Method of evaluating imaging performance

Details Method of evaluating imaging performance Method of evaluating imaging performance

1999-09-13

2003-05-13

Font, Frank G. (Department: 2877)

Optics: measuring and testing

Lens or reflective image former testing

C356S401000

Reexamination Certificate

active

06563573

ABSTRACT:

FIELD OF THE INVENTION AND RELATED ART
This invention relates to a method of evaluating the imaging performance of an imaging optical system. More particularly, the invention is concerned with an imaging performance evaluating method for use in performance inspection of an imaging optical system in a registration inspecting apparatus, for example, which is useable for inspection of an optical performance such as distortion or alignment precision, of a projection optical system in a semiconductor device manufacturing exposure apparatus of step-and-repeat type or step-and-scan type, for example.
In projection exposure apparatuses for the manufacture of semiconductor devices, a performance for printing, through projection exposure, a circuit pattern of a reticle onto a wafer with high resolution is required to meet further increases in density of an integrated circuit. In an attempt to improve the resolving power for the projection of a circle pattern, a method in which the numerical aperture (NA) of a projection optical system is enlarged while keeping the wavelength of exposure light fixed, a method of shortening the wavelength of exposure light (e.g., changing from g-line to i-line, from i-line to the emission wavelength of an excimer laser or to that of an F
2
laser, or a method using SOR light), have been proposed and developed.
On the other hand, in the stream of further miniaturization of a circuit pattern, a performance for aligning a wafer and a reticle having an electronic circuit pattern formed thereon very precisely, is also required. Generally, the required alignment precision is about one-third or less of the linewidth of a circuit pattern. For a 1G bit DRAM, for example, if the circuit pattern is based on 0.18 micron rule, an overlay precision (alignment through the whole exposure region) of 60 nm or less is required.
Further, in a registration inspecting apparatus wherein this overlay precision is measured, a precision of about one-tenth of the overlay precision is required. For a 1G bit DRAM, a precision of 6 nm or less is necessary.
For higher precision measurement in such registration inspecting apparatuses, a method called a TIS (Tool Induced Shift) correction method wherein the influence of TIS, which is a detecting system factor among the measurement error producing factors, is reduced, has been proposed and developed.
FIG. 1A
is a schematic view for explaining this TIS correction method.
FIG. 1A
shows an example wherein a surface step (difference in level) is defined on a silicon wafer
1
by an etching process, and wherein the relationship between an etching pattern (first mark)
2
based on the level difference and a resist image pattern (second mark)
3
having been printed and developed after the alignment process is going to be measured. In accordance with the TIS correction method, the measurement is made twice. Here, the second time measurement is performed while rotating the wafer
1
by 180 degrees as compared with the first time measurement. The result of such a first time measurement is called “0-deg. measured value” (&Dgr;
0-deg
), and the result of a second time measurement is called “180-deg. measured value” (&Dgr;
180-deg
). In the TIS correction method, a value obtainable by dividing, by 2, the difference of a 0-deg. measured value minus a 180-deg. measured value, is used as a measured value. On the basis of this, an error in the detection system factor is reduced, whereby high precision measurement can be accomplished. Here, the value obtainable by dividing, by 2, the sum of the 0-deg. measured value plus the 180-deg. measured value, is called “TIS”.
Use of such a TIS correction method will be effective to reduce the measurement error of the detection system factor. However, it is still insufficient. For example, even if one and the same wafer is measured in accordance with the TIS correction method by using two different measuring machines, there may occur a large difference, as shown in FIG.
1
B. As regards this alignment precision, since the alignment sequence of the exposure apparatus is based on the global alignment method, it does not become worse beyond “root 2” times the precision of a stage system of the exposure apparatus, having an interferometer and for driving the wafer.
The result of the measurement by the registration inspecting apparatus #
2
of
FIG. 1B
is, however, worse by much more than “root 2” times the precision of the stage system for driving the wafer and having an interferometer. Thus, clearly, it is attributable to a factor in the registration inspecting apparatus.
This results from the TIS-WIS interaction on an occasion when there is a WIS (Wafer Induced Shift), which is a wafer factor among the measurement error producing factors. In this example, the insufficiency of the result based on the TIS correction is clearly seen in FIG.
1
B. New semiconductor processes such as a Cu-CMP process, for example, will be introduced successively and, in those cases, there will still be WIS present. In consideration of this, for improvement in precision of a registration inspecting apparatus, it is desirable that the TIS be removed as much as possible to thereby prevent the TIS-WIS interaction.
In the TIS correction method, the measurement has to be done twice, at 0 degree and 180 degree. This is a large problem in relation to the throughput. In current mass-production of semiconductor devices, therefore, the TIS correction method is not used prevalently.
As regards removal of TIS, the cause of production of TIS due to an optical factor as well as conventional examples therefor will be described. Most of currently used registration inspecting apparatuses or alignment detecting systems are based on a bright-field image processing process.
FIG. 2
is a schematic view of an example of a registration inspecting apparatus. In this example, a special mark (first and second marks
2
and
3
) is provided on a wafer
1
, and an image of the mark is formed through an optical system upon an image pickup device such as a CCD, for example. An electric signal therefrom is processed, by which the positions of the first and second marks
2
and
3
are detected. The imaging performance which is most necessary in this optical system is the symmetry of images corresponding to the first and second marks
2
and
3
. If there is something in the optical system (imaging optical system) that deteriorates the image symmetry, it means that there is TIS present.
In these types of alignment detecting systems, the magnification is made high (e.g., about 100×) and, in most cases, it is used on or in the proximity of the axis. For this reason, the major cause for deterioration of symmetry of the mark image is not an off-axis aberration but an eccentric coma aberration close to the axis of the optical system and the non-uniformness in an illumination system.
Further, it has been found that the symmetry of a mark image changes with the amount of surface level difference (height of a surface-step) upon a wafer to be measured.
FIG. 1C
is a schematic view for explaining a first mark
2
and a second mark
3
formed on the wafer
1
of
FIG. 2
, and it illustrates the manner of calculating the symmetry SOI (Symmetry of Image) to be produced by a signal (waveform intensity) which is based on the first and second marks
2
and
3
.
When the waveform intensities at the opposite ends
2
a
and
2
b
of the first mark
2
are denoted by a and b, respectively, and when the waveform intensity based on the wafer top surface is denoted by c, the symmetry SOI can be determined in accordance with:
SOI=100(
a−b
)/
c.
FIG. 1D
shows the results of the symmetry SOI of the optical image as defined with reference to
FIG. 1C
, which were obtained experimentally while changing the amount of surface level difference SH. In the experiments, the symmetry SOI of the image was measured while taking the eccentric coma aberration of the optical system as &lgr;/4 and −&lgr;/20 (where &lgr; is the illumination wavelength).
As seen from
FIG. 1D
,

Affiliated with

Also associated with

Rating

Method of evaluating imaging performance does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Method of evaluating imaging performance, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of evaluating imaging performance will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3017761