Image analysis – Applications – Manufacturing or product inspection
Reexamination Certificate
2000-09-07
2002-01-01
Mehta, Bhavesh (Department: 2621)
Image analysis
Applications
Manufacturing or product inspection
C430S005000, C430S030000, C716S030000
Reexamination Certificate
active
06335981
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a photomask pattern correcting method for correcting a pattern of a photomask used when exposing in a lithography process which is one of the manufacturing processes of a semiconductor device, and to a photomask corrected by the same, and to a photomask pattern correcting device.
BACKGROUND OF THE INVENTION
In a lithography process, which is one of the manufacturing processes of a semiconductor device, various types of light energies such as visible light, UV light, or electron beam are projected on a target for exposure so as to transfer a desired pattern thereon.
In recent years, more refined and highly integrated semiconductor elements have been developed. In response to this, in the lithography technique, in order to achieve a minimum processing dimension of not more than 0.1 &mgr;m, a super resolution lithography technique capable of processing with a dimension substantially the same as or less than the wavelength of the exposing light has been developed for practical use.
The practical limit of a resolution of pattern exposure is determined by various factors. Of those factors, in response to refining of patterns in recent years, the optical proximity effect has been one of the main factors determining the resolution limit. The proximity effect refers to a problem which is caused by the interference effect of a radiation energy such as light between proximate patterns. Such a problem includes deformation of a transfer pattern caused by interference within a single pattern.
In the conventional lithography process, since the size of a transfer pattern is sufficiently large compared with the wavelength of exposing light, the problem of proximity effect is not caused. In the super resolution lithography technique, however, the proximity effect phenomenon is a big problem.
When the size of a transfer pattern is substantially the same as or less than the wavelength of exposing light, due to the proximity effect phenomenon and receding of edges (especially at line end) and a pattern deformation phenomenon of a photoresist during development, a difference in line-width and shape is generated between a pattern of a photomask and a pattern transferred onto the photoresist.
For this reason, in the super resolution photolithography, in order to form a photoresist having a desirable pattern, it is one of the most important techniques to accurately estimate the degree of deformation due to the optical proximity effect, etc., generated when transferring, and to correct a photomask pattern. The same can be said for the lithography technique adopting an electron beam which has a large interaction, or other types of light energies.
To present, various attempts have been made to correct a photomask pattern by accurately estimating the optical proximity effect. For example, Japanese Unexamined Patent publication No. 189913/1990 (Tokukaihei 2-189913) discloses a method of correcting a photomask pattern with respect to the optical proximity effect.
In the above publication, an improvement is made at a semiconductor element level; however, in practice, correction at a chip level, i.e., a large area of approximately several tens of millimeters square is required.
An example of correction with respect to the optical proximity effect at a chip level or at a block level is suggested in the following publication: S. Miyama, K. Yamamoto, et al., “Large area optical proximity correction with a combination of rule-based and simulation-based methods”, Jpn. J. Appl. Phys. Vol. 35 (1996/12) pp. 6370-6373.
The following describes the correction steps with respect to the proximity effect carried out in the above conventional example referring to the flowchart of
FIG. 13
, and FIG.
14
and FIG.
15
.
First, a distribution of light intensity in a projected image is determined from a pattern
30
of the photomask shown in
FIG. 14
, and a critical edge (transparent pattern edge or opaque pattern edge of pattern
30
) subject to correction with respect to the optical proximity effect is extracted (S
41
and S
42
). In
FIG. 14
, the hatched portions indicate opaque portions, and the other portions indicate transparent portions (translucent portions) . Also, in
FIG. 14
, the critical edge is indicated by the heavy broken line E.
Then, an appropriate point for determining a correction amount of the critical edge E is set as a correction point, and a 1D (one dimensional) context of the correction point is determined (S
43
). Namely, a binary judgement is performed with respect to a correction point on the arrow C of
FIG. 14
(for example, correction point P indicated by × in
FIG. 14
) so as to determine the 1D context which is bit map data representing the transparent portion and the opaque portion indicated by “0” and “1”, respectively.
Thereafter, it is judged in S
44
whether the 1D context thus determined coincides with any one of 1D contexts prepared beforehand in a correction table of FIG.
15
. If it is judged in S
44
that the 1D contexts coincide, the correction amount is determined referring to the correction table so as to correct portions of the photomask pattern requiring correction to the correction amount thus determined (S
47
). Note that, the broken line in
FIG. 15
indicates the correction point P.
On the other hand, in the case where the determined 1D context does not coincide with any of the 1D contexts in the correction table, a correction amount appropriate for the determined 1D context is calculated by simulation (S
45
), and the determined 1D context and the correction amount determined in S
45
are added to the correction table so as to update the correction table (S
46
). Then, the correction point is replaced with another correction point appropriate for determining the correction amount for the critical edge E (S
43
), and a correction amount is determined referring to the updated correction table so as to correct, in the same way as above, portions of the photomask pattern requiring correction to the correction amount thus determined (S
47
).
The sequence of S
43
through S
47
is repeated until correction is finished with respect to all the correction points of the edge extracted in S
42
(S
48
), and when correction is finished with respect to all the correction points, the process is finished.
However, in the photomask pattern correcting method of the described conventional example, contrast and gradient of light intensity are determined from a distribution of light intensity in a projected image, and a critical edge is determined with respect to target pattern dimensions so as to carry out correction with respect to the optical proximity effect in the photoresist.
That is to say, in the photomask pattern correcting method of the conventional example, it is impossible to carry out, along with the correction with respect to the optical proximity effect, correction with respect to photoresist development and a difference in underlayer level by extracting a critical pattern range associated with receding of edges and pattern deformation of the photoresist generated during development, and line-width shifting, etc., of the photoresist due to the difference in underlayer level.
As a result, the photoresist pattern deviates from a desired pattern due to the receding of edges and pattern deformation of the photoresist, or the line-width shifting. In other words, the photomask pattern correcting method of the conventional example has a problem in that accurate correction cannot be carried out.
Further, in the photomask pattern correcting method of the conventional example, it is required that (1) simulation of a projected light optical image, (2) a calculation by simulation of photoresist exposure and development, and (3) a preparation of a correction table are carried out with respect to all regions requiring correction. As a result, a large amount of measurement data are required to be prepared in advance and an extremely long time is required for calculation, preventing fast correction from being carried out.
SUMMARY O
Mehta Bhavesh
Nixon & Vanderhye P.C.
Sharp Kabushiki Kaisha
LandOfFree
Photomask pattern correcting method and photomask corrected... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Photomask pattern correcting method and photomask corrected..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Photomask pattern correcting method and photomask corrected... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2832719