U.S. patent application number 11/361677 was filed with the patent office on 2006-12-28 for apparatus and method for enhanced critical dimension scatterometry.
This patent application is currently assigned to Accent Optical Technologies, Inc.. Invention is credited to Steve Hummel, Chris Raymond.
Application Number | 20060289790 11/361677 |
Document ID | / |
Family ID | 36440961 |
Filed Date | 2006-12-28 |
United States Patent
Application |
20060289790 |
Kind Code |
A1 |
Raymond; Chris ; et
al. |
December 28, 2006 |
Apparatus and method for enhanced critical dimension
scatterometry
Abstract
Scatterometers and methods of using scatterometry to determine
several parameters of periodic microstructures, pseudo-periodic
structures, and other very small structures having features sizes
as small as 100 nm or less. Several specific embodiments of the
present Invention are particularly useful in the semiconductor
industry to determine the width, depth, line edge roughness, wall
angle, film thickness, and many other parameters of the features
formed in microprocessors, memory devices, and other semiconductor
devices. The scatterometers and methods of the invention, however,
are not limited to semiconductor applications and can be applied
equally well in other applications.
Inventors: |
Raymond; Chris; (Bend,
OR) ; Hummel; Steve; (Bend, OR) |
Correspondence
Address: |
PERKINS COIE LLP;PATENT-SEA
P.O. BOX 1247
SEATTLE
WA
98111-1247
US
|
Assignee: |
Accent Optical Technologies,
Inc.
Bend
OR
|
Family ID: |
36440961 |
Appl. No.: |
11/361677 |
Filed: |
February 24, 2006 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60656712 |
Feb 25, 2005 |
|
|
|
Current U.S.
Class: |
250/459.1 |
Current CPC
Class: |
G01N 23/2251 20130101;
G01N 21/21 20130101; G03F 7/70625 20130101; G01N 2021/4792
20130101; G01N 21/956 20130101; G01B 11/30 20130101; G01N 21/47
20130101; G01B 11/0616 20130101; G01B 11/24 20130101; G01N 21/274
20130101; G01N 21/9501 20130101 |
Class at
Publication: |
250/459.1 |
International
Class: |
G01T 1/10 20060101
G01T001/10 |
Claims
1. A scatterometer for evaluating microstructures on workpieces,
comprising: a radiation source configured to produce a beam of
radiation having a wavelength; an optical system having a first
optics assembly and an object lens assembly, wherein the first
optics assembly is configured to condition the beam of radiation
such that beam is diffuse and randomized, and wherein the object
lens assembly is configured to (a) focus the beam to an area of at
object focal plane through at least a 15.degree. range of altitude
angles and at least a 90.degree. range of azimuth angles
simultaneously and (b) present a return radiation scattered from a
microstructure in a radiation distribution at a second focal plane;
a detector positioned to receive the radiation distribution of the
return radiation and configured to produce a representation of the
radiation distribution; and a computer operatively coupled to the
detector to receive the representation of the radiation
distribution, wherein the computer includes a database having a
plurality of simulated radiation distributions corresponding to
different sets of parameters of the microstructure and a
computer-operable medium containing instructions that cause the
computer to identify a simulated radiation distribution that
adequately fits the representation of the radiation distribution
produced by the detector.
2. The scatterometer of claim 1, further comprising a navigation
system including a light source separate from the radiation source
and navigation optics having a low magnification capability for
locating a general area of the microstructure and a high
magnification capability for accurately aligning the object lens
with the microstructure.
3. The scatterometer of claim 1 wherein the object lens assembly is
configured to focus the conditioned beam to a spot size not greater
than 30 .mu.m.
4. The scatterometer of claim 1 wherein the radiation source
comprises a laser configured to generate a first beam having a
first wavelength and a second beam having a second wavelength
different than the first wavelength.
5. The scatterometer of claim 4 wherein the first wavelength is
approximately 266 nm and the second wavelength is approximately 405
nm.
6. The scatterometer of claim 1 wherein the wavelength is
approximately 200 nm to approximately 475 nm.
7. The scatterometer of claim 1 wherein the wavelength is
approximately 375 nm to approximately 475 nm.
8. The scatterometer of claim 1 wherein the first wavelength is
approximately 244 nm and the second wavelength is approximately 457
nm.
9. The scatterometer of claim 1 wherein the wavelength is
approximately one of 405 nm or 457 nm.
10. The scatterometer of claim 1, wherein the computer-operable
medium further comprises instructions that cause the computer to
acquire data from only selected pixels of the detector having
sensitivities above a threshold level.
11. The scatterometer of claim 1, further comprising: a calibration
member having a first reflectivity of the wavelength located
proximate to a workpiece site; and a computer operatively coupled
to the detector, wherein the computer includes a computer-operable
medium containing instructions that determine a reference
reflectance using a first detected reflectance from the first
calibration member and a second detected reflectance from an area
separate from the calibration member.
12. The scatterometer of claim 11 wherein the calibration member
comprises a mirror having a reflectance greater than approximately
95% and capable of reflecting radiation through a range of altitude
angles of 0.degree. to 89.degree., and the second reflectance is
from free space.
13. (canceled)
14. The scatterometer of claim 1 wherein the first optics assembly
comprises a diffuser that produces a diffuse randomized beam.
15. The scatterometer of claim 1 wherein the first optics assembly
comprises an order selector configured to limit the angular range
of various diffraction orders.
16. The scatterometer of claim 1 wherein the scatterometer further
comprises a field stop having an aperture and an illumination lens
through which the diffused and randomized beam pass.
17. The scatterometer of claim 1 wherein the object lens assembly
comprises a plurality of achromatic lenses.
18. The scatterometer of claim 1, further comprising a plurality of
optical heads, wherein individual optical heads are suitable for a
unique bandwidth of radiation.
19. The scatterometer of claim 1, further comprising a reference
detector configured to measure changes in the beam from the
irradiation source.
20. The scatterometer of claim 1 wherein the object lens assembly
is configured to simultaneously focus the conditioned beam at the
object focal plane through a range of incidence angles having (a)
altitude angles of 0.degree. to at least about 45.degree. and (b)
azimuth angles of 0.degree. to at least about 90.degree..
21. The scatterometer of claim 20 wherein the altitude angles are
0.degree. to at least 70.degree. and the azimuth angles are
0.degree. to at least 180.degree..
22. The scatterometer of claim 20 wherein the altitude angles are
0.degree. to at least 80.degree. and the azimuth angles are
0.degree. to at least 360.degree..
23. The scatterometer of claim 1, further comprising a polarizing
beam splitter in a path of the return radiation between the object
lens assembly and the detector to separate the p- and s-polarized
components of the return radiation from each other.
24. The scatterometer of claim 23, further comprising a single
detector to receive both the p- and s-polarized components of the
return radiation, and wherein the polarizing beam splitter
comprises a cube-type polarizing beam splitter.
Description
CROSS REFERENCE OF RELATED APPLICATION
[0001] The present application claims the benefit of U.S.
Application No. 60/656,712, filed on Feb. 25, 2005, which is
incorporated by reference herein.
TECHNICAL FIELD
[0002] The present invention is related to evaluating
microstructures on workpieces, such as semiconductor wafers, using
apparatus and methods that can obtain a representation of the
distribution of radiation returning from the workpiece through a
large range of angles of incidence.
BACKGROUND
[0003] Semiconductor devices and other microelectronic devices are
typically manufactured on a workpiece having a large number of
individual dies (e.g., chips). Each wafer undergoes several
different procedures to construct the switches, capacitors,
conductive interconnects and other components of a device. For
example, a workpiece can be processed using lithography,
implanting, etching, deposition, planarization, annealing, and
other procedures that are repeated to construct a high density of
features. One aspect of manufacturing microelectronic devices is
evaluating the workpieces to ensure that the microstructures are
within the desired specifications.
[0004] Scatterometry is one technique for evaluating several
parameters of microstructures. With respect to semiconductor
devices, scatterometry is used to evaluate film thickness, line
spacing, trench depth, trench width, and other aspects of
microstructures. Many semiconductor wafers, for example, include
gratings in the scribe lanes between the individual dies to provide
a periodic structure that can be evaluated using existing
scatterometry equipment. One existing scatterometry process
includes illuminating such periodic structures on a workpiece and
obtaining a representation of the scattered radiation returning
from the periodic structure. The representation of return radiation
is then analyzed to estimate one or more parameters of the
microstructure. Several different scatterometers and methods have
been developed for evaluating different aspects of microstructures
and/or films on different types of substrates.
[0005] Eldim Corporation of France manufactures devices that
measure the photometric and calorimetric characteristics of
substrates used in flat panel displays and other products. The
Eldim devices use an Optical Fourier Transform (OFT) instrument
having an illumination source, a beam splitter aligned with the
illumination source, and a first lens between the beam splitter and
the sample. The first lens focuses the light from the beam splitter
to a spot size on the wafer throughout a large range of angles of
incidence (e.g., .PHI.=0.degree. to 360.degree. and
.THETA.=0.degree. to 88.degree.). The light reflects from the
sample, and the first lens also focuses the reflected light in
another plane. The system further includes an optical relay system
to receive the reflected light and a sensor array to image the
reflected light. International Publication No. WO 2005/026707 and
U.S. Pat. Nos. 6,804,001; 6,556,284; 5,880,845; and 5,703,686
disclose various generations of scatterometers. The scatterometers
set forth in these patents are useful for assessing the photometric
and calorimetric properties of flat panel displays, but they may
have several drawbacks for assessing parameters of extremely small
microstructures on microelectronic workpieces.
[0006] One challenge of scatterometry is properly locating very
small microstructures on a workpiece. This is not particularly
difficult for analyzing the pixels of a flat panel display because
measuring the photometric and colorimetric properties of such
substrates merely requires locating the illumination spot on
relatively large pixel areas instead of very small periodic
structures. As a result, systems used to analyze flat panel
displays may not include navigation systems capable of locating
very small microstructures on the order of 20-40 .mu.m. Moreover,
the devices used to analyze flat panel displays may have relatively
large spot sizes that are not useful to measure the properties of a
20-40 .mu.m grating because such large spot sizes generate
reflections from the surrounding areas that result in excessive
noise. Therefore, devices designed for assessing flat panel
displays may not be well-suited for assessing gratings or other
microstructures having much smaller dimensions on microelectronic
workpieces.
[0007] Another challenge of using scatterometry to evaluate very
small microstructures is obtaining a useful representation of the
radiation returning from such microstructures. Existing
scatterometers that assess the films and surface conditions of flat
panel displays typically use relatively long wavelengths of light
(e.g., 532 nm). In contrast to flat panel displays, many
microstructures on semiconductor wafers have line widths smaller
than 70 nm, and such microstructures are continually getting
smaller and being packed in higher densities. As a result, the
relatively long wavelengths used to assess flat panel displays may
not be capable of assessing very small microstructures on many
microelectronic devices. Therefore, devices used for assessing flat
panel displays may be further inadequate for assessing the
properties of microstructures on microelectronic workpieces.
[0008] Another challenge of assessing microstructures using
scatterometry is processing the data in the representation of the
return radiation. Many scatterometers calculate simulated or
modeled representations of the return radiation and then use an
optimization regression to optimize the fit between the simulated
representations and an actual reflectance signal. Such optimization
regressions require a significant amount of processing time using
high-power computers because the actual reflectance signals for
measurements through a large range of incidence angles contain a
significant amount of data that is affected by a large number of
variables. The computational time, for example, can require several
minutes such that the substrates are typically evaluated offline
instead of being evaluated in-situ within a process tool.
Therefore, many conventional scatterometers may not be well-suited
for evaluating microstructures on microelectronic workpieces.
[0009] Yet another challenge of assessing microstructures using
scatterometry is calibrating the scatterometer. One difficulty of
calibrating scatterometers is that the return radiation can have
both p- and s-polarized components when the input path is off-axis
relative to the microfeature (e.g., a grating). This increases the
complexity of fitting the output to a model because the p- and
s-polarized components must be treated separately. This is also
challenging because the p- and s-polarized components change for
each off-axis azimuth angle, and thus proper calibration requires
measurements and calculations for several different azimuth angles
in more sophisticated applications.
[0010] Calibrating scatterometers that operate over a large number
of azimuth angles is also difficult because it is challenging to
measure the p- and s-polarized components. One existing system for
measuring p- and s-polarized components is a two-camera system that
splits the output beam into separate p- and s-polarized beams which
propagate at a non-parallel angle relative to each other. Such
systems have one camera to detect the p-polarized component and
another camera to detect the s-polarized component. The use of two
cameras, however, is undesirable because the additional camera
increases the cost and form factor of the scatterometer. This may
prevent such two-camera scatterometers from fitting into many
integrated tool sets where metrology is desired. Additionally, it
is time-consuming to calibrate two cameras because of the
additional camera and compensating for the inherent variations in
the cameras. Such two-camera systems are also undesirable because
the separate images must be registered and integrated with each
other to produce a meaningful result. This is a significant,
time-consuming computational procedure. Another system for
measuring the p- and s-polarized components uses a single camera
and a polarizer that alternates between the p- and s-polarized
components. This system may have problems because the serial
presentation of the p- and s-polarized components to the detector
requires more time to obtain the measurements. Moreover, the
polarizer is a mechanical device that moves between p- and
s-polarizing states, and as such it may lack the precision and
accuracy to obtain meaningful measurements. Such mechanical devices
may wear out and further denigrate the precision and accuracy of
the calibration. Therefore, obtaining images of p- and s-polarized
components for calibrating scatterometers or other uses presents a
significant challenge in scatterometry.
[0011] Still another challenge of scatterometry is noise or
inconsequential data in the measurements. In systems that are able
to simultaneously obtain measurements through a large range of
altitude and azimuth angles, the data in many areas of the
resulting image may not be meaningful. Therefore, there is a need
to improve the process of operating scatterometers that
simultaneously obtain measurements for a large range of altitude
and azimuth angles.
BRIEF DESCRIPTION OF THE DRAWINGS
[0012] FIG. 1 is a schematic view illustrating a scatterometer in
accordance with an embodiment of the invention.
[0013] FIG. 2 is a schematic isometric view illustrating a portion
of a three-dimensional convergence beam for irradiating
microstructures on a workpiece in accordance with an embodiment of
the invention.
[0014] FIG. 3A is a schematic view illustrating an optical system
for use in a scatterometer in accordance with an embodiment of the
invention.
[0015] FIG. 3B is a schematic view of a cube-type polarizing beam
splitter for use in a scatterometer in accordance with an
embodiment of the invention.
[0016] FIG. 3C is a schematic view of a CMOS imager for use in a
scatterometer in accordance with an embodiment of the
invention.
[0017] FIG. 4 is a schematic view illustrating an optical system
and an auto-focus system for use in a scatterometer in accordance
with an embodiment of the invention.
[0018] FIG. 5A is a simulated radiation distribution for use in a
scatterometer in accordance with an embodiment of the
invention.
[0019] FIG. 5B is a measured radiation distribution provide by a
scatterometer in accordance with an embodiment of the
invention.
[0020] FIG. 6 is a schematic view illustrating a portion of a
computer system and a computational method for ascertaining
parameters of microstructures using a scatterometer in accordance
with an embodiment of the invention.
[0021] FIG. 7 is a flow chart illustrating a method for determining
one or more parameters of a microfeature using a scatterometer in
accordance with an embodiment of the invention.
[0022] FIG. 8 is a flow chart of a procedure for developing a
predetermined sensitivity record used in a method for assessing a
parameter of a microfeature on a workpiece in accordance with an
embodiment of the invention.
[0023] FIGS. 9A-9C are images illustrating additional aspects of
developing a predetermined sensitivity record, acquiring a measured
selected radiation distribution, and fitting the measured selected
radiation distribution to a modeled selected radiation
distribution.
DETAILED DESCRIPTION
A. Overview
[0024] The present invention is directed toward evaluating
microstructures on microelectronic workpieces and other types of
substrates. Many applications of the present invention are directed
toward scatterometers and methods of using scatterometry to
determine several parameters of periodic microstructures,
pseudo-periodic structures, and other very small structures having
features sizes as small as 100 nm or less. Several specific
embodiments of the present invention are particularly useful in the
semiconductor industry to determine the width, depth, line edge
roughness, wall angle, film thickness, and many other parameters of
the features formed in microprocessors, memory devices, and other
semiconductor devices. The scatterometers and methods of the
invention, however, are not limited to semiconductor applications
and can be applied equally well in other applications.
[0025] One embodiment of the invention is directed toward a
scatterometer for evaluating microstructures on workpieces. In this
embodiment, the scatterometer comprises an irradiation source, a
first optics assembly, and an object lens assembly. The irradiation
source can be a laser that produces a first beam of radiation at a
wavelength. The first optics assembly is aligned with the path of
the beam and configured to condition the beam (e.g., shape,
randomize, select order, diffuse, converge, diverge, collimate,
etc.), and the object lens assembly is positioned between the first
optics assembly and a workpiece site. The object lens assembly is
configured to focus the conditioned beam to a spot at an object
focal plane. The object lens assembly or another optical assembly
of the scatterometer is also configured to (a) receive radiation
scattered from a workpiece and (b) present a distribution of the
scattered radiation at a second focal plane. For example, the
radiation distribution can be the intensity, polarimetric,
ellipsometric and/or reflectance distribution of the scattered
radiation. The scatterometer of this embodiment can further include
a detector, a navigation system, and an auto-focus system. The
detector is positioned to receive at least a portion of the
radiation distribution and configured to produce a representation
of the radiation distribution. The navigation system is operatively
coupled to the lens assembly or a support structure holding the
workpiece, and it is configured to identify and locate the desired
microstructure on the workpiece. The auto-focus system is
operatively coupled to one of the lens assembly or the workpiece
site, and it is configured to position the microstructure at a
desired focal position (e.g., the object focal plane).
[0026] Another embodiment of a scatterometer in accordance with the
invention comprises a laser configured to produce a beam of
radiation having a first wavelength, an optical system having a
first optics assembly configured to condition the beam of
radiation, and a lens assembly. The lens assembly is configured to
focus the beam at an area of an object focal plane or other desired
focal plane having a small spot size such that the beam has angles
of incidence through a range of altitude angles of at least
approximately 0.degree. to 45.degree. and azimuth angles of at
least approximately 0.degree. to 90.degree.. The altitude angle
(.THETA.) is the angle between the illumination ray and a reference
vector normal to the object focal plane, and the azimuth angle
(.PHI.) is the angle between the incident plane and a reference
vector in a plane parallel to the focal plane. The beam more
preferably has angles of incidence through altitude angles of
0.degree. to greater than 70.degree. and azimuth angles of
0.degree. to 360.degree.. The scatterometer is further configured
to collect and present the radiation scattered from the
microstructure at a second focal plane. In one embodiment, the lens
assembly itself presents the scattered radiation at the second
focal plane, but in other embodiments the optical system has
another optic member that presents the radiation distribution at
the second focal plane. The scatterometer of this invention further
includes a detector positioned to receive the radiation
distribution and configured to produce a representation of the
radiation distribution. The scatterometer also includes a computer
operatively coupled to the detector to receive the representation
of the radiation distribution. The computer includes a database and
a computer-operable medium. The database has a plurality of
simulated radiation distributions corresponding to different sets
of parameters of the microstructure. The computer-operable medium
contains instructions that cause the computer to identify a
simulated radiation distribution that adequately fits the
representation of the measured radiation distribution.
[0027] Another embodiment of the invention is a scatterometer for
evaluating a microstructure on a workpiece comprising an
irradiation system, an optical system, and a detector. The
irradiation system includes a laser and/or a lamp, and the
irradiation system is configured to produce a first beam of
radiation having a first wavelength and a second beam of radiation
having a second wavelength. The optical system has a first unit
configured to condition the first and second beams, and a second
unit configured to (a) focus the first and second beams at an area
of an object focal plane having a small spot size, and (b) present
a distribution of scattered radiation returning from a
microstructure at a second focal plane. The detector is positioned
to receive the radiation distribution, and the detector is
configured to produce a representation of the radiation
distribution.
[0028] Another embodiment of a scatterometer in accordance with the
invention comprises a laser configured to produce a beam of
radiation having a wavelength, an optical system, a detector, a
calibration unit, and a computer. The optical system has a first
optics assembly configured to condition the beam of radiation such
that the beam is a diffuse and randomized beam. The optical system
also includes an object lens assembly configured to (a) focus the
beam at an area of an object focal plane and (b) present scattered
radiation returning from a microstructure in a radiation
distribution at a second focal plane. The detector is positioned to
receive the radiation distribution of the scattered radiation and
configured to produce a representation of the radiation
distribution. One embodiment of the calibration unit includes a
first calibration member having a first reflectivity of the
wavelength and a second calibration member having a second
reflectivity different than the first reflectivity. The first and
second calibration members are located to be irradiated by the beam
during a setup procedure to determine a reference reflectance or
other reference radiation distribution. In other embodiments, the
second calibration unit can be eliminated such that the second
reflectance is from free space. The computer is operatively coupled
to the detector and includes a computer-operable medium that
determines the reference reflectance using a first reflectance from
the first calibration member and a second reflectance from the
second calibration member or free space.
[0029] Still another embodiment of a scatterometer in accordance
with the invention comprises a laser and/or lamp configured to
produce a beam of radiation having a wavelength and an optical
system. The optical system has a first optics assembly including an
object lens assembly configured to focus the beam to an area at an
object focal plane and present return radiation scattered from a
microstructure in a radiation distribution at a second focal plane.
The optical system further includes a second optics assembly having
a polarizing beam splitter configured to present separate images of
p- and s-polarized components of the return radiation. The
scatterometer further includes a detector having a single array
positioned to simultaneously receive the separate images of the p-
and s-polarized components of the return radiation. The detector is
also configured to produce a representation of the p- and
s-polarized components of the return radiation.
[0030] Yet another embodiment of a scatterometer in accordance with
the invention comprises a laser and/or lamp configured to produce a
beam of radiation having a wavelength and an optical system having
a first optics assembly. The first optics assembly includes an
object lens assembly configured to focus the beam at an area on an
object focal plane and present return radiation scattered from a
microstructure in a radiation distribution at a second focal plane.
The scatterometer further includes a detector comprising a CMOS
imager have a die with an image sensor, focal optics, and packaging
that defines an enclosed compartment in which the focal optics and
the image sensor are fixed with respect to each other without a
cover having parallel, fiat surfaces between the image sensor and
the focal optics.
[0031] The present invention is also directed toward several
methods for evaluating a microstructure on a workpiece. One
embodiment of such a method comprises generating a laser beam or a
beam from a lamp having a wavelength and irradiating a
microstructure on a workpiece by passing the beam through a lens
assembly that focuses the beam to a focus area at a focal plane.
The focus area can have a dimension not greater than 50 .mu.m or in
other embodiments approximately at least 10 of the periodic
features of the microstructure, and the beam simultaneously has
altitude angles of 0.degree. to at least 15.degree. and azimuth
angles of 0.degree. to greater than 90.degree.. In several
applications, the focus area is not greater than 30 .mu.m, and the
altitude angles are 0.degree. to about at least 45.degree.. The
altitude angles can be from 0.degree. to at least 70.degree. in
other examples. The method further includes detecting an actual
radiation distribution corresponding to radiation scattered from
the microstructure.
[0032] In another embodiment of a method in accordance with the
invention the procedure of irradiating a microstructure comprises
irradiating the focus area with a laser beam having a first
wavelength and irradiating the focus area with a laser beam having
a second wavelength different than the first wavelength. The first
and second wavelengths can be in a range of approximately 200
nm-475 nm, and more specifically a first wavelength can be from 200
nm-300 nm and a second wavelength can be from 375 nm-475 nm. For
example, the first wavelength can be approximately 266 nm and the
second wavelength can be approximately 405 nm, or in another
embodiment the first wavelength can be about 244 nm and the second
wavelength about 457 nm. As such, the workpieces can be irradiated
with one or more beams having one or more wavelengths less than 500
nm, but longer wavelengths may be used in other embodiments. In
particular, a third wavelength of 633 nm may be used. Another
aspect in accordance with another embodiment of the invention
includes calibrating the detector by providing a first calibration
member having a first reflectivity and a second calibration member
having a second reflectivity. The system can be calibrated by
determining a reference reflectance using a first reflectance from
the first calibration member and a second reflectance from the
second calibration member. Other embodiments can use only a single
calibration member and obtain a second reflectance measurement from
free space.
[0033] In yet another embodiment of a method in accordance with the
invention, the magnitude and phase of the scattered radiation is
measured by the detection system to perform an ellipsometric style
measurement described above. The ellipsometric measurement in this
embodiment can be performed by positioning polarizers, wave plates,
or other phase modifying optical devices in the incident and/or
detection optical system assemblies.
[0034] In yet another embodiment of a method in accordance with the
invention, an automated workpiece transport system is incorporated
with the optical system and method to enable automatic high speed
measurements across several workpieces without the need for moving
or otherwise handling the workpieces manually.
[0035] In yet another embodiment of a method in accordance with the
invention, the optical system maintains a sine relationship between
pixels on the image detector and the altitude illumination angle
theta. One example of this relationship is such that displacement,
x, in the image plane corresponds to angle .THETA. so that x=F sin
.THETA., where F is some constant. The advantage to this
implementation is that an adequate number of pixels can be sampled
throughout the entire image plane. More specifically, for critical
sampling of the image plane, the sampling frequency should be twice
the highest spatial frequency in the plane. If the highest spatial
frequency does not depend on the position in the image plane, then
the number of pixels to be averaged (to effectively make a larger
pixel of the correct dimensions for critical sampling) does not
depend on the position in the image plane. For any other
distribution, the number of pixels required to be averaged will
depend on position. Thus, unless a sine relationship or another
suitable relationship between the pixels on the image sensor and
the altitude angle is maintained, then the fixed number of pixels
available could result in some regions of the image plane being
sampled with fewer than the optimal number of pixels.
[0036] Various embodiments of the invention are described in this
section to provide specific details for a thorough understanding
and enabling description of these embodiments. A person skilled in
the art, however, will understand that the invention may be
practiced without several of these details or additional details
can be added to the invention. Well-known structures and functions
have not been shown or described in detail to avoid unnecessarily
obscuring the description of the embodiments of the invention.
Where the context permits, singular or plural terms may also
include the plural or singular term, respectively. Moreover, unless
the word "or" is expressly limited to mean only a single item
exclusive from the other items in reference to a list of two or
more items, then the use of "or" in such a list is to be
interpreted as including (a) any single item in the list, (b) all
of the items in the list, or (c) any combination of items in the
list.
B. Embodiments of Scatterometers and Methods for Evaluating
Microstructures on Workpieces
[0037] FIG. 1 is a schematic illustration of a scatterometer 10 in
accordance with an embodiment of the invention. In this embodiment,
the scatterometer 10 includes an irradiation source 100 that
generates a beam 102 at a desired wavelength. The irradiation
source 100 can be a laser system and/or lamp capable of producing
(a) a beam 102 at a single wavelength, (b) a plurality of beams at
different wavelengths, or (c) any other output having a single
wavelength or a plurality of wavelengths. In many applications
directed toward assessing microstructures on semiconductor
workpieces, the irradiation source 100 is a laser that produces a
beam having a wavelength less than 500 nm, and more preferably in
the range of approximately 266 nm-475 nm. For example, the
wavelength can be about 375 nm-475 nm, or in some specific examples
about 405 nm or 457 nm. In a different embodiment, the irradiation
source 100 can include a plurality of different lasers and/or
filters to produce a first beam having a first wavelength of
approximately 266 nm and a second beam having a second wavelength
of approximately 405 nm, or in another embodiment the first beam
can have a wavelength of 405 nm and the second beam can have a
wavelength of 457 nm. It will be appreciated that the irradiation
source 100 can produce additional wavelengths having shorter or
longer wavelengths in the UV spectrum, visible spectrum, and/or
other suitable spectrum. The irradiation source 100 can further
include a fiber optic cable to transmit the beam 102 through a
portion of the apparatus.
[0038] The scatterometer 10 further includes an optical system 200
between the irradiation source 100 and a workpiece W. In one
embodiment, the optical system 200 includes a first optics assembly
210 that conditions the beam 102 to form a conditioned beam 212.
The first optics assembly 210, for example, can include a beam
diffuser/randomizer that diffuses and randomizes the radiation to
reduce or eliminate the coherence of the beam 102. The first optics
assembly 210 can also include a beam element that shapes the beam
to have a desired cross-sectional dimension, shape, and/or
convergence-divergence. The beam element, for example, can shape
the beam 212 to have a circular, rectilinear, or other suitable
cross-sectional shape for presentation to additional optic elements
downstream from the first optics assembly 210.
[0039] The optical system 200 can further include an object lens
assembly 300 that focuses the conditioned beam 212 for presentation
to the workpiece W and receives radiation reflected from the
workpiece W. The object lens assembly 300 is configured to receive
the conditioned beam 212 and form a convergent beam 310 focused at
a discrete focus area S on a desired focal plane, such as an object
focal plane 320. The convergent beam 310 can have a conical shape
when the conditioned beam 212 has a circular cross-section, but in
other embodiments the convergent beam 310 can have other shapes.
For example, when the conditioned beam 212 has a rectilinear
cross-section, the convergent beam 310 has a pyramidal shape. As
explained in more detail below with reference to Section C, the
convergent beam 310 can have a range of incidence angles having
altitude angles of 0.degree. to greater than approximately
70.degree. and azimuth angles of 0.degree. to greater than
90.degree. and more preferably 0-360.degree.. The altitude angle is
the angle between an incident ray and a reference vector normal to
the object focal plane 320, and the azimuth angle is the angle
between an incident plane and a reference vector in a plane
parallel to the object focal plane 320. The large range of
incidence angles generates a large number of unique data points
that enable accurate evaluations of several parameters of the
microstructure.
[0040] The focus area at the object focal plane 320 preferably has
a size and shape suitable for evaluating the particular
microstructure. For example, when the microstructure is a grating
or other structure on a workpiece having a maximum dimension of
approximately 10-40 .mu.m, then the focus area is also
approximately 10-40 .mu.m. The size of the focal area is preferably
not greater than the size of the microstructure so that the
radiation does not reflect from features outside of the particular
microstructure. In many applications, therefore, the object lens
assembly 300 is configured to produce a spot size generally less
than 40 .mu.m, and more preferably not greater than 30 .mu.m. The
scatterometer 10 can have larger focus areas in other embodiments
directed to assessing larger structures.
[0041] The object lens assembly 300 is further configured to
collect the scattered radiation reflecting or otherwise returning
from the workpiece W and present the scattered radiation on a
second focal plane 340. The object lens assembly 300, more
particularly, presents the scattered radiation in a manner that
provides a radiation distribution of the scattered radiation at the
second focal plane 340. In one embodiment, the object lens assembly
300 directs the scattered radiation coming at particular angles
from the object focal plane 320 to corresponding points on the
second focal plane 340. Additional aspects of specific embodiments
of the object lens assembly 300 are further described below with
reference to Section C.
[0042] The optical system 200 can further include a beam splitter
220 through which the conditioned beam 212 can pass to the object
lens assembly 300 and from which a portion of the return beam
propagating away from the second focal plane 340 is split and
redirected. The optical system 200 can optionally include a second
optics assembly 230 that receives the split portion of the return
beam from the beam splitter 220. The second optics assembly 230 is
configured to prepare the return beam for imaging by an imaging
device. Additional aspects of specific embodiments of the second
optics assembly 230 are described below with reference to Section
C.
[0043] The scatterometer 10 further includes a detector 400
positioned to receive the radiation distribution propagating back
from the second focal plane 340. The detector 400 can be a CCD
array, CMOS imager, other suitable cameras, or other suitable
energy sensors for accurately measuring the radiation distribution.
The detector 400 is further configured to provide or otherwise
generate a representation of the radiation distribution. For
example, the representation of the radiation distribution can be
data stored in a database, an image suitable for representation on
a display, or other suitable characterizations of the radiation
distribution. Several embodiments of the detector 400 are described
below in greater detail with reference to Section D.
[0044] The scatterometer 10 can further include a navigation system
500 and an auto-focus system 600. The navigation system 500 can
include a light source 510 that illuminates a portion of the
workpiece W and optics 520 that view the workpiece W. As explained
in more detail below, the navigation system 500 can have a low
magnification capability for locating the general region of the
microstructure on the workpiece (e.g., global alignment), and a
high magnification capability for precisely identifying the
location of the microstructure. Several embodiments of the
navigation system can use the irradiation source 100 and components
of the optical system 200. The navigation system 500 provides
information to move the object lens assembly 300 and/or a workpiece
site 510 to accurately position the focus area of the object lens
assembly 300 at the desired microstructure on the workpiece W.
[0045] The auto-focus system 600 can include a focus array 610, and
the optical system 200 can include an optional beam splitter 240
that directs radiation returning from the workpiece W to the focus
array 610. The auto-focus system 600 is operatively coupled to the
object lens assembly 300 and/or the workpiece site 510 to
accurately position the microstructure on the workpiece W at the
object focal plane 320 of the object lens assembly 300 or another
plane. As explained in more detail below with reference to Section
E, the navigation system 500 and the auto-focus system 600 enable
the scatterometer 10 to evaluate extremely small features of very
small microstructures on semiconductor devices or other types of
microelectronic devices.
[0046] The scatterometer 10 further includes a calibration system
for monitoring the intensity of the beam 102 and maintaining the
accuracy of the other components. The calibration system (a)
monitors the intensity, phase, wavelength or other beam property of
the beam 102 in real time, (b) provides an accurate reference
reflectance for the detector 400 to ensure the accuracy of the
scatterometer, and/or (c) provides angular calibration of the
system. In one embodiment, the calibration system includes a
detector 700 and a beam splitter 702 that directs a portion of the
initial beam 102 to the detector 700. The detector 700 monitors
changes in the intensity of the beam 102 in real-time to
continuously maintain the accuracy of the measured radiation
distribution. The detector 700 can also or alternatively measure
phase changes or a differential intensity. The calibration system,
for example, can use the polarity of the return radiation to
calibrate the system.
[0047] The calibration system can further include a calibration
unit 704 having one or more calibration members for calibrating the
detector 400. In one embodiment, the calibration unit 704 includes
a first calibration member 710 having a first reflectance of the
wavelength of the beam and a second calibration member 720 having a
second reflectance of the wavelength of the beam. The first
calibration member 710 can have a very high reflectance, and the
second calibration member 720 can have a very low reflectance to
provide two data points for calibrating the detector 400. In
another embodiment, the second calibration member 720 can be
eliminated and the second reflectance can be measured from free
space.
[0048] The scatterometer 10 further includes a computer 800
operatively coupled to several of the components. In one
embodiment, the computer 800 is coupled to the irradiation source
100, the detector 400, the navigation system 500, the auto-focus
system 600, and the reference detector 700. The computer 800 is
programmed to operate the irradiation source 100 to produce at
least a first beam having a first wavelength and preferably to also
produce a second beam having a second wavelength, as described
above. The computer 800 can also control the source 100 to control
the output intensity of the beam. The computer 800 further includes
modules to operate the navigation system 500 and auto-focus system
600 to accurately position the focus area of the convergent beam
310 at a desired location on the wafer W and in precise focus.
[0049] In several embodiments, the computer 800 further includes a
computer-operable medium for processing the measured radiation
distribution to provide an evaluation of the microstructure on the
workpiece W. For example, the computer 800 can include a database
having a plurality of simulated radiation distributions
corresponding to known parameters of the microstructure. The
computer 800 can include computer-operable media to process the
measured radiation distribution in conjunction with the database of
simulated radiation distributions in a manner that selects the
simulated radiation distribution that best fits the measured
radiation distribution. Based upon the selected simulated radiation
distribution, the computer stores and/or presents the parameters of
the microstructure corresponding to those of the simulated
radiation distribution, or an extrapolation or interpolation of
such parameters. In another embodiment, the computer 800 can scan
or otherwise acquire data from pixels of the detector only where
there is a high sensitivity to changes in the parameter(s). Such a
selective input to the computer reduces the amount of data and
increases the quality of the data for processing in the computer
800. Several aspects of the computer 800 and methods for processing
the measured radiation distribution are set forth below in greater
detail with reference to Section G.
C. Embodiments of Optics and Lens Assemblies
[0050] The scatterometer 10 can have several different embodiments
of optics assemblies and lens assemblies for optimizing the
scatterometer for use with specific types of microstructures. The
object lens assembly 300, for example, can be achromatic to
accommodate a plurality of beams at different wavelengths, or it
can have a plurality of individual assemblies of lenses that are
each optimized for a specific wavelength. Such individual lens
assemblies can be mounted on a turret that rotates each lens
assembly in the path of the beam according to the wavelength of the
particular beam, or such lenses may be mounted in separate, fixed
positions that correspond to the incident beam paths of the
respective wavelengths. In either case, the object lens assembly
300 is useful for applications that use different wavelengths of
radiation to obtain information regarding the radiation returning
from the workpiece.
[0051] The object lens assembly 300 can also include reflective
lenses that are useful for laser beams in the UV spectrum. Certain
types of glass may filter UV radiation. As such, when the beam has
a short wavelength in the UV spectrum, the object lens assembly 300
and other optic members can be formed from reflective materials
that reflect the UV radiation. In another embodiment, the first
optics assembly 210 or the object lens assembly 300 may have a
polarizing lens that polarizes the radiation for the convergent
beam 310 (FIG. 1).
[0052] FIG. 2 illustrates one embodiment of the convergent beam 310
explained above with reference to FIG. 1 formed by an embodiment of
the object lens assembly 300. The convergent beam 310 illustrated
in FIG. 2 has a frusto-conical configuration that results in a
focus area S. The focus area S is smaller than the area of the
microstructure under evaluation, but it generally covers at least
8-10 of the periodic structures of the microfeature. In several
particular applications for the semiconductor industry, the focus
area S is approximately 10-40 .mu.m in diameter, and more
preferably approximately 20-30 .mu.m in diameter. The focus area S,
however, is not limited to these ranges in other embodiments. The
focus area S may not necessarily be circular, and thus the
convergent beam 310 is typically configured such that the focus
area S has a maximum dimension less than 30 .mu.m (e.g.,
approximately 50 nm to approximately 30 .mu.m).
[0053] The convergent beam 310 simultaneously illuminates a
microfeature M through a wide range of incidence angles having
large ranges of altitude angles .THETA. and azimuth angles .PHI..
Each incidence angle has an altitude angle .THETA. and an azimuth
angle .PHI.. The object lens assembly is generally configured to
focus the beam to an area at the object focal plane through at
least (a) a 15.degree. range of altitude angles and (b) a
90.degree. range of azimuth angles simultaneously. For example, the
incidence angles can be simultaneously focused through altitude
angles .THETA. of 0.degree. to at least 45.degree., and more
preferably from 0.degree. to greater than 70.degree. (e.g.,
0.degree. to 88.degree.), and azimuth angles .PHI. of 0.degree. to
greater than approximately 90.degree., and more preferably
throughout the entire range of 0.degree. to 360.degree.. As a
result, the object lens assembly 300 can form a conical beam having
a large range of incidence angles (.THETA., .PHI.) to capture a
significant amount of data in a single measurement of the workpiece
W. This is expected to enhance the utility and throughput of
scatterometry for measuring critical dimensions in submicron
microstructures in real time and in-situ in a process tool.
[0054] FIG. 3 is a schematic diagram illustrating a specific
embodiment of the optical system 200 in accordance with the
invention. In this embodiment, the first optics assembly 210
includes a beam conditioner 214 that produces a conditioned beam
212 including diffused and randomized radiation. The beam
conditioner 214 can be a fiber optic line that transmits the beam
from the irradiation source (not shown in FIG. 3) and an actuator
that moves the fiber optic line to randomize the laser beam. The
actuator can move the beam conditioner 214 in such a way that it
does not repeat its movement over successive iterations to
effectively randomize the radiation.
[0055] The beam conditioner 214 can further include or
alternatively be an order sorter for removing undesired diffraction
orders from the output. For example, the beam conditioner 214 may
form a conditioned beam that provides a limited input to the object
lens assembly 300 so that only a single, specific diffraction
illuminates pre-selected parts of the detector. The beam
conditioner 214 may include a carousel of apertures placed at the
input of the optical system 200 so that different input apertures
may be selected according to the desired diffraction order of the
conditioned beam 212.
[0056] The first optics assembly 210 can further include a field
stop 216 and an illumination lens 218. The field stop 216 is
positioned in the first focal plane of the illumination lens 218,
and the field stop 216 can have an aperture in a desired shape to
influence the spot size and spot shape in conjunction with the
illumination lens 218. In general, the illumination lens 218
collimates the radiation for presentation to the object lens
assembly 300.
[0057] The embodiment of the object lens 300 illustrated in FIG. 3
can include a plurality of separate lenses. For example, the object
lens assembly 300 can include a divergent lens 302, a first
convergent lens 304, and a second convergent lens 306. The first
convergent lens 304 can have a first maximum convergence angle, and
the second convergent lens 306 can have a second maximum
convergence angle (see FIG. 4). In operation, the object lens
assembly 300 (a) focuses the conditioned beam 212 to form the
convergent beam 310 and (b) presents the return radiation from the
workpiece W on the second focal plane 340. The location of the
second focal plane 340 depends upon the particular configurations
of the lenses 302, 304 and 306. For purposes of illustration, the
second focal plane 340 is shown as coinciding with the location of
the first convergent lens 304.
[0058] The object lens assembly 300 is configured such that the
angle (.THETA..sub.x, .PHI..sub.y) of rays within the convergent
beam 310 will pass through corresponding points (x, y) in the
second focal plane 340. As a result, radiation passing through any
given point (x, y) in the second focal plane 340 toward the
workpiece W will pass through the object focal plane 320 at a
particular corresponding angle (.THETA..sub.x, .PHI..sub.y), and
similarly radiation reflecting from the object focal plane 320 at a
particular angle (.theta..sub.x, .PHI..sub.y) will pass through a
unique point (x, y) on the second focal plane 340. The reflected
radiation passing through the second focal plane 340 propagates to
the beam splitter 220 where it is directed toward the second optics
assembly 230.
[0059] The second optics assembly 230 includes a relay lens 232, an
output beam splitter 234, and an image-forming lens 236. The relay
lens 232 and output beam splitter 234 present the reflected and/or
diffracted radiation (i.e., return radiation) from the beam
splitter 220 to the image-forming lens 236, and the image-forming
lens 236 "maps" the angular distribution of reflectance and/or
diffraction (i.e., the radiation distribution) from the second
focal plane 340 to the imaging array of the detector 400. In a
particular embodiment, the image-forming lens 236 preferably
presents the image to the detector 400 such that the pixels of the
imager in the detector 400 can be mapped to corresponding areas in
the second focal plane 340.
[0060] The second optics assembly 230 can further include a
polarizing beam splitter 238 to separate the return radiation into
the p- and s-polarized components. In one embodiment, the
polarizing beam splitter 238 is positioned between the output beam
splitter 234 and the image-forming lens 236. In another embodiment,
the beam splitter 238 is positioned at a conjugate of the focal
spot on the wafer along a path between the image-forming lens 236
and the detector 400 (shown in dashed lines). In still another
embodiment, the polarizing beam splitter 238 can be located between
the relay lens 232 and the output beam splitter 234 (shown in
dotted lines). The polarizing beam splitter 238 is generally
located to maintain or improve the spatial resolution of the
original image of the focal spot on the workpiece. The location of
the polarizing beam splitter 238 can also be selected to minimize
the alteration to the original optical path. It is expected that
the locations along the optical path between the relay lens 232 and
the image-forming lens 236 will be the desired locations for the
polarizing beam splitter 238.
[0061] The polarizing beam splitter 238 provides the separate p-
and s-polarized components of the return radiation to improve the
calibration of the scatterometer 10 and/or provide additional data
for determining the parameter(s) of the microfeature on the
workpiece. For example, because the optics may perturb the
polarization of the input and output radiation, the polarizing beam
splitter 238 provides the individual p- and s-polarized components
over the large range of incidence angles. The individual p- and
s-polarized components obtained in this system can accordingly be
used to calibrate the scatterometer 10 to compensate for such
perturbations caused by the optical elements. Additionally, the p-
and s-polarized components can be used for obtaining additional
data that can enhance the precision and accuracy of processing the
data.
[0062] FIG. 3B is a schematic view of a cube-type polarizing beam
splitter for use in the scatterometer 10 shown in FIGS. 2 and 3A.
The cube-type polarizing beam splitter 238 receives a return
radiation beam 239 and splits it into a p-polarized component beam
239a and an s-polarized component beam 239b. The cube-type
polarizing beam splitter 238 can be a crystal with birefringence
properties, such as calcite, KDP or quartz. The p- and s-polarized
component beams 239a-b exit from the cube-type polarizing beam
splitter 238 along at least substantially parallel paths. The p-
and s-polarized beams 239a and 239b are also spaced apart from each
other such that they form separate images on the detector 400. To
increase the distance between the p- and s-polarized component
beams 239a-b, the size of the polarizing beam splitter 238 can be
increased. For example, as shown in dashed lines in FIG. 3B, a
larger polarizing beam splitter 238 results in at least
substantially parallel p- and s-polarized component beams 239a-b
that are spaced apart from each another by a larger distance than
the polarizing beam splitter 238 shown in solid lines 238. However,
large cube-type polarizing beam splitters can alter the p- and
s-polarized beams, and thus the size of polarizing beam splitter
238 is generally limited. As with the non-polarized return
radiation, the individual p- and s-polarized component beams 239a-b
impinge upon pixels of the detector 400 in a manner that they can
be mapped to corresponding areas in the second focal plane 340
shown in FIG. 3A.
[0063] One advantage of several embodiments of scatterometers
including cube-type polarizing beam splitters it that they provide
fast, high-precision measurements of the p- and s-polarized
components with good accuracy. The system illustrated in FIGS. 3A-B
use a single camera in the detector 400 to simultaneously measure
both of the p- and s-polarized components of the return radiation
239. This system eliminates the problems of properly calibrating
two separate cameras and registering the images from two separate
cameras to process the data from the p- and s-polarized components.
This system also eliminates the problems associated with serially
polarizing the return radiation beam using a mechanically operated
device because the polarizing beam splitter 238 can be fixed
relative to the return beam 239 and the detector 400.
[0064] Another aspect of several embodiments of the optics is that
a sine relationship or another suitable relationship is maintained
between the pixels on the image sensor and the altitude angles of
the beam. This allows a linear relationship between pixels on the
image sensor and altitude angles. As such, the optics enable good
sampling of the return radiation even at the peripheral regions of
an image sensor.
D. Embodiments of Detectors
[0065] The detector 400 can have several different embodiments
depending upon the particular application. In general, the detector
is a two-dimensional array of sensors, such as a CCD array, a CMOS
imager array, or another suitable type of "camera" or energy sensor
that can measure the intensity, color or other property of the
scattered radiation from the workpiece W corresponding to the
distribution at the second focal plane 340. The detector 400 is
preferably a CMOS imager because it is possible to read data from
only selected pixels with high repeatability instead of having to
read data from an entire frame. This enables localized or selected
data reading, which is expected to (a) reduce the amount of data
that needs to be processed and (b) eliminate data that does not
have a meaningful contrast. Additional aspects of using CMOS images
for image processing are described in more detail below. The p- or
s-polarized components can be measured with a single CMOS imager to
determine certain characteristics that are otherwise undetectable
from non-polarized light. As such, using a CMOS imager and
polarizing the reflected radiation can optimize the response to
increase the resolution and accuracy of the scatterometer 10.
[0066] FIG. 3C is a schematic view showing a CMOS imager assembly
for use in the detector 400 in accordance with an embodiment of the
invention. In this example, the CMOS imager assembly includes a die
410 having an image sensor 412, focal optics 420, and packaging 430
defining an enclosed compartment 432 between the die 410 and the
focal optics 420. The focal optics 420 typically have curved
surfaces or other configurations such that they are not merely a
plate having parallel, flat surfaces. Additionally, the CMOS imager
assembly does not have a glass cover or other optical member with
parallel, flat surfaces between the image sensor 412 and the focal
optics 420. As such, the CMOS imager assembly illustrated in FIG.
3C does not have any flat optics in the compartment 432 between the
image sensor 412 and the focal optics 420. In this embodiment, the
polarizing beam splitter 238 is just upstream of the CMOS imager
assembly 400 relative to the return radiation beam 239.
[0067] The CMOS imager assembly 400 illustrated in FIG. 3C is
expected to provide several advantages for use in scatterometers.
In several embodiments, for example, the lack of a cover or other
flat optical member between the image sensor 412 and the focal
optics 420 is expected to reduce perturbations in the return
radiation beam 239 at the image sensor 412. More specifically, a
glass member with parallel, flat surfaces between the focal optics
420 and the image sensor 412 can alter the return radiation just
before it reaches the image sensor 412. By eliminating such glass
members with parallel, flat surfaces, the CMOS imager assembly
illustrated in FIG. 3C is expected to eliminate distortion or
interference caused by a glass member with parallel surfaces.
E. Navigation and Auto-Focus Systems
[0068] Referring back to FIG. 1, the navigation system 500
accurately aligns the beam 310 with a desired area on the workpiece
W, and the auto-focus system 600 adjusts the object lens assembly
300 or workpiece site 510 so that the object focal plane 320 is at
the microstructure. In one embodiment, the navigation system 500
has a separate illumination source, lens and measurement optics for
determining the precise location of the microstructure on the
workpiece W. The light source of the navigation system 500 can be a
LED, and the lens and optics can be a two-stage system having low
and high magnifications. The low magnification stage identifies the
general area on the wafer where the microstructure is located, and
the high magnification stage refines the location. In other
embodiments, the navigation system 500 can include additional relay
optics introduced to image the surface directly through the object
lens assembly 300.
[0069] The auto-focus system 600 can be a camera correlation focus
system having a dihedral mirror that simultaneously splits the
illumination pupil in two and redirects the light from the two
halves of the dihedral mirror to different sections of a CCD array.
The displacement between the two images is used to automatically
determine the focus. A field stop can be incorporated to prevent
overlap of the two images on the focus camera. The field stop is
included in the illumination beam of the microscope of the
auto-focus system.
[0070] FIG. 4 is a schematic illustration of an embodiment of the
navigation system 500 and auto-focus system 600 for use in the
scatterometer. Several aspects of FIG. 4 are similar to those
explained above with reference to FIGS. 1 and 3A, and thus like
reference numbers refer to like components in these figures. The
navigation system 500 can have a high magnification system
associated with the metrology system. For example, the high
magnification system includes a light source 550, such as an LED,
that injects light via a beam splitter 552 and is focused on the
second focal plane by a relay lens 553 via beam splitter 240. This
light illuminates the workpiece and is reflected back through the
object lens assembly 300. The reflected light is directed by beam
splitter 220 and through lenses 232 and 554 to camera 560. The
lenses 232 and 554 form an image of the microstructure on the
camera 560.
[0071] The auto-focus system 600 in this embodiment shares the
relay lens 553 and the beam splitter 552 with the navigation
system. The beam splitter 552 directs a beam 620 to a dihedral
mirror 630, an image lens 632, and a steering mirror 634. The first
beam 620 is then received by an auto-focus detector 640, such as a
CCD array or other type of camera.
F. Calibration
[0072] The calibration system is used to monitor the properties of
the initial beam 102 (FIG. 1) and calibrate the system efficiency
for accurately detecting the radiation distribution. The beam
properties are monitored by a reference detector 700 that receives
a portion of the beam 102 in real time. As the beam fluctuates, the
reference detector 700 detects the changes in the beam 102 and
sends a signal to the computer 800. The computer 800 accordingly
normalizes and/or performs other computational operations to the
measured intensities, or it adjusts the measured radiation
distribution by the variances in the intensity of the initial beam
102, to compensate for small changes in the beam 102. Unlike some
systems that do this periodically, the computer 800 continuously
receives signals from the reference detector 700 to maintain the
accuracy of the system in real time. This is expected to
significantly enhance the accuracy and precision with which the
scatterometer 10 can evaluate extremely small features in
microstructures.
[0073] The calibration system can also include a calibration unit,
such as the calibration unit 704 (FIG. 1) with one or more
calibration members, for providing photometric calibration of the
system. In one embodiment, the first calibration member 710 can be
a highly reflective mirror having a reflectance greater than 95%,
and more preferably a reflectance of approximately 99.99%. The
first calibration member 710 can be configured to have a consistent
reflectance through a wide range of altitude angles. The second
calibration member 720 can be black glass having a low reflectance
(e.g., 0% to 10%). In operation, the detector 400 is calibrated by
measuring the reflectance of the beam from the first calibration
member 710 and from the second calibration member 720 to provide
two data points corresponding to the known 99.99% reflectance of
the first calibration member 710 and the known 0% reflectance of
the second calibration member 720. Using these two data points, a
straight line can be obtained to provide a reference reflectance of
the detector 400.
[0074] In another embodiment of the calibration unit 704, the
second calibration member 720 is not included or the second
calibration member can be free space such that there is no
reflectance of the illumination radiation. In this embodiment, the
scatterometer 10 is calibrated by obtaining a first reflectivity
from the first calibration member 710 and a second reflectivity
from an area separate from the first calibration member 710. When
the second reflectivity is obtained from free space, there is
approximately zero percent reflectance such that a straight line
can be obtained from these two measurements to provide a reference
reflectance of the detector 400.
[0075] The scatterometer can be calibrated further using several
different methods. For example, a known grating with a known
radiation distribution can be measured using the scatterometer 10
to determine whether the detector 400 accurately produces a
representation of the radiation distribution. In another
embodiment, a thin film having a known thickness can be irradiated
to determine whether the detector 400 provides an accurate
representation of the radiation distribution from such a thin film.
Both of these techniques can also be combined for yet another
calibration method.
G. Computational Analyses
[0076] The computer 800 can use several different processes for
determining one or more parameters of the microstructure based on
the measured radiation distribution from the detector 400. In
general, the computer 800 compares the measured radiation
distribution with one or more simulated radiation distributions
corresponding to selected parameters of the features and materials
of the microstructure (e.g., height, width, line edge roughness,
roundness of edge corners, spacing, film thickness, refraction
index, reflection index, and/or other physical properties). Based
on the comparison, the computer 800 then stores and/or provides an
output of one or more parameters of the microstructure.
[0077] FIG. 5A is an image illustrating a simulated radiation
distribution 810 having a first interference pattern 812 including
a plurality of thin arcs, a second interference pattern 814
including a plurality of different arcs, and a third interference
pattern 816 in a configuration of a "bulls-eye." The first
interference pattern 812 can correspond to the specular
reflections, the second interference pattern 814 can correspond to
higher order diffractions, and the third interference pattern 816
can correspond to the film thickness. The symmetry of the image can
also be assessed to provide additional information regarding the
microstructure. For example, in overlay applications, asymmetry in
the image can be used to evaluate the skew between overlay
structures (e.g., the extent of misregistration). Another example
of using the image symmetry is determining the asymmetry of
sidewall angles of grating lines. The interference patterns of the
simulated radiation distribution 810 are unique to each set of
feature parameters, and thus changing one or more of the feature
parameters will produce a different simulated radiation
distribution.
[0078] FIG. 5B is an image of a measured radiation distribution 820
of an actual microstructure on a workpiece. The measured radiation
distribution 820 includes a corresponding first interference
pattern 822, a second interference pattern 824, and a third
interference pattern 826. In operation, the computer 800 ascertains
the parameters of the microstructure by selecting and/or
determining a simulated radiation distribution 810 that best fits
the measured radiation distribution 820.
[0079] FIG. 6 illustrates one embodiment for ascertaining the
feature parameters of the microstructure. In this embodiment, the
computer 800 includes a database 830 including a large number of
predetermined simulated reference radiation distributions 832
corresponding to different sets of feature parameters. The computer
800 further includes a computer-operable medium 840 that contains
instructions that cause the computer 800 to select a simulated
radiation distribution 832 from the database 830 that adequately
fits a measured radiation distribution 850 within a desired
tolerance. The computer-operable medium 840 can be software and/or
hardware that evaluates the fit between the stored simulated
radiation distributions 832 and the measured radiation distribution
850 in a manner that quickly selects the simulated radiation
distribution 832 having the best fit with the measured radiation
distribution 850 or at least having an adequate fit within a
predetermined tolerance. In the case where a plurality of the
simulated radiation distributions 832 have an adequate fit with the
measured radiation distribution 850, the computer 800 can
extrapolate or interpolate between the simulated distributions.
Once the computer has selected a simulated radiation distribution
with an adequate fit or the best fit, the computer selects the
feature parameters associated with the selected simulated
distribution.
[0080] In an alternative embodiment, the computer calculates a
simulated radiation distribution and performs a regression
optimization to best fit the measured radiation distribution with
the simulated radiation distribution in real time. Although such
regressions are widely used, they are time consuming and they may
not reach a desired result because the regression may not converge
to within a desired tolerance.
[0081] In still other embodiments, the computer 800 may perform
further processing or different processing such as finite element
models for evaluating non-periodic or pseudo-periodic structures.
The computer 800 may also be able to solve for the refraction index
and reflectivity index of the particular materials by determining
the film thickness. Therefore, the enhanced data in the measured
radiation distribution enables the computer 800 to more accurately
determine the feature parameters of the microstructure and may
enable more feature structures to be monitored (e.g., line edge
roughness, refraction index, reflectivity index, etc.).
[0082] FIG. 7 is a flow chart showing a method 900 for ascertaining
one or more parameters of a microstructure using the computer 800
in accordance with another embodiment of the invention. In this
embodiment, the method 900 includes a first stage 910 comprising
irradiating a microstructure on a workpiece by passing a beam
through an object lens assembly that simultaneously focuses the
beam to a focus area at an object plane through a large number of
incidence angles. The beam for example, can be focused
simultaneously through angles of incidence having altitude angles
of 0.degree. to at least 45.degree. and azimuth angles of 0.degree.
to greater than 90.degree.. The method 900 further includes a
second stage 920 comprising obtaining an actual radiation
distribution of scattered light or other radiation returning from
the microstructure through the angles of incidence. The actual
radiation distribution can be obtained using an array of
addressable pixels that can be scanned or read individually. The
method 900 further includes a third stage 930 comprising acquiring
a measured selected radiation distribution by reading data from
selected pixels in the array that have sufficient sensitivity to
changes in the parameter based upon a predetermined sensitivity
record, and a forth stage 940 comprising fitting the measured
selected radiation distribution to simulated or modeled selected
radiation distributions corresponding to the selected pixels to
determine a value of the parameter. The method 900 is expected to
be particularly useful because selected areas of the pixel array
with high sensitivity to changes in the measured parameter can be
scanned, and then this smaller amount of data that is more
sensitive to changes in the measured parameter can be fitted to
modeled distributions.
[0083] The first stage 910 of the method 900 can be performed using
a scatterometer as described and shown above with reference to
FIGS. 1-4. As such, an input beam can be passed through an object
lens assembly that forms a beam having a large range of incident
angles (.THETA., .PHI.) to capture a significant amount of data in
a single measurement of the workpiece. The altitude angles .PHI.
can be from 0.degree. to approximately 80.degree. to 88.degree. and
the azimuth angles .THETA. can be from 0.degree. to 360.degree. as
explained above.
[0084] After irradiating the microstructure, the second stage 920
of the method 900 can include obtaining the actual radiation
distribution of radiation returning from the microstructure through
the same angles of incidence as described above with reference to
the detector 400. In this embodiment, the detector is preferably a
CMOS imager that has an array of addressable pixels in which
individual pixels can be independently scanned.
[0085] The third stage 930 of the method 900 comprises scanning or
otherwise acquiring the actual intensity measurements from selected
pixels of the array corresponding to angles of incidence that are
highly sensitive to changes in the measured parameter(s) based upon
a predetermined sensitivity record. The sensitivity record can be a
pixel-by-pixel analysis of a plurality of pixels in the array that
correspond to individual incidence angles (.PHI..sub.n,
.THETA..sub.n). At each pixel, model intensities of the return
radiation are calculated for different values of the parameter, and
then the model intensities are subtracted from each other to
determine the magnitudes of the changes in the intensity for the
incremental changes in the parameter. Larger intensity changes
correspond to pixels and angles of incidence that are more
sensitive to changes in the parameter compared to smaller changes
in the intensity.
[0086] FIG. 8 is a flow chart illustrating an embodiment of the
third stage 930 for use in the method 900. In this embodiment, the
third stage 930 includes identifying a pixel of the array
(procedure 932), and then calculating model intensities of the
return radiation at the pixel that correspond to different values
of the parameter (procedure 934). This embodiment of the third
stage 930 further includes differencing the model intensities
(procedure 936) to determine the magnitude of change in the
intensities corresponding to changes in the parameter. Based upon
the magnitudes of the changes in the model intensities determined
in procedure 936, the third stage 930 further includes assigning a
sensitivity value (procedure 938) to a corresponding pixel. A third
stage 930 can be repeated for any number of pixels in a CMOS imager
array to develop a predetermined sensitivity record that associates
the pixels in the array with the sensitivity to changes in the
parameter.
[0087] FIGS. 9A and 9B graphically illustrate an embodiment of
developing a predetermined sensitivity record for a parameter and
selecting pixels in the array that have sufficient sensitivity to
changes in the parameter for use in the measured selected radiation
distribution. FIG. 9A, more specifically, shows model intensity
distributions for three different values of a parameter (e.g., the
critical dimension) through angles of incidence where .THETA. is
from 0.degree. to approximately 65.degree. and .PHI. is from
0.degree. to approximately 90.degree.. In FIG. 9A, the intensity
distribution 952 corresponds to a first value of the parameter, the
intensity distribution 954 corresponds to a second value of the
parameter, and the intensity distribution 956 corresponds to a
third value of the parameter. Although the data in the intensity
distributions 952, 954 and 956 is useful, it does not provide an
indication of which incidence angles (e.g., pixels in a CMOS
imager) are more sensitive to changes in the parameter. FIG. 9B is
a sensitivity map 960 illustrating a sensitivity record associating
the sensitivity of various incidence angles to changes in the
parameter. The sensitivity map is obtained by differencing the
model intensities at corresponding incidence angles between the
intensity distributions 952, 954 and 956 shown in FIG. 9A. The
sensitivity map 960 shown in FIG. 9B assigns a sensitivity value
according to the magnitude of the difference between the model
intensity distributions. In this embodiment, bright regions
indicate angles of incidence that are more sensitive to changes in
the parameter and dark regions indicate angles of incidence that
are less sensitive to changes in the parameter. The sensitivity
values for the angles of incidence can then be associated with the
pixels in the CMOS imager array corresponding to the angles of
incidence.
[0088] One aspect of the method 900 is that the sensitivity map
960, which is a graphical representation of the sensitivity record,
provides an indication of the angles of incidence that may provide
the most valuable data corresponding to changes in the parameter.
Additionally, because specific pixels of a CMOS imager can be
scanned or read individually, one aspect of the method 900 is
determining which pixels in the CMOS imager array have high
sensitivities from which the measured intensity values can be
acquired (e.g., scanned).
[0089] FIG. 9C illustrates one embodiment of selecting pixels in
the array that have sufficient sensitivity to changes in the
parameter based upon the predetermined sensitivity record or
sensitivity map. Referring to FIG. 9C, pixels in the CMOS imager
array corresponding to angles of incidence in the bright region 962
are expected to provide good data because this area indicates a
region of high sensitivity, and pixels associated with angles of
incidence in the dark region 964 are likely to have low sensitivity
to changes in the parameter. As such, the measured selected
radiation distribution can be acquired by reading data from the
pixels in the array corresponding to angles of incidence that have
a high sensitivity to changes in the parameter based on the
predetermined sensitivity record. After scanning the pixels that
correspond to angles of incidence which have a high sensitivity,
the measured selected radiation distribution from such pixels is
compared to modeled radiation distributions from the same angles of
incidence until the measured selected radiation distribution
adequately fits with one of the modeled selected radiation
distributions. At this point, a value of the parameter is
determined according to the corresponding best fit of the modeled
selected radiation distribution.
[0090] The method 900 can be applied to applications in which
several parameters of the microfeature are to be assessed using
scatterometry. In such multi-parameter applications, separate
sensitivity records are established for individual parameters by
varying the parameter of interest while keeping the other
parameters constant. The regions of the CMOS imager array that are
scanned can then be determined by selecting the sets of pixels
corresponding to high sensitivity values for the parameters that
are to be assessed using the scatterometer.
[0091] In practice, the sensitivity record can be used to optimize
the scan path to retrieve data from only the pixels that have
sufficient sensitivity, and to also fit the data from the selected
pixels to corresponding models in a library. This is expected to
significantly reduce the simulation times for arriving at a value
of one or more parameters because it reduces the number of
(.THETA., .PHI.) combinations that need to be stored in a library
and used to fit the measured selected radiation distribution to a
modeled selected radiation distribution. The system is also
expected to increase the acquisition rate because only a portion of
the pixels in the CMOS imager need to be scanned for each
measurement. This provides more measurements in a given exposure
period, which can lead to better averaging and lower
signal-to-noise ratios. This method is also expected to enhance the
precision (e.g., repeatability) because it uses only data from
highly sensitive pixels. Additionally, this method is useful
because it is possible to overlay known pixel noise for a
particular CMOS imager array with the sensitivity array such that
pixels with high sensitivities caused by noise can be eliminated
from the measurements. Therefore, such an image fitting procedure
that scans only selected pixels of a CMOS imager array with high
sensitivity values is expected to significantly improve the
computational analysis for determining values of parameters in
scatterometry.
[0092] From the foregoing, it will be appreciated that specific
embodiments of the invention have been described herein for
purposes of illustration, but that various modifications may be
made without deviating from the spirit and scope of the invention.
Accordingly, the invention is not limited except as by the appended
claims.
* * * * *