed to be the same as in the examples of FIGS. 12 and 13
(feature size 200.times.200 nm and pitch Px=Py=800 nm). If the
overall size of each continuous array is the same as in FIG. 12,
then of course each target structure will be smaller.
Alternatively, the overall size of the target can be increased to
achieve a desired size of individual target structure. The overall
size in one example is 16.times.16 .mu.m.
[0167] In FIG. 16 (b), the corresponding portions of the
diffraction image are again numbered, to show which diffraction
signals yield which of the asymmetry values As.sub.1 to As.sub.8 in
the equations above. Each different programmed offset occurs at
least four times in different target structures across the array.
Four of the programmed offsets occur more times. This provides
redundancy of signals for solving the eight equations. This
redundancy can be exploited in a number of different ways, which
will now be described.
[0168] The 44 target structures (regions having non-zero programmed
offsets) can be viewed in different groups. These different groups
can be used as overlay targets for different layer pairs in a
device manufacturing process. The first features can be made in a
continuous array in a first layer, for example, and then the second
features in a can be added in a second layer over a first part of
the array to define a first group of target structures. Second
features can be added over a second part of the array to define a
second group of target structures, and so on. Provided each group
has a set of five or more different programmed offsets, for example
seven or eight different programmed offsets, the first group of
target structures can be used to measure overlay for the second
layer over the first layer, while the second group of target
structures can be used to measure overly for the third layer over
the first layer.
[0169] Referring now to FIG. 17, two different options are
illustrated for choosing such groups, given the basic design of
target 1600. Referring to FIG. 17 (a), it will be noticed that some
of the target structures define closed rings with their neighbors,
each one similar in layout to the target 1400. Four groups can be
defined as shown by the boundaries labeled 1700-1 to 1700-4. Each
of these groups contains the full set of eight programmed offsets.
FIG. 17 (b) shows another sub-division of the same basic target
design into groups. As shown by the boxes labeled 1702-1 to 1702-4,
the first, third, fifth and seventh rows of regions can be used to
provide four groups of target structures in which the target
structures of each group are arranged in a line. Each of the lines
contains seven of the set of eight different programmed offsets. As
mentioned above, anything from five to eight different programmed
offsets will be enough to solve the system of equations necessary
for an overlay measurement.
[0170] Within each group, any target structure bordering two
neighboring target structures has a programmed offset intermediate
between the programmed offsets of those two neighboring target
structures. Therefore, each group of target structures effectively
can be used as an independent metrology target and can be assigned
to a different top- or bottom-grating (multi-layer target). As
mentioned already, if it suits a particular process, it may be the
top grating that is formed as a continuous array without positional
offsets, while the programmed offsets are included in lower layers
formed before the top layer.
[0171] Instead of using different groups for different layers, the
redundancy in the target of FIG. 16 can be exploited to obtain
additional information relating to a single layer pair. For
example, the results from a single bias value, measured at
different locations on the target, can be used for correcting
process variations that may occur over the area of the target, or
variations in the inspection tool itself, such as in homogeneity of
the radiation spot. While groups are shown in FIG. 17 as squares or
lines of neighboring target structures, other schemes are possible
in which a group comprises target structures distributed widely,
even randomly across the larger target.
[0172] The numerous variations described above in relation to the
third embodiment can be applied equally in the modified third
embodiment.
[0173] The central no-bias regions can again be used to measure
bottom grating asymmetry or other parameters. Where the larger
target provides multiple such regions, the variation of these
parameters can be measured across the target. The central no-bias
regions can of course be used as additional input to overlay
calculations, to improve robustness against effects such as grating
asymmetry.
CONCLUSION
[0174] The principles disclosed above allow measurement accuracy to
be maintained when target structures have two-dimensional
characteristics in both first features and second features. The
technique is suitable for application in asymmetry measurements to
be made by dark field imaging methods, using segmented detection
systems, as well as other methods. Use of two or more sets of
capture conditions, and/or two or more different designs of target
structure allows the simple and efficient inspection apparatus
based on a segmented detection system to operate with a wider range
of target designs, including those having significant diffraction
in a second direction in both layers.
[0175] Additionally, the disclosed method and apparatus can deliver
information about the two-dimensional character of the target
structures. Such information may in practice be unknown, prior to
inspection.
[0176] With regard to the third embodiment, a single acquisition
from a composite metrology target can determine 2D overlay. This
third embodiment can be extendibility to multi-layer target
design.
[0177] The embodiments based on continuous array structures can
reduce noise and process dependency, as well as enabling
intra-target/inter-grating correction of process effects.
[0178] The arrangement of target structures with progressive bias
differences helps reduce edge-effects between gratings, enabling
larger ROI selection and/or reduced overall target size.
[0179] Additional measurements of parameters such as bottom grating
asymmetry can be integrated into the target design.
[0180] While specific embodiments of the invention have been
described above, it will be appreciated that the invention may be
practiced otherwise than as described.
[0181] While the inspection apparatus or tool illustrated in the
embodiments comprises a particular form of scatterometer having
first and second branches for simultaneous imaging of pupil plane
and substrate plane by parallel image sensors, alternative
arrangements are possible. Rather than provide two branches
permanently coupled to objective lens 16 with beam splitter 17, the
branches could be coupled selectively by a movable optical element
such as a mirror. The optical system could be made having a single
image sensor, the optical path to the sensor being reconfigured by
movable elements to serve as a pupil plane image sensor and then a
substrate plane image sensor.
[0182] While the optical system illustrated in FIG. 2 comprises
refractive elements, reflective optics can be used instead. For
example the use of reflective optics may enable the use of shorter
wavelengths of radiation.
[0183] While the target structures described above are metrology
targets specifically designed and formed for the purposes of
measurement, in other embodiments, properties may be measured on
targets which are functional parts of devices formed on the
substrate. Many devices have regular, grating-like structures. The
terms `target grating` and `target structure` as used herein do not
require that the structure has been provided specifically for the
measurement being performed.
[0184] In association with the inspection apparatus hardware and
suitable target structures realized on patterning devices and on
patterned substrates, an embodiment may include a computer program
containing one or more sequences of machine-readable instructions
implementing methods of measurement of the type illustrated above
to obtain information about a target structure and/or about a
lithographic process. This computer program may be executed, for
example, within image processor and controller PU in the apparatus
of FIG. 2 and/or the control unit LACU of FIG. 1. There may also be
provided a data storage medium (e.g., semiconductor memory,
magnetic or optical disk) having such a computer program stored
therein.
[0185] Further embodiments according to the invention are described
in below numbered clauses:
[0186] 1. A method of determining overlay performance of a
lithographic process, the method including the following steps:
[0187] (a) obtaining a plurality of target structures that have
been formed by the lithographic process, each target structure
comprising a set of first features arranged periodically in at
least a first direction and a set of second features arranged
periodically in at least the first direction and being subject to
overlay error in the placement of the second features relative to
the first features,
[0188] (b) using a detection system to capture first diffraction
signals comprising selected portions of radiation diffracted by at
least a subset of the target structures;
[0189] (c) using the detection system to capture second diffraction
signals comprising selected portions of radiation diffracted by at
least a subset of the overlay targets;
[0190] (d) processing asymmetry information derived from the first
diffraction signals and the second diffraction signals to calculate
at least a measurement of said overlay error in at least the first
direction,
[0191] wherein said target structures have been formed with
programmed offsets in the placement of the second features relative
to the first features in addition to said overlay error, the
programmed offsets within each subset differing in both the first
direction and in a second direction, the first and second
directions being non-parallel,
[0192] and wherein the calculation of overlay error in step (d)
combines said asymmetry information with knowledge of said
programmed offsets while making no assumption whether asymmetry in
a given target structure results from relative displacement of the
second features in the first direction, in the second direction or
both directions.
[0193] 2. A method according to clause 1 wherein the first
diffraction signals are captured in step (b) under first capture
conditions and the second diffraction signals are captured in step
(c) under second capture conditions different from the first
capture conditions.
[0194] 3. A method according to clause 2 wherein said first capture
conditions and said second capture conditions differ in one or more
of the wavelength, polarization, and angular distribution of
radiation used for illumination and/or detection of the target
structures.
[0195] 4. A method according to clause 1, 2 or 3 wherein the first
diffraction signals captured in step (b) comprise radiation
diffracted by a first subset of target structures and the second
diffraction signals captured in step (c) comprise radiation
diffracted by a second subset of target structures different from
the first subset of target structures.
[0196] 5. A method according to clause 4 wherein the target
structures of said first subset and the target structures of said
second subset differ in one or more of pitch, feature size,
relative placement, and segmentation in the second direction.
[0197] 6. A method according to clause 4 wherein the target
structures of said first subset and the target structures of said
second subset differ only in the combinations of programmed offsets
in both the first direction and the second direction, the number of
combinations of programmed offsets available over both subsets
being greater than four.
[0198] 7. A method according to clause 6 wherein target structures
of the first and second subsets are arranged together a composite
metrology target, the layout of target structures being such that a
bias vector defined by the programmed offsets in the first and
second directions varies progressively from each target structure
to its neighbors.
[0199] 8. A method according to any of clauses 1 to 7 wherein each
of said first features and said second features comprises a feature
whose dimension is the same in the first direction as in the second
direction.
[0200] 9. A method according to any preceding clause wherein each
of said first features and said second features comprises an
elongate feature extending transverse to the first direction and
being segmented periodically in the second direction.
[0201] 10. A method according to clause 9 wherein the segmentation
of the elongate first and second features has a period different to
a period of spacing of the first and second features.
[0202] 11. A method according to any preceding clause wherein the
first features of at least the first subset of target structures
are formed in a first continuous array and the second features of
the first subset of target structures are formed in a second
continuous array of features, the different target structures being
defined by variation of said positional offsets over one or other
of said continuous arrays.
[0203] 12. A method according to any preceding clause wherein the
calculation of overlay error in step (d) derives from the first
diffraction signals a first asymmetry value for each of at least
four target structures and derives from the second diffraction
signals a second asymmetry value for each of at least four target
structures, and uses at least the derived first and second
asymmetry values to solve equations in more than four unknowns, one
of said unknowns being the measurement of overlay error in the
first direction.
[0204] 13. A method according to any preceding clause wherein the
calculation of overlay error in step (d) derives from the first and
second diffraction signals an asymmetry value for five or more
target structures, and uses at least the derived asymmetry values
to solve equations in more than four unknowns, one of said unknowns
being the measurement of overlay error in the first direction.
[0205] 14. A method according to clause 13 wherein the calculation
of overlay error in step (d) derives from the first and second
diffraction signals an asymmetry value for seven or more target
structures, and uses at least the derived first and second
asymmetry values to solve equations in more than four unknowns, one
of said unknowns being the measurement of overlay error in the
first direction.
[0206] 15. A method according to any preceding clause further
comprising a step:
[0207] (c2) using the detection system to capture third diffraction
signals comprising selected portions of radiation diffracted by at
least a subset of the overlay targets,
[0208] and wherein the step (d) includes processing asymmetry
information derived from the first diffraction signals, the second
diffraction signals and the third diffraction signals to calculate
a measurement of said overlay error in at least the first
direction.
[0209] 16. A method according to clause 15 wherein the calculation
of overlay error in step (d) uses the first diffraction signals to
derive an asymmetry value for each of at least four target
structures, uses the second diffraction signals to derive an
asymmetry value for each of at least four target structures and
uses the third diffraction signals to derive an asymmetry value for
each of at least three target structures, and uses more than eight
of the derived asymmetry values to solve equations in more than
eight unknowns, one of said unknowns being the measurement of
overlay error in the first direction.
[0210] 17. A method according to any of clauses 12 to 16 to wherein
the calculation of overlay error in step (d) calculates a
measurement of overlay error in the second direction.
[0211] 18. A method according to any preceding