 |
Description  |
|
|
RELATED APPLICATION
The present application is related to application Ser. No. 154,469, field
Feb. 8, 1988, now U.S. Pat. No. 4,810,895, which is a continuation of
application Ser. No. 07/003,055 filed Jan. 13, 1987 (now abandoned) by
Oded Kafri and Ilana Glatt, and assigned to the same assignee as the
present application.
BACKGROUND OF THE INVENTION
The above-cited patent application Ser. No. 07/154,469 relates to a method
and apparatus for optical examination of an object, particularly by moire
ray deflection mapping. The present application is directed to the
extension of the basic setup described in that patent application so as to
enable the apparatus also to operate as a Fizeau interferometer and also
as a schlieren device, as well as a moire ray deflection mapper.
There are two different approaches to optical metrology, namely:
interferometry, which measures phase retardation between two light beams,
and ray-deflection analysis. The latter was first utilized in schlieren
photography, and later in schlieren interferometry and moire
deflectometry. patent application Ser. No. 07/154,469 describes an
instrument which operates as a moire deflectometer. The invention of the
present application shows how the basic telescopic instrument of that
patent application can be used to produce an instrument which can operate
independently not only as a moire deflectometer, but also as a Fizeau
interferometer or as a schlieren device. As will be described more
particularly below, such an instrument can be used in all three modes of
operation without moving any optical components, and therefore offers a
unique opportunity to compare the three methods. Moreover, the basic
instrument can be used to operate according to all three methods
simultaneously, and therefore the combinations of Fizeau and moire, moire
and schlieren, and Fizeau and schlieren, can complement each other to
produce improved results.
More particularly, the above-cited application Ser. No. 07/154,469
discloses a method, and also apparatus, for optical examination of an
object involving the steps: providing a point source of light producing a
diverging beam of direct light; directing the diverging beam of direct
light to a first optical system including the object to be examined, which
system retraces the light in the form of a beam of reflected light from
the examined object back towards the point source of light; intercepting
the converging beam of reflected light before reaching the point source of
light; passing the intercepted converging beam of reflected light through
a second optical system which collimates the beam of reflected light; and
examining the collimated beam of reflected light. That patent application
describes the use of the method for moire ray deflection mapping, wherein
the examination of the collimated beam of reflected light is effected by
directing the collimated beam through first and second gratings at a
preselected angular orientation and distance with respect to each other to
produce moire fringe patterns providing an indication of the properties of
the examined object.
BRIEF SUMMARY OF THE PRESENT INVENTION
As indicated earlier, the invention of the present application may be used
for extending the basic setup described in patent application Ser. No.
07/154,469 so as not only to operate as a moire deflectometer, but also to
operate as a Fizeau interferometer, or as a schlieren device.
According to the present invention, the first optical system in the setup
described above includes a partial transmittance reference plate, whereby
a contour map of the object topography is obtained from the interference
between the direct light beam and the reflected light beam.
The second optical system includes a schlieren filter, such as a knife-edge
spatial filter or any other spatial filter, at the focal point of the
reflected beam to cut down the Fourier transform of the image, whereby a
schlieren image is produced at the viewing device. The apparatus further
includes means for selectively locating the partial transmittance
reference plate or the schlieren filter in its respective position in the
first and second optical systems.
Further features and advantages of the invention will be apparent from the
description below.
BRIEF DESCRIPTION OF THE DRAWINGS
The invention is herein described, by way of example only, with reference
to the accompanying drawings, wherein:
FIG. 1 illustrates the basic telescope setup used in the apparatus of the
above-cited application Ser. No. 0./154,469, and also used in the
invention of the present application;
FIG. 2 illustrates the basic setup of FIG. 1 modified to operate as a moire
deflectometer, as also described in application Ser. No. 07/154,469, FIGS.
2a-2c illustrating moire fringe patterns produced by the moire
deflectometer;
FIG. 3 illustrates the basic setup of FIG. 1 modified to operate as a
Fizeau interferometer, FIGS. 3a-3c illustrating the patterns produced when
so operated; and
FIG. 4 illustrates the basic setup of FIG. 1 modified to operate as a
schlieren device, FIGS. 4a, 4b, illustrating the patterns produced when so
operated.
DESCRIPTION OF PREFERRED EMBODIMENTS Basic Telescope Setup of FIG. 1
FIG. 1 illustrates the basic telescope setup for examining phase objects.
As described in application Ser. 07/154,469, the same basic setup can be
used, with minor modification, also for examining specular objects having
curved specular surfaces.
The apparatus illustrated in FIG. 1 comprises a point source of light 2
which produces a diverging beam of light., This point source of light 2
may be a laser producing a collimated beam, which is subsequently passed
through a divergent lens to produce the diverging beam. The diverging beam
passes through a beam splitter 4 and is directed to an optical system,
generally designated 10, which includes the phase object to be examined,
the latter being designated 12.
Optical system 10 further includes an objective lens 14 which collimates
the light from the point source 2 before the light passes through the
phase object 12, and a flat reflective surface 16 which reflects the light
passing through the phase object 12 back through the phase object to the
objective lens 14. The latter lens converges the reflected beam after
passing through the phase object 12 and directs back towards the point
source 2. Optical system 10, including objective lens 14, the examined
phase object 12, and the reflective surface 16, thus causes the light
beam, after passing twice through the phase object 12, to be retraced in
the form of a converging beam back towards the point source of light 2.
Beam splitter 4 intercepts the converging beam of reflected light before
reaching the point source 2 and directs the beam to a second optical
system, generally designated 20, including a second objective lens 22.
Optical system 20 collimates the beam reflected from beam splitter 4 and
directs the collimated beam to a viewing device in the form of a matt
screen 24.
The above-described basic telescope setup illustrated in FIG. 1, together
with the setup modified to operate as a moire deflectometer as illustrated
in FIG. 2, is described in more detail in application Ser. No. 07/003,055.
Thus, to operate the system as a moire deflectometer, optical system 20,
after collimating the beam reflected from beam splitter 4, directs the
collimated beam through first and second ratings G.sub.1, G.sub.2, at a
preselected angular orientation and distance with respect to each other,
to produce moire fringe patterns on the matt screen 24, the matter being
attached to grating G.sub.2. The moire fringe patterns may be used,
according to known techniques, to provide an indication of the properties
of the examined phase object 12.
The focal length of optical system 10 is larger than that of optical system
20, thereby reducing the image directed through the first and second
gratings G.sub.1, G.sub.2. This increases the sensitivity. In addition,
lens 14 in optical system 10 may be mounted so as to be movable towards
and away from the examined phase object 12, to maintain their axial
approximation, thereby enabling the apparatus to be used for short
focal-length lenses.
As further described in application Ser. No. 07/154,469, the setup
illustrated in FIG. 2 is unlike the classical deflectometer setup where
beam expansion is achieved by a reverse telescope, comprising a microscope
objective and an off-axis telescope mirror attached to a laser. Rather,
the setup illustrated in FIG. 2 uses a Newtonian-type telescope, like the
Fizeau interferometer. After the laser beam is expanded to the required
width and has passed through the large objective lens 14, it passes
through the phase object 12 (assuming that it remains parallel with the
paraxial approximation), and is then reflected back into the telescope by
the flat mirror 16, thus passing twice through the phase object. The
retraced beam is diverted 90.degree. by beam splitter 4 to the smaller
objective lens 14 where it is recollimated. Now, a small diameter
deflectometer may be used to detect redeflections.
As further described in patent application Ser. No. 07/154,469 the setup
illustrated in FIG. 2 can be easily modified to measure flat specular
objects simply by replacing the flat reflector by the object to be
examined. The optical system including the examined object may also
include a large objective lens, corresponding to lens 14 in FIGS. 1 and 2,
but this is not essential in a setup for examining a concave specular
object.
Further information concerning the structure of the moire deflectometer
illustrated in FIG. 2, including its mode of operation and results
produced, are set forth in application Ser. No. 07/154,469.
The Basic Setup used as a Fizeau Interferometer
FIG. 3 illustrates the basic setup of FIG. 1 modified so as to operate as a
Fizeau interferometer. This may be done merely by adding a partial
transmittance reference plate 30 to the optical system 10 in the basic
setup of FIG. 1. A contour map of the object topography is thus obtained
from the interference between the reflected beam and the direct beam. Each
fringe represents a change of elevation of one-half wavelength
(.lambda./2). When testing a phase object 12, it is placed between the
reference plate 30 and the reflector surface 16, and a contour map of the
optical path will be obtained instead of the height contour map.
It will thus be seen that by simply adding the partial transmittance
reference plate 30 to optical system 10, the basic setup of FIG. 1 is
converted for use as a Fizeau interferometer. However, when so operated,
the following considerations should be noted:
1. The height contour map is highly sensitive to the tilt of the test
object. If the slope of the object relative to the reference plate is
changed, a different contour map is obtained. Therefore, one must
calculate the derivatives which are not affected by a constant slope.
2. The ability to distinguish between hills and valleys is also a problem.
A contour map cannot detemine whether the fringes represent increased,
decreased, or equal elevation. This must be determined by mechanically
shifting the object and observing the relative movement of the fringes.
3. The reference plate and the test object (and the mirror for phase
objects) must be stable within .lambda./10 during the measurement time.
This condition mandates the use of vibration isolated tables when working
in an industrial environment.
4. Because of the laser coherence length, the test object must be placed
close to the reference plate in order to obtain good quality fringes. This
might not be so easy in large phase objects because the object is placed
between the reference plate and the mirror. Moreover, the reflectivity of
the test object and the reference plate must be similar in intensity.
The main advantage of interferometry is its high sensitivity which results
in the ability to measure very small deviations. Unfortunately, this high
sensitivity also results in an inability to analyze results from
relatively large (compared to .lambda.) deviations.
The first object tested was a moderate quality beam splitter, and the
contour map obtained by the Fizeau interferometer is shown in FIG. 3a. A
second beam splitter contained relatively large slopes, and a complete
mapping was unobtainable. A slight tilt of the object exhibited a
completely different mapping, as shown in FIGS. 3b and 3c, and reproducing
the results was difficult.
Basic Setup Operated as a Schlieren Device (FIG. 4)
The conventional schlieren technique is a semi-quantitative one. The basic
setup illustrated in FIG. 1 can be altered so as to operate as a schlieren
device by placing a knife-edge spatial filter, or any other spatial
filter, at the focal point of the reflected beam. This spatial filter
simply cuts the Fourier transform of the image.
Thus, as shown in FIG. 4, the illustrated schlieren device includes the
basic setup illustrated in FIG. 1, but with the addition of a knife-edge
spatial filter 40 in the second optical system 20 between the beam
splitter 4 and the objective lens 22. Ignoring diffraction effects, if the
test object is flat the reflected beam will remain collimated, and all of
the rays will be focussed to a infinitely small point. If the object
contains gradients in the X-direction, some rays will be focussed below or
above the focal point 40. To determine the slopes in the Y-direction,
either the object or the knife-edge must be rotated by 90.degree..
The knife-edge filter 40 causes the image to appear brighter or darker
depending the direction of the gradient. The relative magnitudes can be
estimated by slowly bringing the knife-edge towards the focal point and
observing when an area becomes dark. Hills and valleys can be
distinguished in that the first areas to become dark will be the positive
gradients, followed by the flat areas, and finally the negative gradients.
Exact magnitudes can be determined by precalibrating the knife-edge
position to a given gradient of the object. The more quantitative
measurement will be sensitive to the tilt of the test object.
The main disadvantage of the schlieren device is that it is only
semi-quantitative, and in order to receive more quantitative measurements
the test object alignment becomes a factor. Furthermore, the measurement
is dynamical, namely the movement of the knife-edge during the test. Some
advantges are the low stability requirement, the easy determination of
relative slope, and the high contrast results. FIGS. 4a and 4b illustrate
two schlieren photographs of the object of FIG. 3a. In FIG. 4a, the
knife-edge cuts a small part of the deflected rays from the Fourier
transform of the object. In FIG. 4b, a larger portion of the rays are cut.
The high positive slopes of FIG. 3a are detected in FIG. 4a, and in FIG.
4b some smaller slopes are also detected.
Comparison and Combined Operation
As shown above, the operation of the basic telescopic setup may be greatly
modified by simple alternations. By inserting a reference plate in the
area of the test object, a Frizeau interferometer is achieved; by adding a
knife-edge to the focal plane, schlieren setup is obtained; and by placing
two Ronchi rulings behind the small objective lens, a moire deflectometer
is obtained. Since each of these additions is placed in a different area
of the setup, it is possible to apply any combination of the techniques
simultaneously. As will be demonstrated, one can benefit from this
property. Since it would be desirable to utilize the relative advantages
of each method while limiting the disadvantages, the three methods will
first be compared.
1. Fringe Interpretation. The basic difference between interferometric
techniques which measure phase retardation, and ray delfection techniques
such as schlieren and deflectometry, is the quantity which is mapped.
Heights are mapped in interferometry and slopes in deflectometry and
schlieren. It is true that it is much easier to visualize the shape of an
object from a map of height gradients, but this technique is not alway
accurate since any small tilt of the object will produce a change in the
map of the height gradients. For a field such as flatness analysis (of
silicon wafers, hard disks, mirrors, etc.) one would like to map a
quantity that is invariant to the tilt of the object and will not map a
constant slope. This quantity is the curvature of the fringes, in other
words, the second derivitive of the height.
In interferometry, the measured quantity is height contour, and a curved
object will produce curved fringes. A finite tilt of the object will
change the number of fringes, but not their curavture. Therefore, the
analysis is done on the curvature of the fringes. In moire deflectometry a
slope contour map is obtained, and a constant slope will cause a change
only in the phase of the fringes. In schlieren, as stated earlier, in
order to achieve more quantitative results, the position of the knife-edge
is precalibrated and therefore affected by a constant tilt of the object.
Therefore, because of its invariance to object tilt, deflectometry is
preferred.
2. Mechanical Stability Requirement. Moire delfectometry and schlieren
photography are ray tracing methods, and the averaging is done at the
viewing device. This means that the system stability should be one order
of magnitude greater than the required measurement sensitivity. In Fizeau
interferometry, the stability requirements depend on the interference
phenomenon between two waves. Therefore, no matter what the required
sensitivity is, the two interfering beams should be stable within
.lambda./10. In other words, the refernce plate and the test object must
be in unison up to .lambda./10 requiring expensive tables and a
"laboratory like" environment.
3. Sensitivity. In interferometry, the sensitivity is determined by the
light wavelength , The way to calculate the slope is to divide the actual
increment between two fringes .lambda./2 by the observed distance between
them. If one measures the fringe intensity, and we assume that one can
resolve 1/2.pi. of a fringe without imaging processing, we obtain the
equation
.delta..beta.=.lambda./4.pi.a
where .delta..beta. is the minimum slope measurement and a is the test
object diameter. The minimum sensitivity of moire deflectometry, which is
identical to shearing interferometry, is determined by the uncertainty
principle
.delta..phi..delta..chi..gtoreq..lambda./2.pi.
where .delta..phi. is the error in the deflection angle determination and
.delta..chi. is the spatial resolution. If we substitute
2.delta..beta.=.delta..phi. and .delta..chi.=a, we obtain the identical
result as in interferometry, but the spatial resolution is sacrificed.
In order to receive maximum quality results, a system must be tuned to the
appropriate sensitivity. As it was mentioned, the deflectometer can be
tuned to the exact desired sensitivity by merely changing the distance
between the two gratings. The sensitivity in interferometry can be reduced
using a technique called grazing incidence, but this technique is costly
compared to its results. The sensitivity in schlieren is similar to that
of interferometry and is determined by the micrometer moving the
knife-edge.
(a) Combined Operation--Moire and Schlieren
The infinite fringe mode of moire cannot distinguish between hills and
valleys, and therefore, it is helpful to add a schlieren device that can
easily distinguish between the positive and negative slopes. For example,
in the test object of FIGS. 3a and the infinite fringe moire pattern in
FIG. 2a, one can see all the slope deviations but cannot determine the
signs of the slope. The schlieren device was added, and the first area
darkened, which can be seen in FIG. 4a, is the area of the largest
positive slope. By moving the knife-edge towards the focal point,
additional areas are darkened that follow along the lines of the moire
fringes. Therefore while the schlieren device determines relative slopes,
the moire infinite fringe mode determines the exact slope increments.
(b) Combined Operation--Interferometry and Moire
As mentioned earlier, both an advantage and disadvantage of interferometry
is its high sensitivity. As seen in FIG. 3b, the areas of relatively small
height gradients produce a sensitive contour map. In the areas where the
height gradients were relatively large, compared to .lambda., a blur of
indistinguishable fringes was produced. When combining the infinite fringe
moire pattern with interferometry, we achieved a slope contour map in the
areas where it was impossible to obtain a height contour map (see FIG.
2c). From this map we now have an idea of what the object looks like in
the areas of large height deviations. Although the two sets of fringes are
not measuring the same quantities, the moire fringes are only a derivative
and therefore it is relatively simple to correlate the results.
(c) Combined Operation--Interferometrv and Schlieren
Although in Fizeau interferometry the hills and valleys problem can be
solved by a dynamic movement of the object during testing, it is much
simpler to apply the schlieren technique and observe the gradual darkening
of the object. The schlieren does not follow along the lines of the
fringes as in the moire-schlieren combination, rather, since schlieren is
the derivative of interferometry, the denser fringes, that represent the
highest positive slope, will darken first. These areas are followed by the
areas of very few fringes, and lastly, the denser fringes that represent
the highest negative slope.
SUMMARY
The above description shows how the basic telescope setup of FIG. 1 may be
operated as a moire deflectometer, Fizeau interferometer, schlieren
device, and various combinations. Fizeau interferometry was found to be
most suitable for high sensitivity measurements of near perfect objects.
Deflectometry must be used for objects that require a lower sensitivity
analysis, although it can also be used for high sensitivity measurements.
Deflectometry is also preferred in flatness analysis where the object tilt
will affect the interferometric results. Schlieren is a desirable
non-qualitative add on which can determine hills and valleys in a very
simple manner without the need to do fringe interpretation (finite fringe
deflectometry) or dynamically move the object (interferometry).
The combined operation of deflectometry and interferometry is an ideal
solution for analyzing objects with a wide range of slopes. One can
utilize the fixed high slopes of interferometry for the areas of the
object with relatively small height deviations, and reduced sensitivity
moire for the areas of relatively high slopes as was shown in FIG. 2c.
The same instrument of FIG. 1 can also serve as a shearing interferometer
by removing the two gratings G.sub.1, G.sub.2 in FIG. 2, and substituting
a shearing device, as known in shearing interferometry. The instrument of
FIG. 1 can also serve as a device to determine stress via the photoelastic
effect. Linearly polarized laser light combined with a polarized beam
splitter will yield high quality fringes representing different levels of
stress. This phenomenon is due to the photoelastic effect.
While the invention has been described with respect to one preferred
embodiment, it will be appreciated that many other variations,
modifications and applications of the invention may be made.
* * * * *
|
|
 |
|
 |
Description |
|
