Radiation imagery chemistry: process – composition – or product th – Including control feature responsive to a test or measurement
Reexamination Certificate
2002-04-11
2004-01-13
Young, Christopher G. (Department: 1756)
Radiation imagery chemistry: process, composition, or product th
Including control feature responsive to a test or measurement
C430S005000, C430S296000, C430S942000
Reexamination Certificate
active
06677089
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to a charged particle beam exposure method, more particularly, to a charged particle beam exposure method performing a proximity effect correction in preparation of exposure data for a charged particle beam exposure apparatus in order to improve size accuracy of a transferred pattern, and a method for converting a rectangular pattern data of a charged particle beam exposure mask pattern to a lattice pattern data thereof in order to reduce an exposure dose for a relatively large pattern and a charged particle beam exposure method using the conversion method.
2. Description of the Related Art
In a case where a resist film on a substrate is irradiated with a charged particle beam, for example an electron beam, to draw a circuit pattern thereon, an electron beam incident on the resist film is partly scattered forward and the electron beam transmitted through the resist film is partly scattered backward to again impinge on the resist film. For this reason, even if an electron beam impinges at one point on the resist film, an influence thereof spreads around, causing a so-called proximity effect.
An energy intensity distribution (EID) function f (X, Y) on a resist film when an electron beam impinges at a point of X=0 and Y=0 on the resist film is expressed by the following equation in which a forward scattering term and a backscattering term approximate respective Gaussian functions:
f
(
X
,
Y
)
=
1
π
(
1
+
η
)
{
1
β
f
2
exp
(
-
X
2
+
Y
2
β
f
2
)
+
η
β
b
2
exp
(
-
X
2
+
Y
2
β
b
2
)
}
(
1
)
wherein &eegr; denotes a backscattering coefficient, &bgr;
f
denotes a forward scattering radius and &bgr;
b
denotes a backscattering radius. The values are dependent on energy of an electron beam, a thickness of a resist film, material of a substrate and others, each being determined by an experiment. With increase in acceleration voltage of the electron beam, &bgr;
f
decreases and &bgr;
b
increases.
In a prior art proximity effect correction method, fixed sample points were set at the middle points of sides or corners of each pattern to be exposed and an exposure intensity at each of the fixed sample points when pattern was exposed was calculated using the equation (1) and an exposure dose was determined such that the sum of the squares of differences each between an exposure intensity and a corresponding target value over all the fixed sample points is minimized.
However, in company with a progress in integration to a high degree of LSI, a rapid increase has occurred in the number of patterns, resulting in an excessively extended calculation time.
Hence, there has been a desire for a proximity effect correction method capable of reducing the calculation time and confining size error of a developed pattern (transferred pattern) within an allowable range.
As one of such methods, there has been disclosed a method in, for example, JP No. 2502418 and Journal of Vacuum Science Technology, Vol. B10, No. 6, pp. 3072-3076, 1992; in which method a layout plane of an LSI exposure pattern is divided into squares in mesh and a pattern area density is calculated for each of the squares to thereby obtain an approximate value of a scattering exposure intensity of a square of interest in consideration of influences of peripheral squares on the square of interest on the basis of the backscattering term of the equation (1). In this method, an exposure dose is determined such that the sum of the half value of a peak of a forward scattering intensity and a backscattering intensity is constant.
According to the method, using a simple and easy-to-use algorithm, it is possible to prevent a global variation in size of a transferred pattern caused by an influence of backscattering.
However, since no consideration is given to a spread of absorbed energy distribution due to forward scattering, it is not guaranteed that a size of a transferred pattern is equal to a design width. That is, as pattern elements become finer, a spread of an absorbed energy distribution at a half value intensity cannot be neglected, thereby making a size of a transferred pattern larger than a design width due to forward scattering.
Therefore, a proximity effect correction method as described below is proposed in JP 11-151330 A.
(A) A pattern width is adjusted such that a half-width of a forward scattering intensity distribution, determined by surface integration of the forward scattering term of the energy intensity distribution function over a pattern of interest, is equal to a design width and a forward scattering intensity &egr;
p
giving the half-width is determined as a reference forward scattering intensity;
(B) an exposure intensity &agr;
p
·&eegr; due to the backscattering term of the energy intensity distribution function is determined using a pattern area density map method; and
(C) a corrected exposure dose Qcp is determined such that Qcp times (&egr;
p
+&agr;
p
·&eegr;) is equal to a threshold value Eth of pattern developing.
For example, when, as shown by dotted lines in FIG.
51
(A), design widths in the X direction of a large width pattern and a small width pattern are (X
2
−X
1
) and (X
4
−X
3
), respectively, the pattern widths are narrowed as shown by solid lines in FIG.
51
(A) in the processing of the above step (A). The large width pattern has &egr;
p
=½ and &agr;
p
=1 and a pattern having such values is referred to as a reference pattern. If corrected exposure doses Qcp of the large width pattern and the small width pattern are expressed by Q
1
and Q
2
, respectively, the following equation holds in the step (C):
(½+&eegr;)Q
1
=(&egr;
p
+&agr;
p
·&eegr;)Q
2
where Q
1
>Q
2
.
When the rectangular regions
13
and
14
shown by dotted lines, surrounding the rectangular transmission holes
11
and
12
shown by solid lines on the mask
10
, are irradiated with an electron beam at respective exposure doses Q
1
and Q
2
, an exposure intensity distribution on a wafer coated with a photoresist is determined as shown in FIG.
51
(B).
According to this method, since a slant at the threshold value Eth of an exposure intensity distribution of each of the patterns is sharp, a variation in pattern width relative to a variation in exposure condition decreases, enabling to achieve a high accuracy pattern. Further, a corrected exposure dose can be determined in relatively short time. However, since the above method obtains a corrected exposure dose Qcp for each pattern, it cannot be applied in a case where a block exposure pattern in a small region of, for example, 4.5×4.5 &mgr;m
2
on a stencil mask is selected to collectively expose.
Therefore, in JP 12-166465 A, the minimum value of corrected exposure doses for respective patterns in a block exposure pattern is determined as a corrected exposure dose Qcp for the block exposure pattern, and then auxiliary exposure is applied to exposure intensity-short regions in the block.
For example, when the rectangular transmission holes
11
and
12
of FIG.
51
(A) are irradiated with an electron beam at an exposure dose Q
1
, covering a rectangular region
15
shown by a dotted line in FIG.
52
(A), an exposure intensity distribution on a wafer coated with a photoresist is determined as shown in FIG.
52
(B). In this state, the narrow line pattern cannot be developed due to shortage of exposure. Then, a rectangular transmission hole not shown for auxiliary exposure (ghost exposure) is irradiated with an electron beam to determine an exposure intensity distribution shown by an alternate long and short dash line in
FIG. 53
, thereby enabling developing of the narrow line pattern having a width (X
4
−X
3
).
As is apparent by comparison of
FIG. 53
with FIG.
51
(B), however, since a slant of an exposure intensity distribution at a pattern edge of a narrow line pattern is gentle, a variation in width of a transferred pattern image re
Ogino Kozo
Osawa Morimi
Arent Fox Kintner & Plotkin & Kahn, PLLC
Fujitsu Limited
Young Christopher G.
LandOfFree
Rectangle/lattice data conversion method for charged... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Rectangle/lattice data conversion method for charged..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Rectangle/lattice data conversion method for charged... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3228867