JP2002313693A - Forming method for mask pattern - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a forming method for a mask pattern which corrects proximity effect by using simple algorithm reducing a computation load. SOLUTION: The intensity computation is performed on a forward scatter term and a backward scatter term of an exposure intensity distribution function, the proximity effect is corrected by varying the size of a mask pattern so that the width of an exposure intensity distribution of specific energy intensity becomes equal to design size, and an area density method is used for the backward scatter intensity computation.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method of forming a mask pattern, and more particularly to a method of forming a mask pattern for correcting a proximity effect for charged particle beam exposure.

[0002]

2. Description of the Related Art In recent years, with the improvement in the degree of integration of semiconductor devices, it has been required to fabricate devices having a design rule of 100 nm or less, making it difficult to achieve resolution by a conventional exposure method using light such as ultraviolet rays. It is getting. Therefore, an exposure method using a charged particle beam having a shorter wavelength than light, in particular, an electron beam (electron beam) has been used.

As an exposure method using an electron beam, there are a point beam exposure method using a finely focused electron beam and a variable shaping exposure method using a rectangular electron beam of an arbitrary size. In a conventional electron beam exposure apparatus, a method of scanning the electron beam using a deflector and sequentially irradiating a target position on a sample substrate to form a pattern, that is, a so-called one-stroke exposure method is generally used. is there.

Although this method has an advantage in that an arbitrary pattern can be generated, it has a fatal disadvantage that throughput is low due to a low exposure speed. Therefore, in order to improve the throughput, a partial batch exposure apparatus for exposing a part of a target pattern through a mask has been developed. This mask is a metal thin film or a silicon substrate in which holes are partially formed in accordance with the shape of a pattern to be formed, and is called a stencil mask.

This partial batch exposure apparatus does not perform batch exposure of the whole chip through a mask, but reduces and projects only a 5 × 5 μm partial area of a repetitive pattern through a mask to change an arbitrary pattern. Exposure is performed by a molding exposure method. For example, an arbitrary pattern other than a memory cell is exposed with a variable shaped beam, while a repetitive pattern such as a memory cell is formed as a stencil mask and arranged in a variable shaped beam generating section, so that the pattern can be processed at high speed. Can be exposed.

When exposing a sample substrate with an electron beam, the influence of the proximity effect caused by the forward scattering of the incident electrons in the resist and the back scattering of the substrate becomes a problem.
That is, a phenomenon occurs in which the resolution line width varies depending on the pattern size and the pattern density due to the influence of the proximity effect. Normally, in an electron beam exposure method in which direct exposure is performed without using a mask, an exposure distribution intensity (EID: Exp) determined in advance by an experiment in order to correct the proximity effect
OSURE Intensity Distribut
ion) Based on the function, the exposure amount of each exposure pattern is calculated by self-alignment calculation, and finally the proximity effect is corrected by correcting the exposure amount for each pattern so that all the patterns obtain the same absorption energy. Has been corrected. In addition, in order to further increase the exposure margin and stably draw a fine pattern, a method of correcting the proximity effect by using a combination of a graphic change and a correction of an exposure amount in which the exposure data dimension of the electron beam is smaller than a design dimension is used. Used.

In recent years, a projection type electron beam exposure apparatus (batch projection exposure apparatus) has been developed for the purpose of further improving the throughput.
In this apparatus, all the patterns of the chip are formed as masks as in the case of the light exposure, and the patterns can be collectively exposed over a large area through the mask. This allows
A dramatic improvement in throughput can be expected.

[0008] The projection type electron beam exposure apparatus is a method in which a larger number of patterns are exposed collectively through a mask over a wide area. Therefore, the proximity effect is corrected by changing the exposure amount for each pattern in the same mask. It is difficult. Therefore, a method called the ghost method for correcting the proximity effect has been proposed. In this method, first, a pattern to be drawn is exposed using an electron beam that is normally focused on an exposed portion of the drawn pattern, and then the non-exposed portion, which is an inverted portion of the exposed portion, is defocused by the electron beam. By subjecting the portion to auxiliary exposure, the influence of the proximity effect is reduced.

[0009]

However, in this method, it is necessary to produce a mask of an inverted pattern of the drawing pattern in addition to a regular mask having the drawing pattern, so that the cost is increased. The throughput is reduced. Furthermore, this method has a problem that the exposure margin is reduced.

In a projection type electron beam exposure apparatus, since all the patterns of the chip are exposed through a mask, a method of correcting (changing a figure) the pattern of the mask is an effective method for correcting the proximity effect. is there. However, conventionally, there is no simple algorithm for correcting all the patterns on the mask by changing the figure without correcting the exposure amount, and there is a problem that an enormous calculation time is required.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and uses a simple algorithm that reduces the calculation load so that the size of the finally completed pattern is equal to the design size. It is an object of the present invention to provide a method of creating a mask pattern capable of performing the above-mentioned graphic change.

[0012]

In order to solve the above-mentioned problems, the present invention relates to a method of forming a mask pattern, calculates the intensity of a forward scattering term and a back scattering term in an exposure intensity distribution function, and obtains a predetermined energy intensity. The proximity effect correction is performed by changing the dimension of the mask pattern so that the width of the exposure intensity distribution matches the design dimension.

According to the present invention, first, a predetermined energy intensity of the exposure intensity distribution is set. For example, a forward scatter intensity distribution calculated based on forward scatter based on the width of a certain reference pattern is calculated, and the exposure energy intensity at which the width becomes the same width as the design dimension is calculated and defined as a predetermined energy intensity. . After that, the exposure intensity distribution is calculated based on the intensity calculation of the forward scattering term and the back scattering term in the exposure intensity distribution function for each pattern, and the mask is set so that the width of the exposure intensity distribution at a predetermined energy intensity matches the design dimension. Change the dimensions of the pattern.

In this manner, since the calculation is performed not only for the backscattering but also for the spread of the absorbed energy distribution due to the forward scattering, the accuracy of the proximity effect correction can be improved. As a result, when exposing all chip patterns to a resist film through a mask using a projection type electron beam exposure apparatus, the dimensions of the mask pattern are adjusted so that the pattern dimensions do not deviate from the design dimensions due to the proximity effect. Since the correction is performed, a resist pattern with small dimensional fluctuation is formed.

In the above-described method for forming a mask pattern for charged particle beam exposure, the calculation of the intensity of the backscattering is performed based on an area density method. Backscattering of charged particles, for example, when the acceleration voltage of electrons is 100 KV, affects a range of a radius of about 30 μm from the reference point, and a huge number of patterns exist in this range. It is not realistic to capture all of these effects for each pattern.

In the area density method according to the present invention, the mask pattern is divided into correction meshes, the area density of the pattern of each correction mesh is obtained, and the area of, for example, about twice the backscattering radius is calculated with respect to the correction mesh focused on. It is based on summing all the contributions of the scatter, ie smoothing. This makes it possible to easily and accurately calculate the effective backscattering intensity.

In the above-described method of forming a mask pattern for charged particle beam exposure, it is preferable that the calculation of the area density method is repeatedly performed based on the changed dimensions of the mask pattern. For example, when the backscattering intensity is obtained by one calculation using the above-described area density method, the area density and the smoothing are calculated using the dimensions of the mask pattern before correcting the proximity effect. Will be included based on the difference in the size of the mask pattern.
However, according to the present invention, the area density is approximated to a value based on the actually used mask dimensions by repeating the calculation of the area density and smoothing by taking in the dimensions of the mask pattern after correction, The accuracy of correction of the proximity effect can be improved.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a flowchart illustrating an algorithm according to a mask pattern creation method according to an embodiment of the present invention, and FIG. 2 is a diagram illustrating the principle of a proximity effect correction method according to the mask pattern creation method according to the embodiment of the present invention. It is.

To change the figure of the mask pattern so that the pattern size actually obtained by exposing the electron beam resist to the electron beam becomes the same as the design size, EID
It is necessary to calculate an exposure intensity profile for each of the mask patterns having a plurality of patterns based on the function, and change the figure based on the calculated exposure intensity profile. In addition,
In the present embodiment, figure change refers to correction of the size of a max pattern.

This EID function is generally expressed as in equation (1).

[0021]

(Equation 1)

That is, it is known that the EID function is represented by the sum of two Gaussian functions. The forward scattering term in the equation (1) represents the energy intensity distribution given to the resist film by the electrons incident on the resist film, while the back scattering term is
The energy intensity distribution given to the resist by electrons scattered by atoms in the substrate below the resist film and returned to the resist film is shown.

Here, β _f and β _b are forward scattering lengths, respectively.
The backscattering length, η, which represents the spread of the forward scattering distribution and the backscattering distribution, is the forward scattering / backscattering ratio, and represents the ratio of backscattering to forward scattering. These values depend on the energy of the charged particles, the thickness of the resist, the material of the substrate, and the like, and are determined by calculation or experiment.
For example, as the acceleration voltage of the electron beam increases, the forward scattering length beta _f decreases backscattering length beta _b increases.

For example, when the acceleration voltage of the charged particles is 100 k
V, when the resist film thickness is 200 nm, the forward scattering length β_fIs
7 nm, backscattering length β_bIs about 30 μm,
Scattering length β _fAnd backscattering length β_bIs very different in length
ing. Thus, β_fΒ compared to_bIs large enough
Uses the area density method to calculate the backscattering intensity.
Is effective. This area density method has high accuracy,
This is because the calculation time is short. That is,
This area density method is described in F. Murai et al., J. Vac. Sci. Techno.
l.Exposure data as described in B10 (6), 3072 (1992)
Data is divided into correction calculation mesh areas, and each correction calculation
The area density of the pattern in the
Smoothing according to the backscattering contribution
This is a method for obtaining an effective area density.

Next, with reference to FIGS. 1 and 2, an algorithm according to the mask pattern creating method of the present embodiment will be described. The algorithm according to the mask pattern creating method of the present embodiment includes four steps. That is, as shown in FIG. 1, first, the reference exposure intensity E _th is determined for the uncorrected data in a first step S10, and then the area density is calculated by an area density method in a second step S20. Then, in a third step S30, the area density map is smoothed, and then a fourth step S4
Corrected data is obtained by calculating the dimension shift amount with 0, and the pattern on the mask can be changed based on this data.

First, in a first step S10, FIG.
As shown in (a), for example, a pattern 10 having a minimum width and an isolated pattern is selected from a plurality of mask patterns (10 to 60), and this pattern 10 is formed with the same beam size as the design size. The intensity at which the width of the beam profile at the time of exposure matches the design size of the pattern 10 is determined as the reference exposure intensity _Eth .

Thereafter, as shown in FIG. 2B, for each of a plurality of mask patterns (10 to 60),
Calculating the exposure energy intensity distribution based on the equation (1),
The above reference exposure intensity E _{th of} this exposure energy intensity distribution
Exposure data size W is determined such that the width at is equal to the design size Wo. Note that the exposure data dimension W is the width of the electron beam, and in a projection type electron beam exposure apparatus, indicates the dimension of the mask pattern after correction.

The exposure data size W can be easily calculated by solving the equation (2) for the exposure data size W.

[0029]

(Equation 2)

Here, Wo is the design dimension, and α 'is the area density after smoothing. Erf in the equation (2) is an error function, and is a function defined by the equation (3).

[0031]

(Equation 3)

The left side of the equation (2) indicates the exposure energy intensity, and the first term is a function obtained by integrating the forward scattering term of the equation (1) with the exposure data dimension W. When the electron beam width is W, Means the forward scattering intensity of On the other hand, the second term α′η is
It represents the effective backscattering intensity obtained by the areal density method, and is obtained by multiplying the effective areal density α ′ by the backscattering ratio η.

In this manner, the exposure data dimension (mask pattern dimension) W is calculated, and the mask pattern can be changed so as not to be affected by the proximity effect. (Example) The method for forming a mask pattern according to the present embodiment will be described in further detail based on the principle of the above-described method for forming a mask pattern.

FIG. 3 is a diagram showing a relationship between an exposure intensity distribution due to forward scattering and a pattern size according to the method of forming a mask pattern according to the embodiment of the present invention, and FIG. 4 is a diagram illustrating the formation of a mask pattern according to the embodiment of the present invention. And FIG. 5 is a diagram illustrating a figure changing method according to a mask pattern creating method according to an embodiment of the present invention. First, it determined as follows the reference exposure intensity E _th in the first step S10.

First, as shown in FIG. 2A, the thinnest and isolated pattern 10 is selected from a plurality of uncorrected patterns (10 to 60) formed on the mask, and I do. Then, as shown in FIG. 3, an energy intensity distribution based on forward scattering of electrons when this reference pattern is exposed with an exposure data dimension (dimension of a pattern on a mask) according to a design dimension Wo is calculated. Exposure intensity Ep equal to design dimension Wo is calculated as reference exposure intensity E
_Set to _th .

Specifically, it is obtained by the equation (4).

[0037]

(Equation 4)

When the reference exposure intensity E _th is determined in this manner, the exposure data dimension W of the other pattern becomes the design dimension Wo.
It becomes thinner, and a decrease in the exposure margin is minimized. When determining the reference exposure intensity E _th , since the reference pattern is not changed in shape, the dimension of the reference pattern may be made smaller than the design dimension in order to increase the exposure margin of the reference pattern. In addition, when the isolated minimum width pattern selected as the reference pattern is extremely thin compared to the target design dimension and is also thin compared to the forward scattering length, the reference exposure intensity _Eth becomes extremely small. Would. In such a case, it becomes difficult to correct the dense pattern. Therefore, the reference exposure intensity _Eth is determined by using the pattern of the target design dimension as the reference pattern without using the isolated minimum width pattern as the reference pattern. You may.

Next, in the second step and the third step, the effective backscattering intensity is calculated using the area density method. As shown in FIG. 4, the surface on which the pattern to be exposed is arranged is divided into correction meshes of dimensions A × A. The backscattering intensity distribution has a large spread of 30 μm in an electron beam exposure apparatus with an acceleration voltage of 100 kV, for example. Therefore, the correction mesh (i,
The pattern at j) is (iA, mA) apart (i
The backscattering is also affected by the pattern in the (+1, j + m) -th mesh. How much is affected
The pattern area ratio α ( _{i + 1, j + m} ) in the (i + 1, j + m) th correction mesh (hereinafter, referred to as area density), the correction mesh (i, j), and the correction mesh (i +
1, j + m) and a distance coefficient a (l, m).

The distance coefficient a (l, m) is obtained by the equation (5).

[0041]

(Equation 5)

Here, the (i + 1, j + m) -th correction method
Area density in the_{i + l, j + m})
Backscatter from (i + 1, j + m) th correction mesh
Is a (l, m) × α (_{i + l, j + m}). (I,
the effective backscattering intensity in the j) th correction mesh
α ′ (i, j) is calculated from the (i, j) th correction mesh
Multiple complements whose radius is within about twice the backscattering length
This is the sum of all contributions from the positive mesh. this
Area density where the step of addition is the third step S30
Which can be calculated according to equation (6).
Wear.

[0043]

(Equation 6)

When the backscattering intensity distribution is obtained without using the area density method, the backscattering of the electron beam affects a range of, for example, a radius of 30 μm from the reference point.
Since there are a huge number of patterns in this range, it is necessary to incorporate all of these effects into each pattern. However, such a calculation requires a huge amount of calculation time, and is not practical.

The area density method according to the mask pattern forming method of the present embodiment focuses on the fact that the range of the spatially varying length of the forward scattering intensity and the back scattering intensity of the electron beam is significantly different. Therefore, the following two are assumed. (1) The forward scatter and the back scatter can be calculated separately. (2) When the correction mesh size is sufficiently smaller than the backscattering length, the correction mesh (i + 1, j + m)
The backscattering intensity exerted on the periphery by the pattern inside is (i +
It can be approximated to the backscattering intensity when there is one pattern having an area equal to the sum of the pattern areas in the mesh at the center of the (l, j + m) -th correction mesh.

Accordingly, the mask pattern is divided into correction meshes, and the area density of the pattern of each correction mesh is obtained.
This is performed on the basis of adding all the contributions of the backscatter to the focused correction mesh, that is, smoothing the region, for example, about twice as large as the backscatter radius. This makes it possible to easily and accurately calculate the effective backscattering intensity.

The graphic change of the mask pattern performed in the fourth step S40 can be obtained by solving the above equation (2) for the exposure data dimension (dimension of the mask pattern) W after the graphic change. That is, at this point,
The forward scattering length β _f , the back scattering length β _b , and the design dimension W in the equation (2)
Since o, the area density α ′ after smoothing and the backscattering ratio η are known, the exposure data dimension (dimension of the mask pattern) W of the equation (2) can be calculated.

The pattern change of the mask pattern is basically performed for each pattern. The size of one mesh is, for example, about 3.0 × 3.0 μ which is about 1/10 of the backscattering length.
m, and one case is assumed to be included in this mesh, and one case is assumed to be arranged over a plurality of correction meshes. When one pattern is arranged over a plurality of correction meshes, since the effective backscattering intensity differs for each correction mesh, it is necessary to find the figure change amount of the mask pattern required for each correction mesh.

That is, in the case of a pattern extending over a plurality of correction meshes, the pattern is once divided by the plurality of correction meshes, and the figure change amount of the mask pattern is obtained for each correction mesh. However, when the figure change amount is obtained for each correction mesh in this way, as shown in FIG.
A pattern 70a divided by a plurality of meshes as shown in FIG. In other words, a large number of thin beams are generated in the pattern, making it difficult to create a mask pattern.

Therefore, with respect to the pattern which is originally one, the figure is not changed by the contact recognition at the boundary between the respective correction meshes. By performing this contact recognition, as shown in FIG. 5C, the figure is changed to a pattern 70b in which no beam is generated in the pattern. This facilitates creation of a mask pattern.

Here, since the corrected exposure data dimension (dimension of the mask pattern) is smaller than the design dimension due to the figure change, the area density of the mask pattern used for actual exposure is not calculated in the figure change calculation. It is smaller than the area density used. In other words, for example, the exposure data size calculated by one calculation is not calculated based on the area density based on the dimension of the mask pattern after correction, and therefore includes an error based on the deviation of the area density. Will be.

Therefore, when high accuracy is required,
It is preferable to repeatedly perform the area density calculation, the smoothing process, and the calculation of the figure change amount based on the corrected exposure data size after the figure change. Thereby, the figure change amount is calculated with the area density approximated to the area density based on the actual exposure data size, so that the proximity effect can be corrected with high accuracy. It goes without saying that as the number of times of repeating this calculation is increased, the accuracy of the proximity effect correction is improved.

As described above, the mask creation method according to the present embodiment can simplify the algorithm of the proximity effect correction, so that the calculation load can be reduced. Further, since the calculation is performed including the spread of the absorbed energy distribution due to the forward scattering, the accuracy of the proximity effect correction can be improved. Therefore, when forming a pattern by exposing an electron beam to a resist film through a mask, it becomes possible to correct the proximity effect,
Variations in pattern dimensions can be suppressed.

(Results of Calculation by the Inventor of the Present Invention) The inventor of the present application obtained the exposure data dimension (mask pattern dimension) W by using the mask pattern forming method of the present embodiment. Variations in pattern dimensions due to the proximity effect mainly occur when electrons pass through the resist film and enter the substrate, and some of the backscattered electrons in the substrate return to the resist film and expose the resist film. to cause.

Therefore, the proximity effect is, for example, a line & space (L & S: Line &Space; hereinafter, L & S).
) Becomes remarkable in a dense pattern region formed continuously. In addition, the proximity effect depends on the distance from the end of the L & S, that is, the area that is not exposed, in the area of the dense pattern composed of the L & S. From this, attention was paid to the dependence of the post-shift exposure data dimension (mask pattern dimension after correction) W on the distance from the end of the L & S.

FIG. 6A shows the distance d from the end of the L & S.
FIG. 6B shows the post-shift exposure data dimension W with respect to FIG.
FIG. 6 is a graph showing the distance d from the end of the L & S and the post-shift exposure data dimension W based on the data of FIG. As a mask pattern for correcting the proximity effect, a 1: 1 L <"
S was assumed. In this case, the area density becomes 50%. As the calculation conditions, the forward scattering length β _f was 40 nm, the back scattering length β _b was 30 μm, the back scattering ratio η was 0.6, and the size of the correction mesh was 3 μm × 3 μm.

Next, a description will be given of the result of changing the dimensions of the mask pattern using the algorithm according to the mask pattern creation method of the present embodiment under the above-described calculation conditions. In FIG. 6B, the exposure data dimension obtained as a result of performing the proximity effect correction processing is the post-shift exposure data dimension, and the resist dimension when exposure is performed using this corrected mask pattern. Is a simulation result showing the result calculated by simulation.

As shown in FIGS. 6A and 6B, L &
It has been found that as the distance d from the terminal end of S increases, the shift dimension with respect to the finished size gradually increases, and when the distance d exceeds 36 μm, the shift dimension is saturated. For example, in the case of a line at a position where the distance d from the end of the L & S is 10.08 μm, the post-shift exposure data dimension is set to 4 for the completed dimension of 60 nm.
For a line at a position where the distance d is set to 0.8 nm and the distance d is set to 20.04 μm, it means that the proximity effect can be corrected by setting the exposure data dimension to 38 nm. In a region where the distance d from the end of the L & S is 36 μm or more, the post-shift exposure data dimension is 35.6 nm, and it is found that the shift dimension of the mask pattern becomes the same.

As described above, in the projection type electron beam exposure apparatus, for example, the design dimension is 6
It has been found that a mask pattern may be created by shifting the design dimension as shown in FIG. 6 in order to suppress the dimensional variation due to the proximity effect of the wiring of 1: 1 L & S of 0 nm. As described above, the details of the present invention have been described with the embodiments. However, the scope of the present invention is not limited to the examples specifically shown in the embodiments, and the scope of the present invention is not limited to the scope of the present invention. Modifications of the embodiment are included in the scope of the present invention.

For example, a dense pattern in which L & S are arranged as a pattern easily affected by the proximity effect has been described as an example. However, the present invention is not limited to this, and may be applied to a hole pattern or the like. (Supplementary Note 1) Calculate the intensity of the forward scattering term and the backscattering term in the exposure intensity distribution function, and change the dimension of the mask pattern so that the width of the exposure intensity distribution at a predetermined energy intensity matches the design dimension. A method for creating a mask pattern, comprising performing correction. (1) (Supplementary note 2) The method according to claim 1, wherein the calculation of the intensity of the backscattering is performed based on an area density method. (2) (Supplementary note 3) The method according to claim 2, wherein the calculation of the area density method is repeatedly performed based on the changed dimensions of the mask pattern.
(3) (Supplementary note 4) The supplementary note 1, wherein the predetermined energy intensity is an energy intensity at which a dimension of an isolated pattern having a minimum width among the mask patterns can be as designed. How to create a mask pattern. (4) (Supplementary note 5) The mask pattern creating method according to Supplementary note 1, wherein the mask pattern dimension is corrected for each area divided into the correction mesh. (5) (Supplementary note 6) The mask pattern creating method according to claim 5, wherein the correction of the mask pattern dimension is not performed on a side adjacent to the pattern at the boundary of the correction mesh.

[0061]

As described above, according to the present invention,
The exposure intensity distribution is calculated based on the intensity calculation of the forward scattering term and the back scattering term in the exposure intensity distribution function, and the dimension of the mask pattern is adjusted so that the width of the exposure intensity distribution at a predetermined energy intensity matches the design dimension. The method is characterized in that the proximity effect correction is performed by changing, and the area density method is used for calculating the backscatter.

By doing so, the algorithm of the proximity effect correction can be simplified, and the calculation load can be reduced. Further, since the calculation is performed including the spread of the absorbed energy distribution due to the forward scattering, the accuracy of the proximity effect correction can be improved.

[Brief description of the drawings]

FIG. 1 is a flowchart illustrating an algorithm according to a mask pattern creating method according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating the principle of a proximity effect correction method according to the mask pattern creation method according to the embodiment of the present invention;

FIG. 3 is a diagram showing a relationship between an exposure intensity distribution due to forward scattering and a pattern size according to the mask pattern creating method according to the embodiment of the present invention.

FIG. 4 is a diagram illustrating an area density method according to the embodiment of the present invention for creating a mask pattern.

FIG. 5 is a diagram for explaining a figure changing method according to a method of creating a mask pattern according to the embodiment of the present invention.

6A shows the post-shift exposure data size with respect to the distance from the end of the L & S, and FIG. 6B shows FIG.
FIG. 7 is a graph of the distance from the end of the L & S and the post-shift exposure data size based on the data of FIG.

[Explanation of symbols]

 10, 20, 30, 40, 50, 60: Pattern W: Exposure data dimension (dimension of mask pattern) Wo: Design dimension 70: Pattern before correction 70b: Pattern after correction without contact recognition 70c: Pattern after correction in case of contact recognition d: Distance from the end of L & S

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2H095 BA01 BA08 BB01 5B046 AA08 BA06 5F056 AA22 CA30 CC11 FA05

Claims (5)

[Claims] An intensity calculation of a forward scattering term and a back scattering term in an exposure intensity distribution function is performed, and a dimension of a mask pattern is changed so that a width of an exposure intensity distribution at a predetermined energy intensity matches a design dimension. A method of creating a mask pattern, wherein effect correction is performed. 2. The method according to claim 1, wherein the calculation of the intensity of the backscattering is performed based on an area density method. 3. The method according to claim 2, wherein the calculation of the area density method is repeatedly performed based on the changed dimensions of the mask pattern. 4. The energy intensity according to claim 1, wherein the predetermined energy intensity is an energy intensity of a minimum width of the mask pattern, and the size of an isolated pattern can be as designed. How to create a mask pattern. 5. The method according to claim 1, wherein the correction of the mask pattern dimension is performed for each area divided into a correction mesh.