JP2004078115A - Pattern correction method - Google Patents

Abstract

A pattern is corrected in accordance with a design value of the pattern, and a calculation time required for calculating a correction amount of the pattern is reduced.
The present invention relates to a line width of a correction target pattern to be corrected, an inter-pattern distance between an adjacent pattern adjacent in the width direction of the correction target pattern and the correction target pattern, The correction value is set based on the edge to be corrected and the distance between the edges to the far side edge of the adjacent pattern adjacent in the edge direction opposite to the edge to be corrected. By setting the correction value by referring to the distance between edges as a parameter, it is possible to perform pattern correction in consideration of the influence of a pattern formed around the pattern to be corrected.
[Selection diagram] FIG.

Description

【００１９】
また、補正テーブルを構成するパラメータについて図３を参照しながら定義すると、パターン間距離Ｗ１は、補正対象エッジ３１ａに対して、パターン３１の幅方向であってエッジ３１ａの外側に向かう方向であるエッジ方向に隣接するパターンとの距離であり、線幅Ｗ２は、補正対象となるパターン３１の線幅である。但し、本例のように補正対象パターン３１が、一次元的に配列されたパターンのうち最も端に形成されている場合は、パターン間距離Ｗ１は無限遠とされる。補正対象パターン３１の補正対象エッジ３１ａの補正量に殆ど寄与しない程度にパターン間距離Ｗ１が大きい場合にもパターン間距離Ｗ１を無限遠とすることができる。また、補正対象エッジ３１ａに対して反対側のエッジ方向に隣接するパターンのエッジのうち補正対象エッジ３１ａから遠い側のエッジ３２ａと、補正対象エッジ３１ａとの距離をエッジ間距離Ｗ３とする。
【００２０】
さらに、補正対象パターン３１に対して補正対象エッジ３１ａと反対側に形成されるパターンの配置を考慮して、パターン間距離Ｗ１と補正対象パターンの線幅Ｗ２とをパラメータとし、補正対象エッジに対する補正量を修正した補正量から構成される補正テーブルを作成しておく。具体的には、エッジ間距離Ｗ３をパラメータとし、パターン間距離Ｗ１と線幅Ｗ２とにより設定される補正量がエッジ間距離Ｗ３ごとに修正された補正量を作成することになる。エッジ間距離Ｗ３ごとに作成された補正テーブルに基づいて補正対象エッジ３１ａを補正することにより、補正対象パターン３１の周囲のパターン密度の状況が考慮された補正量に基づいたパターン補正を行うことができる。
【００２１】
続いて、上述のように予め作成された補正テーブルに基づいて補正量を選択するためのパラメータを取得手順について説明する。補正対象エッジから線幅に含まれるエッジを抽出する（Ｓ２）と、同一線幅Ｗ２を有する各パターンのエッジを全て補正対象エッジとして抽出することができる。
【００２２】
次に、パターン間距離Ｗ１を抽出する（Ｓ３）。パターン間距離Ｗ１は、マスクに形成されたパターンのうち隣り合うパターンの間の距離であり、抽出されたエッジそれぞれをパターン間距離Ｗ１に応じて分類する。例えば、図４に示すように、孤立パターン２１に対しては左右両側に隣り合うパターンが形成されていないことから、孤立パターン２１のエッジ２１ａ、２１ｂから隣接パターンまでの距離であるパターン間距離Ｗ１を無限遠として分類する。また、図５に示すように、一次元的に配置されたパターンのうち最も端に形成されたパターン３１の右側の補正対象エッジ３１ａから隣接するパターンまでの距離であるパターン間距離Ｗ１を無限遠として分類し、それ以外のパターン間距離を全て等しいものとして分類する。ここで、パターン間距離Ｗ１以外のパターン間距離は、隣接するパターン間の距離である最小間隔とされる。
【００２３】
次に、上述の補正対象パターンの線幅Ｗ２以上のエッジ間距離Ｗ３を有するエッジを抽出する（Ｓ４）。最小間隔レンジに分類されたエッジを抽出した場合と異なり、エッジに対して陰線処理を行わない。通常、パターン間の領域についてはパターンの形状に関するデータが存在せず、この領域を含む幅を線幅と認識させないよう処理されるが、パターンの線幅を形成する各パターンのエッジを抽出した場合と異なり、隣接するパターンのパターン間距離を含み、異なるパターンのそれぞれのエッジ同士の距離を線幅として定義することにより、パターン間の領域を含む線幅としてのエッジ間距離Ｗ３を形成するエッジを抽出する。
【００２４】
ここで、陰線処理とは一次元的に配列されたパターンの線幅を抽出するに際してのエッジの選択方法に関する処理である。例えば、抽出対象とされる一のエッジに対し、エッジをインエッジ及びアウトエッジの如き２種類のエッジとして定義する。インエッジは、その法線ベクトルがパターンの図形内に向くエッジとされ、アウトエッジはその法線ベクトルがパターンの図形外に向くエッジとされる。つまり、パターンが矩形状である場合には、線幅Ｗ２はパターンの長手方向に沿ったエッジのそれぞれのインエッジ間の距離として定義されることになる。
【００２５】
また、マスクに形成された別のパターンのインエッジとの間の距離がエッジ間距離Ｗ３とされ、一のパターンの線幅Ｗ２は、インエッジとインエッジとの間の距離のうち一のパターンが有するインエッジ間の距離と定義されることになる。
【００２６】
例えば、孤立したパターンの場合にインエッジとインエッジとの距離はそのパターンの線幅Ｗ２に相当するが、一次元的にパターンが配列された場合には、異なるパターン間のインエッジとインエッジとの間の距離であるエッジ間距離Ｗ３は、これらパターンの線幅Ｗ２より大きい寸法となる。よって、線幅Ｗ１、パターン間距離Ｗ２及びエッジ間距離Ｗ３の３つのパラメータを設定する場合には、一のエッジをインエッジ及びアウトエッジの２種類で定義することになる。
【００２７】
さらに図６及び図７に示すように、エッジ間距離Ｗ３がパターン３１の線幅Ｗ２より大きいという条件を付けることにより、図６に示す孤立パターン２１の線幅Ｗ２のエッジ２１ａ、２１ｂは補正の対象外とされる。このとき、線幅Ｗ２以上の間隔を形成するエッジ間距離Ｗ３の寸法は、少なくとも、「補正対象パターンの線幅Ｗ２＋パターン間距離Ｗ１＋補正対象パターンの線幅Ｗ２」の寸法以上とする。ここで、パターン間距離Ｗ１の寸法は、隣接するパターンの間に関するパターン間距離である。従って、補正対象パターン線幅Ｗ２以上のエッジ間距離を形成するエッジを抽出することにより、図７に示すように右から２番目のエッジ３１ｂ以外の全てのエッジを抽出することになる。
【００２８】
次に、線幅Ｗ２以上のエッジ間距離Ｗ３を有するエッジ間の領域どうしが重複するか否かを判定する（Ｓ５）。重複しない場合には、パターンの線幅と、隣接するパターン間距離とにより決められる補正量だけ補正対象エッジを補正する（Ｓ６）。このとき補正値を取得するために参照される補正テーブルは、パターンの線幅と、隣接するパターン間の距離との組み合わせのみにより作成された補正テーブルである。
【００２９】
重複する場合には、補正対象パターン３１に対して補正対象エッジ３１ａと反対側の領域に他のパターンが形成されていることになり、補正対象パターン３１の片側に他のパターンが形成された片側密の状態であることを確認することができるとともに、補正対象エッジ３１ａに対して反対方向に他のパターンが配置された密状態であることが示される。従って、パターンの設計値に対して、設計値とのずれが生じる原因となるパターンが周囲に存在することを確認することができ、これら周囲に配置されたパターンの光学的或いはプロセス的な影響に応じた補正量が考慮された補正テーブルに基づいて、補正対象パターンのエッジ、つまり線幅を補正する（Ｓ７）。エッジ間の領域が重複する補正対象エッジに対しては、エッジ間距離ごとに作成され、パターンの線幅と隣接するパターン間のパターン間距離とによりマトリクス状に設定される片側密状態用の補正量に基づいて補正対象エッジを補正する。このとき、補正量を設定するに際しては、線幅と隣接パターン間の間隔とで決まる補正量からなる補正テーブルと、重複する領域に基づいて補正量が決まる補正テーブルとを参照して補正対象エッジ３１ａを補正する。最後に、これら補正量に関するデータは補正処理を行う処理装置と装備される記憶装置に保存される（Ｓ８）。
【００３０】
以上を説明したパターン補正を全てのパターンの線幅について行うことによりマスクに形成されたパターンの全てを補正することができる。また、線幅、パターン間距離及びエッジ間距離をパラメータとして補正量を設定することにより、マスクに形成された全てのパターンについて形状などのデータを取得することなく、簡便に補正量を設定することができ、正確且つ迅速にパターンを補正することが可能となる。
【００３１】
【発明の効果】
補正対象エッジのエッジ方向のパターン間距離、パターンの線幅、及びエッジ反対方向の隣接パターンの遠い側のエッジとの距離であるエッジ間距離とをパラメータとして補正対象となる補正対象パターンのエッジの補正量を設定することにより、光学的或いはプロセス的に補正対象パターンの補正量にずれを生じさせるパターンの影響を考慮した補正を行うことができ、高い精度で補正を行うことができる。特に、主に補正対象エッジとその反対側エッジにおいて補正対象パターンを除いたほかのパターンの密度差が大きい場合に有効である。
【００３２】
さらに、補正対象エッジの反対方向に隣接するパターンの反対側エッジとの距離とすることにより汎用ＤＲＣ（Ｄｅｓｉｇｎ　Ｒｕｌｅ　Ｃｈｅｃｋ）ツールによるインプリメントが可能となり、簡便且つ高速なパターン補正を行うことが可能となる。
【図面の簡単な説明】
【図１】本発明のパターン補正方法を示すフローチャートである。
【図２】孤立パターンが形成されたマスクの一例を示す平面図である。
【図３】一次元的に配置されたパターンが形成されたマスクの一例を示す平面図である。
【図４】孤立パターンが形成されたマスクの一例を示す平面図である。
【図５】一次元的に配置されたパターンが形成されたマスクの一例を示す平面図である。
【図６】孤立パターンが形成されたマスクの一例を示す平面図である。
【図７】一次元的に配置されたパターンが形成されたマスクの一例を示す平面図である。
【図８】従来のパターン補正方法により補正される孤立したパターンが形成されたマスクを示す平面図である。
【図９】従来のパターン補正方法により補正される一次元的に配置されたパターンが形成されたマスクを示す平面図である。
【符号の説明】
２０，マスク、　２１ａ，エッジ、　２１，３１　パターン、　３１ａ，補正対象エッジ[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a pattern correction method capable of correcting a pattern formed on a mask used in a manufacturing process of a semiconductor device or the like according to a design value of a line width after transfer.
[0002]
[Prior art]
With the progress of high integration of LSIs and miniaturization of element sizes to be built in the LSIs, fidelity of pattern transfer in a lithography process has become a problem. Specifically, phenomena such as rounding of a corner which should be 90 degrees in design and widening or narrowing of a line width occur (optical proximity effect). If the absolute value of the permissible dimensional error decreases as the pattern becomes finer, the permissible dimensional error may exceed the allowable dimensional error due to the optical proximity effect.
[0003]
The optical proximity effect originally meant an effect due to light at the time of transfer, but recently, in addition to an optical effect, it has been used including various effects that occur throughout the entire wafer process such as a resist phenomenon and etching.
[0004]
The causes of the optical proximity effect include optical factors in exposure (interference of transmitted light between adjacent patterns), resist process (bake temperature, time, development time, etc.), reflection and unevenness of the substrate, influence of etching, and the like. No. Specifically, in transfer, effects caused by optical factors such as light diffraction and interference, the pattern dependence of the resist dissolution rate during resist development, the microloading effect when etching the resist, and the etching rate Various causes such as the effect of the pattern dependency can be considered.
[0005]
When a mask used in the manufacture of a semiconductor device is created, a correction process involving a change in pattern shape typified by optical proximity effect correction is often performed for the purpose of wafer manufacturing margin and the like. For example, in the case of rule-based one-dimensional optical proximity effect correction (hereinafter referred to as OPC), the line width of each of the line widths of the pattern to be corrected is corrected by the relationship between the line width to the adjacent pattern. Determine the amount. The line width correction OPC that determines the correction amount of the edge of the line width based on two parameters, that is, the line width of the pattern and the distance between the patterns to the adjacent patterns, is general-purpose. Among them, the inter-pattern distance to the adjacent pattern in the left direction for the left edge and the right direction for the right edge is obtained, and correction is performed by shifting each edge by a predetermined correction amount. At this time, for example, the correction amounts A and B for the edges 81a and 81b of the isolated pattern 81 formed on the mask 80 as shown in FIG. 8 and having no pattern formed around the mask 80 and the mask 90 as shown in FIG. The correction amount C for the edge 91a of the one-dimensionally arranged patterns 91 is set to the same value.
[0006]
[Problems to be solved by the invention]
However, when the arrangement of the pattern formed around the correction target edge to be corrected is different, optical causes such as interference of light passing through the pattern formed around the correction target pattern, and the final cause. Due to the process of fixing the pattern on the transfer destination substrate, the correction amounts A and B of the edges 81a and 81b of the isolated pattern 80 shown in FIG. 8 and the plurality of patterns shown in FIG. In some cases, the correction amount C of the edge 91a of the pattern 91 arranged in a different manner may be different, and it is important to set the correction amount of the correction target edge in consideration of the influence of the pattern arranged around. Become.
[0007]
For example, in order to transfer a pattern in accordance with a design value, it is difficult to correct the edges of all the patterns based on only two parameters, the line width of the pattern and the distance between the patterns. There is a need for a technique that can correct the deviation of the correction amount by considering the pattern formed around the pattern and correct the pattern so that the pattern according to the design value is transferred when the pattern is transferred. I have.
[0008]
Further, in order to calculate the correction amount of the correction target edge, for example, according to Japanese Patent Application Laid-Open No. H11-102062, one-dimensional arrangement information of a pattern is extracted, and a shift amount which is a correction value of the correction target edge is calculated. Although a technique for strictly obtaining the correction is disclosed, the accuracy of correction can be improved, but the calculation time required for calculating the correction value is increased, and the design efficiency of the pattern is reduced. The reduction in pattern design efficiency leads to a reduction in manufacturing efficiency in a semiconductor manufacturing process using a mask on which these patterns are formed.
[0009]
Further, according to the technique disclosed in Japanese Patent Application Laid-Open No. 2000-020564, the precision of the area density of a pattern formed in a region centered on the center of a correction target side formed on a mask is calculated. Although the pattern can be corrected well, the calculation time of the area density and the like is increased as in the technique disclosed in Japanese Patent Application Laid-Open No. H11-102062, and the manufacturing efficiency of the semiconductor device is reduced.
[0010]
Accordingly, an object of the present invention is to provide a pattern correction method which can correct a pattern in accordance with a design value of the pattern and which does not increase a calculation time required for calculating a correction amount of the pattern. And
[0011]
[Means for Solving the Problems]
In the pattern correction method according to the present invention, a correction value is set by an optical proximity effect correction to correct a line width of a pattern formed on a mask, and the line correction of a transferred pattern is made closer to a design value. The line width of the target correction target pattern, the inter-pattern distance between an adjacent pattern adjacent in the width direction of the correction target pattern and the correction target pattern, the correction target edge forming the correction target pattern, and the correction The correction value is set based on an inter-edge distance to an edge on a far side of an adjacent pattern adjacent in an edge direction opposite to the target edge. By referring to the distance between edges, which is the distance between the edge of the correction target pattern and the edge on the far side of the adjacent pattern, as a parameter for considering the pattern arranged around the correction target edge, Compared to the case where the correction amount is calculated by individually calculating the data, the calculation time for calculating the correction amount of the correction target edge does not increase, and the correction amount with high accuracy can be calculated. It is possible to correct the pattern so that the pattern according to the value is transferred.
[0012]
In addition, a correction table is created in which the line width of the pattern to be corrected, the distance between the patterns, and the distance between the edges are parameters according to the distance between the edges, and correction is performed based on the correction value obtained from the correction table. By correcting the target edge, it is possible to perform correction in consideration of the influence of the pattern that causes an error in the correction amount of the correction target pattern.
[0013]
In particular, it is possible to make corrections that take into account the difference in the arrangement of the patterns formed around the pattern compared to other patterns, especially for the pattern arranged at the end of the pattern arranged one-dimensionally on the mask It becomes.
[0014]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, the pattern correction method of the present invention will be described with reference to FIG. Note that each step will be described sequentially while comparing an isolated pattern having no pattern formed therearound and a one-dimensionally arranged pattern as shown in FIGS. 2 to 7.
[0015]
FIG. 1 is a flowchart showing a pattern correction method according to the present invention. FIG. 2 is a plan view of a mask 20 having an isolated pattern 21 around which no other pattern is formed. FIG. FIG. 3 is a plan view of masks 30 arranged one-dimensionally. In this example, as shown in FIGS. 2 and 3, the isolated pattern 21 and the one-dimensionally arranged pattern 31 have a rectangular shape, and the pattern 31 shown in FIG. It is formed so as to be arranged one-dimensionally.
[0016]
As shown in FIGS. 2 and 3, the line widths of the isolated rectangular isolated pattern 21 and the pattern in which the rectangular pattern 31 are arranged one-dimensionally are all equal. Further, the isolated pattern 21 is in a sparse state in which no other pattern is formed on both sides of the pattern 21 in the width direction perpendicular to the edge 21a along the longitudinal direction of the pattern 21. In addition, since the rectangular pattern is formed one-dimensionally on one side of the rectangular pattern, the pattern density is reduced in one side of the pattern 31 formed at the end of the one-dimensionally arranged pattern. One side is dense and one side is sparse with no pattern formed on the opposite side.
[0017]
When correcting a pattern, first, data relating to the design value of the pattern is stored in a storage device in advance (S1). The design value of the pattern is data of the pattern to be finally transferred, and does not include an error generated between the pattern to be actually transferred and the design value of the pattern. Further, the data relating to the design value of the pattern includes data relating to a correction table which is referred to when correcting the pattern, and the correction table includes a line width of the pattern to be corrected and a pattern to be corrected. It is composed of a correction value determined by a combination of the line width and the inter-pattern distance, using the inter-pattern distance, which is the distance to an adjacent pattern, as a parameter. For example, in the correction table, correction values determined by a combination of the inter-pattern distance W1 and the line width W2 of the correction target pattern are arranged in a matrix, and are created in a matrix format as shown in Table 1.
[0018]
[Table 1]

[0019]
Further, when the parameters constituting the correction table are defined with reference to FIG. 3, the inter-pattern distance W1 is the width direction of the pattern 31 and the direction toward the outside of the edge 31a with respect to the correction target edge 31a. The line width W2 is the distance between the adjacent patterns in the direction, and the line width W2 is the line width of the pattern 31 to be corrected. However, when the correction target pattern 31 is formed at the end of the one-dimensionally arranged pattern as in this example, the inter-pattern distance W1 is set to infinity. Even when the inter-pattern distance W1 is large enough to hardly contribute to the correction amount of the correction target edge 31a of the correction target pattern 31, the inter-pattern distance W1 can be set to infinity. Further, the distance between the edge 32a farther from the correction target edge 31a and the correction target edge 31a among the edges of the pattern adjacent to the correction target edge 31a in the edge direction opposite to the correction target edge 31a is defined as a distance between edges W3.
[0020]
Further, taking into account the arrangement of the pattern formed on the side opposite to the correction target edge 31a with respect to the correction target pattern 31, the inter-pattern distance W1 and the line width W2 of the correction target pattern are used as parameters to correct the correction target edge. A correction table composed of correction amounts whose amounts have been corrected is created. Specifically, the correction amount set by the inter-edge distance W1 and the line width W2 using the inter-edge distance W3 as a parameter creates a correction amount corrected for each inter-edge distance W3. By correcting the correction target edge 31a based on the correction table created for each edge-to-edge distance W3, it is possible to perform pattern correction based on the correction amount in consideration of the situation of the pattern density around the correction target pattern 31. it can.
[0021]
Next, a procedure for acquiring a parameter for selecting a correction amount based on the correction table created in advance as described above will be described. When the edge included in the line width is extracted from the correction target edge (S2), all the edges of each pattern having the same line width W2 can be extracted as the correction target edge.
[0022]
Next, a pattern distance W1 is extracted (S3). The inter-pattern distance W1 is a distance between adjacent patterns among the patterns formed on the mask, and classifies each extracted edge according to the inter-pattern distance W1. For example, as shown in FIG. 4, since adjacent patterns are not formed on the left and right sides of the isolated pattern 21, the inter-pattern distance W1 which is the distance from the edges 21 a and 21 b of the isolated pattern 21 to the adjacent pattern. Is classified as infinity. Further, as shown in FIG. 5, the inter-pattern distance W1, which is the distance from the correction target edge 31a on the right side of the pattern 31 formed at the end of the one-dimensionally arranged pattern to the adjacent pattern, is set to infinity. And the other patterns are classified as equal. Here, the inter-pattern distance other than the inter-pattern distance W1 is a minimum interval that is the distance between adjacent patterns.
[0023]
Next, an edge having an inter-edge distance W3 greater than or equal to the line width W2 of the above-described correction target pattern is extracted (S4). Unlike the case where edges classified into the minimum interval range are extracted, no hidden line processing is performed on the edges. Normally, there is no data on the shape of the pattern in the region between the patterns, and processing is performed so that the width including this region is not recognized as the line width. However, when the edge of each pattern forming the line width of the pattern is extracted In contrast to this, by defining the distance between the edges of different patterns as the line width including the distance between adjacent patterns, the edge forming the edge-to-edge distance W3 as the line width including the area between the patterns is defined. Extract.
[0024]
Here, the hidden line process is a process relating to a method of selecting an edge when extracting the line width of a one-dimensionally arranged pattern. For example, for one edge to be extracted, the edge is defined as two types of edges such as an in-edge and an out-edge. The in-edge is an edge whose normal vector points in the figure of the pattern, and the out edge is an edge whose normal vector points out of the figure of the pattern. That is, when the pattern is rectangular, the line width W2 is defined as the distance between the in edges of the edges along the longitudinal direction of the pattern.
[0025]
The distance between the in-edge of another pattern formed on the mask is defined as an inter-edge distance W3, and the line width W2 of one of the patterns is the in-edge of one of the distances between the in-edge and the in-edge. Will be defined as the distance between them.
[0026]
For example, in the case of an isolated pattern, the distance between the in-edges corresponds to the line width W2 of the pattern, but when the patterns are arranged one-dimensionally, the distance between the in-edges between different patterns is different. The distance W3 between the edges, which is the distance, is larger than the line width W2 of these patterns. Therefore, when setting three parameters of the line width W1, the inter-pattern distance W2, and the inter-edge distance W3, one edge is defined by two types of an in-edge and an out-edge.
[0027]
Further, as shown in FIGS. 6 and 7, by giving a condition that the distance W3 between edges is larger than the line width W2 of the pattern 31, the edges 21a and 21b of the line width W2 of the isolated pattern 21 shown in FIG. Not applicable. At this time, the dimension of the edge-to-edge distance W3 forming an interval equal to or larger than the line width W2 is at least the dimension of "line width W2 of pattern to be corrected + distance W1 between patterns + line width W2 of pattern to be corrected". Here, the dimension of the inter-pattern distance W1 is the inter-pattern distance between adjacent patterns. Therefore, by extracting the edges forming the distance between edges equal to or larger than the line width W2 of the pattern to be corrected, all the edges other than the second edge 31b from the right are extracted as shown in FIG.
[0028]
Next, it is determined whether or not areas between edges having an edge distance W3 greater than or equal to the line width W2 overlap (S5). If they do not overlap, the correction target edge is corrected by a correction amount determined by the line width of the pattern and the distance between adjacent patterns (S6). At this time, the correction table referred to for obtaining a correction value is a correction table created only by a combination of a line width of a pattern and a distance between adjacent patterns.
[0029]
In the case of overlapping, another pattern is formed in a region opposite to the correction target edge 31a with respect to the correction target pattern 31, and one side in which another pattern is formed on one side of the correction target pattern 31 It is possible to confirm that the pattern is in a dense state, and it is indicated that the pattern is in a dense state in which another pattern is arranged in a direction opposite to the correction target edge 31a. Therefore, it can be confirmed that there is a pattern around the design value of the pattern, which may cause a deviation from the design value, and the optical or process influence of the pattern arranged around the pattern can be confirmed. The edge of the correction target pattern, that is, the line width is corrected based on the correction table in which the corresponding correction amount is considered (S7). For the correction target edge where the area between edges overlaps, correction for one-sided dense state is created for each distance between edges and set in a matrix by the line width of the pattern and the distance between patterns between adjacent patterns. The correction target edge is corrected based on the amount. At this time, when setting the correction amount, the correction target edge is referred to by referring to a correction table including a correction amount determined by a line width and an interval between adjacent patterns and a correction table determining a correction amount based on an overlapping area. 31a is corrected. Finally, the data relating to the correction amount is stored in a storage device provided with a processing device for performing the correction process (S8).
[0030]
By performing the pattern correction described above for the line widths of all the patterns, it is possible to correct all the patterns formed on the mask. Further, by setting the correction amount using the line width, the distance between patterns, and the distance between edges as parameters, the correction amount can be easily set without acquiring data such as shapes for all patterns formed on the mask. This makes it possible to correct the pattern accurately and quickly.
[0031]
【The invention's effect】
The distance between patterns in the edge direction of the edge to be corrected, the line width of the pattern, and the distance between edges, which is the distance to the edge on the far side of the adjacent pattern in the direction opposite to the edge, are used as parameters to determine the edge of the pattern to be corrected. By setting the correction amount, it is possible to perform the correction in consideration of the influence of the pattern that causes a shift in the correction amount of the correction target pattern optically or in a process, and it is possible to perform the correction with high accuracy. This is particularly effective when the density difference between the correction target edge and the other edge thereof, excluding the correction target pattern, is large.
[0032]
Furthermore, by setting the distance to the opposite edge of a pattern adjacent in the opposite direction to the correction target edge, implementation using a general purpose DRC (Design Rule Check) tool becomes possible, and simple and high-speed pattern correction can be performed. .
[Brief description of the drawings]
FIG. 1 is a flowchart illustrating a pattern correction method according to the present invention.
FIG. 2 is a plan view showing an example of a mask on which an isolated pattern is formed.
FIG. 3 is a plan view showing an example of a mask on which patterns arranged one-dimensionally are formed.
FIG. 4 is a plan view showing an example of a mask on which an isolated pattern is formed.
FIG. 5 is a plan view showing an example of a mask on which patterns arranged one-dimensionally are formed.
FIG. 6 is a plan view showing an example of a mask on which an isolated pattern is formed.
FIG. 7 is a plan view showing an example of a mask on which patterns arranged one-dimensionally are formed.
FIG. 8 is a plan view showing a mask on which an isolated pattern corrected by a conventional pattern correction method is formed.
FIG. 9 is a plan view showing a mask on which a one-dimensionally arranged pattern to be corrected by a conventional pattern correction method is formed.
[Explanation of symbols]
20, mask, 21a, edge, 21, 31 pattern, 31a, edge to be corrected

Claims (7)

In a pattern correction method for correcting the line width of a pattern formed on a mask by setting a correction value by optical proximity effect correction and bringing the line width of a transferred pattern closer to a design value,
The line width of the correction target pattern to be corrected, the inter-pattern distance between an adjacent pattern adjacent in the width direction of the correction target pattern and the correction target pattern, the correction target edge forming the correction target pattern, A pattern correction method, wherein a correction value is set based on an inter-edge distance to an edge on a far side of an adjacent pattern adjacent in an edge direction opposite to a correction target edge. A correction table is created in advance using the line width, the inter-pattern distance, and the inter-edge distance as parameters, and the correction target edge is corrected based on a correction value obtained from the correction table. The pattern correction method according to claim 1. 3. The pattern correction method according to claim 2, wherein the correction target edge is shifted in the width direction. 2. The pattern correction method according to claim 1, wherein the correction target pattern is a pattern formed at an end among patterns formed to be arranged one-dimensionally on the mask. 2. The pattern correction method according to claim 1, wherein the distance between the edges is at least larger than a dimension obtained by adding at least twice the line width and the distance between the patterns. A region between an adjacent pattern adjacent to the correction target pattern in the width direction and the correction target pattern, and an adjacent correction target edge forming the correction target pattern and an adjacent detection target edge adjacent to the correction target edge in an opposite edge direction. 3. The pattern correction according to claim 2, wherein a correction target edge in which a region up to a far edge of the pattern does not overlap is corrected based on a correction table using the line width and the inter-pattern distance as parameters. Method. A region between an adjacent pattern adjacent to the correction target pattern in the width direction and the correction target pattern, and an adjacent correction target edge forming the correction target pattern and an adjacent detection target edge adjacent to the correction target edge in an opposite edge direction. Correcting an edge to be corrected in which an area up to an edge on a far side of the pattern overlaps, based on a correction table created based on the distance between edges, using the line width and the distance between patterns as parameters. 3. The pattern correction method according to claim 2, wherein:

Images