JP2014115415A - Method for forming a reticle pattern - Google Patents

Abstract

A method of forming a reticle pattern that is unlikely to cause an error in reticle inspection.
A first step of performing OPC processing on a first pattern and a second pattern, a first vertex of a first interior angle of less than 180 ° included in the first pattern subjected to OPC processing, and the second processing subjected to OPC processing. A second step of checking whether the interval between the second vertices of the second interior angle less than 180 ° included in the pattern is narrower than a reference value; and the first vertex including the first vertex when the interval is smaller than the reference value And a third step of deleting the second vertex portion including the second vertex from the second pattern while deleting the portion from the first pattern.
[Selection] Figure 7

Description

  The present invention relates to a method for forming a reticle pattern.

  A semiconductor integrated circuit is formed by processing a semiconductor substrate or an insulating film on the semiconductor substrate based on a pattern obtained by layout design (hereinafter referred to as a layout pattern).

  The layout pattern is formed on a reticle after being subjected to OPC (Optical Proximity Correction) processing, and transferred to a resist film on a semiconductor substrate by a photolithography technique. A pattern transferred to the resist film (hereinafter referred to as a resist pattern) is used as a mask for processing (etching, ion implantation, etc.) a semiconductor substrate or an insulating film.

  When a layout pattern having a width of several tens of nanometers is directly formed on a reticle and transferred to a resist film, a resist pattern with rounded corners is formed. The OPC process is a process of correcting the layout pattern to bring the resist pattern shape closer to the layout pattern.

JP-A-2005-156606

  When the OPC process is completed, a resist pattern corresponding to the OPC-processed layout pattern is predicted by simulation (hereinafter referred to as lithography simulation).

  The predicted resist pattern is compared with the layout pattern before the OPC process. If the two patterns substantially match, an OPC-processed layout pattern is formed on the reticle. If the patterns do not match, the OPC parameter is changed and the OPC process is performed again.

  The pattern formed on the reticle is called a reticle pattern. The reticle is manufactured, for example, by forming a light shielding film (for example, a Cr film) on a transparent substrate by electron beam exposure. The manufactured reticle is inspected by reticle inspection.

  In the reticle inspection, first, the light shielding film pattern is photographed by the reticle inspection apparatus. Furthermore, a light shielding film pattern formed on the reticle is predicted by reticle simulation.

  The photographed light shielding film pattern (hereinafter referred to as an inspection image) is compared with the pattern predicted by the reticle simulation. If the two substantially match, the reticle inspection is passed, and an integrated circuit is manufactured using the passed reticle. If the two do not substantially match, the reticle inspection fails and the reticle is recreated.

  For example, when the isolated pattern obtained by the reticle simulation is combined with other patterns on the inspection image to form one pattern, the reticle inspection fails. Even when a single pattern obtained by the reticle simulation is separated into a plurality of patterns on the inspection image, the reticle inspection fails.

  By the way, even if there is no problem with the light-shielding film pattern on the reticle, the reticle inspection may be rejected if the resolution of the reticle inspection apparatus is low or the accuracy of the reticle simulation is low. Such an error is likely to occur when the pattern interval is narrowed by the OPC process or when the pattern width is narrowed by the OPC process.

  The reticle that has failed the reticle inspection is remade by a series of operations starting from the OPC processing of the layout pattern. For this reason, a great deal of time and effort is spent on remaking the reticle.

  Increasing the resolution of the reticle inspection system and the accuracy of the reticle simulation will reduce inspection errors and reduce reticle rework. However, it is not easy to increase the resolution of the reticle inspection apparatus and the accuracy of reticle simulation.

  In order to solve the above problem, according to one aspect of the present method, a first step of performing OPC processing on the first pattern and the second pattern, and a first step of less than 180 ° included in the OPC-processed first pattern. A second step of inspecting whether the interval between the first vertex of one interior angle and the second vertex of the second interior angle less than 180 ° included in the second pattern subjected to OPC processing is narrower than a reference value; and the interval is the reference value A reticle pattern having a third step of deleting a first vertex portion including the first vertex from the first pattern and deleting a second vertex portion including the second vertex from the second pattern if narrower A forming method is provided.

  According to the disclosed method, a reticle pattern that is unlikely to cause an error in reticle inspection is formed.

FIG. 1 is a configuration diagram of a pattern forming apparatus that executes the reticle pattern forming method according to the first embodiment. FIG. 2 is a flowchart showing a flow of the reticle pattern forming method according to the first embodiment. FIG. 3 is a diagram for explaining a reticle pattern forming method. FIG. 4 is a diagram for explaining a reticle pattern forming method. FIG. 5 is a diagram for explaining a reticle pattern forming method. FIG. 6 is a diagram for explaining a reticle pattern forming method. FIG. 7 is a flowchart of the vertex portion correction process. FIG. 8 is a diagram for explaining the convex vertex detection / concave vertex detection step. FIG. 9 is a diagram for explaining the convex vertex detection / concave vertex detection step. FIG. 10 is a flowchart of the convex vertex correction step. FIG. 11 is a flowchart of the correction step of the convex vertex portion. FIG. 12 is a diagram for explaining the convex vertex correction step. FIG. 13 is a diagram for explaining the convex vertex correction step. FIG. 14 is a diagram for explaining the correction step of the convex vertex portion. FIG. 15 is a diagram for explaining the correction step of the convex vertex portion. FIG. 16 is a diagram for explaining the correction step of the convex vertex portion. FIG. 17 is a diagram for explaining an example of the pattern structure of the first vertex portion and the second vertex portion. FIG. 18 is a diagram for explaining a resist pattern corresponding to the first pattern. FIG. 19 is a diagram illustrating step S46. FIG. 20 is a flowchart of the step of correcting the concave vertex portion. FIG. 21 is a flowchart of the correction step for the concave apex portion. FIG. 22 is a diagram for explaining the step of correcting the concave vertex. FIG. 23 is a diagram for explaining the step of correcting the concave vertex portion. FIG. 24 is a diagram for explaining the step of correcting the concave vertex. FIG. 25 is a diagram for explaining a convex vertex detection method and a concave vertex detection method. FIG. 26 is a diagram for explaining a method of determining whether or not the side of the opponent vertex is visible from the own vertex. FIG. 27 is a diagram for explaining a method of detecting an opponent vertex that faces the own vertex. FIG. 28 is a diagram for explaining the second embodiment. It is a figure which shows the relationship between the level | step difference of a 1st pattern in which the 1st vertex part was deleted, and retraction amount. FIG. 30 is a diagram for explaining the third embodiment. FIG. 31 is a diagram for explaining the fourth embodiment. FIG. 32 is a diagram for explaining the fifth embodiment. FIG. 33 is a diagram illustrating selection criteria for the shape and size of the first vertex portion and the shape and size of the second vertex portion.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the technical scope of the present invention is not limited to these embodiments, but extends to the matters described in the claims and equivalents thereof. In addition, the same code | symbol is attached | subjected to the corresponding part even if drawings differ, and the description is abbreviate | omitted.

(Embodiment 1)
(1) Pattern Forming Apparatus FIG. 1 is a configuration diagram of a pattern forming apparatus 2 that executes the reticle pattern forming method according to the first embodiment. The pattern forming apparatus 2 is a computer, for example.

  The pattern forming apparatus 2 includes, for example, a CPU (Central Processing Unit) 4, a ROM 6, a RAM (Random Access Memory) 8, and an HDD (Hard Disk Drive) 10 having a hard disk. Further, the pattern forming apparatus 2 includes a bus 12, an input device 14, and an output device 16.

  The input device 14 is a device that takes in data from the outside. The output device 16 is a device that outputs data to the outside.

  The CPU 4 controls the HDD 10, loads a program recorded in the HDD 10 into the RAM 8, and executes the loaded program. The HDD 10 is a computer-readable recording medium.

  The ROM 6 stores basic programs executed by the CPU 4. In addition to the program, the RAM 8 temporarily records intermediate data when the CPU 4 executes the program.

  In the HDD 10, a reticle pattern forming program (a program executable by the CPU 4) 18 for causing the CPU 4 to execute the layout pattern forming method of the first embodiment is recorded. The computer 2 becomes a pattern forming apparatus when the CPU 4 executes the reticle pattern forming program 18.

  Further, layout pattern data 20a (hereinafter referred to as layout data) is temporarily recorded in the HDD 10.

(2) Pattern Forming Method (i) Processing Flow FIG. 2 is a flowchart showing the flow of the reticle pattern forming method according to the first embodiment. 3 to 6 are diagrams for explaining a reticle pattern forming method.

-Layout pattern input (S2)-
First, the layout pattern 22 a is input to the pattern forming apparatus 2 via the input device 14. Now, it is assumed that the layout pattern 22a includes a first pattern 24a and a second pattern 25a as shown in FIG. 3, for example. The first pattern 24a and the second pattern 25a are preferably polygonal patterns.

  Specifically, for example, layout data having pattern data of the first pattern 24a and pattern data of the second pattern 25a is input. The layout data format is, for example, GDS II (Graphic Design System II) or OASIS (Open Artwork System Interchange Standard). The input layout data 20a is recorded in the HDD 10, for example.

  The CPU 4 converts the input layout data into an internal format. FIG. 4 is a table showing an example of the layout data 20b converted into the internal format.

  As shown in FIG. 4, the layout data 20b includes the number of patterns 30 included in the layout pattern 22a, the first pattern data 32b of the first pattern 24a and the second pattern data 33b of the second pattern 25a included in the layout pattern 22a. Have Each pattern data 32b and 33b has the number 34 of vertices contained in each pattern and the vertex coordinates 36 of each pattern. The pattern number 30, the vertex number data 34, and the vertex coordinates 36 are ordered. Each piece of data (number of patterns 30, number of vertices 34, vertex coordinates 36) has a pointer (not shown) indicating the position of the next data.

  In the example shown in FIG. 4, the number of patterns is 30, the first pattern data 32b, and the second pattern data 33b are ordered in this order. Further, the data of the first pattern data 32b and the second pattern data 33b are ordered in the order of the number of vertices 34 and the coordinates 36.

-Input of OPC parameters (S4)-
Next, the OPC parameters are input from the input device 14 to the pattern forming device 2. The input OPC parameters are recorded in the HDD 10, for example.

  The OPC parameter includes a parameter that defines a pattern portion (for example, the leading end portion of the pattern) to be OPC processed and a parameter that defines a pattern to be added to the pattern portion (or a pattern to be deleted from the pattern portion).

-Graphic processing (S6)-
Next, the CPU 4 converts the input layout pattern 22a into a form suitable for OPC processing. Specifically, for example, layout data 20b (FIG. 4) is converted.

  In the first embodiment, it is assumed that the layout data 20b in FIG. 4 is directly subjected to the OPC process without performing pattern conversion.

-OPC processing (S8)-
Next, the CPU 4 performs OPC processing on the first pattern 24a and the second pattern 25a. FIG. 5 shows an example of the layout pattern 22b subjected to the OPC process.

  For example, as shown in FIG. 5, the tip of the first pattern 24b is thickened by adding the pattern by OPC processing. Similarly, the tip of the second pattern 25b is also thickened. The internal angle of the first pattern 24b and the second pattern 25b is, for example, 90 ° or 270 °.

  Specifically, for example, the CPU 4 changes the layout data 20b. FIG. 6 shows an example of the layout data 20c converted by the OPC process. As shown in FIG. 6, the pattern data 32c, 33c vertex number 34 and vertex coordinates 36 are changed, added, or deleted.

-Vertex correction processing (S10)-
Next, the CPU 4 corrects the convex vertex portion included in a different pattern and approaching less than the first reference value el. Further, the CPU 4 corrects the concave vertex portion included in the same pattern and approaching less than the second reference value EL. Details of the correction process (S10) will be described later.

  By the correction process (S10), a reticle pattern that hardly causes an error in reticle inspection is formed.

-Lithography simulation (S12)-
Next, the CPU 4 predicts a resist pattern corresponding to the layout pattern 22b after the correction process by lithography simulation.

-OPC inspection (S14)-
Next, the CPU 4 determines whether or not the OPC process is properly performed. Specifically, for example, if the resist pattern obtained by the lithography simulation S12 substantially matches the layout pattern 22a before OPC (see FIG. 3), the CPU 4 determines that the OPC process has been performed appropriately. The CPU 4 determines that the OPC process has not been properly performed when the two patterns do not substantially match.

-Reticle pattern output (S16)-
When the CPU 4 determines that the OPC processing is properly performed, the CPU 4 outputs a layout pattern that has been subjected to the OPC processing and further corrected for the vertex portion as a reticle pattern.

Specifically, for example, the CPU 4 outputs reticle data having pattern data of the first pattern in which the OPC process is performed and the vertex portion is corrected, and pattern data of the second pattern in which the OPC process is performed and the vertex portion is corrected. At this time, the CPU 4
Convert the pattern data into a format suitable for reticle manufacturing.

-Change of OPC parameters (S18)-
If the CPU 4 determines in step 14 that the OPC process is inappropriate, the CPU 4 changes the OPC parameter. Thereafter, the CPU 4 returns to the OPC process (S8).

(Ii) Vertex correction processing (S10)
FIG. 7 is a flowchart of the vertex portion correction process (S10).

-Convex vertex detection / concave vertex detection step (S20)-
First, the CPU 4 detects convex vertices and concave vertices included in the OPC-processed first pattern 24b (S20). The CPU 4 further detects a convex vertex and a concave vertex included in the second pattern 25b that has been subjected to the OPC process (S20). 8 and 9 are diagrams for explaining the convex vertex detection / concave vertex detection step.

  FIG. 8 is an example of the first pattern 24b subjected to the OPC process. As shown in FIG. 8, the OPC-processed first pattern 24b has a plurality of vertices 40 (hereinafter referred to as convex vertices) of an inner angle 38 having an angle of less than 180 °. The inner angle is an angle formed by two adjacent sides of the polygon inside the polygon. The first pattern 24b further includes a plurality of vertices 44 (hereinafter referred to as concave vertices) having an inner angle 42 having an angle greater than 180 °.

  The vertex of the inner angle is also the vertex of the polygon having the inner angle. The CPU 4 detects the convex vertex 40 of the first pattern 24b and the convex vertex 40 of the second pattern 25b, and records the detected convex vertex. The CPU 4 further detects the concave vertex 44 of the first pattern 24b and the concave vertex 44 of the second pattern 25b, and records the detected concave vertex.

  Specifically, as shown in FIG. 9, the CPU 4 adds a flag 46 after each vertex coordinate 36 of the layout data 20c (FIG. 6) subjected to the OPC process. When the CPU 4 detects the convex vertex 40, the CPU 4 sets the flag after the coordinate 36 corresponding to the detected convex vertex to “0”. On the other hand, when the CPU 4 detects the concave vertex, it sets the flag after the coordinate 36 corresponding to the detected concave vertex to “1”.

-Correction of convex vertex (S22)-
10 and 11 are flowcharts of the convex vertex correction step (S22). 12-16 is a figure explaining the correction | amendment step (S22) of a convex vertex part.

-Step S30-
First, the CPU 4 determines whether or not there is a pattern having an unprocessed convex vertex (S30). If there is no pattern having an unprocessed convex vertex, the CPU 4 ends the convex vertex correction step (S22). The “processing” in the convex vertex correction step (S22) refers to steps S36 to S46 described later.

  If there is a pattern having an unprocessed convex vertex, the CPU 4 selects a pattern to be processed (for example, the first pattern 24b) from the unprocessed patterns, and proceeds to the next determination step (S32). When there are a plurality of patterns having unprocessed convex vertices, the CPU 4 selects a processing target pattern according to the order recorded in the layout data 20d (see FIG. 9), for example. The selected pattern is hereinafter referred to as a selection pattern. A pattern that is not selected (for example, the second pattern 25b) is hereinafter referred to as a non-selected pattern.

-Step S32-
Next, the CPU 4 determines whether or not an unprocessed vertex exists at the convex vertex of the selected pattern (hereinafter referred to as the selected pattern convex vertex) (S32). If there is no unprocessed selection pattern convex vertex, the process returns to step S30. Whether or not each vertex is a convex vertex is determined by a flag 46 (see FIG. 9).

  When there is an unprocessed selection pattern convex vertex, the CPU 4 selects a processing target candidate convex vertex (for example, the first convex vertex 48a in FIG. 5) from among the unprocessed selection pattern convex vertices, and proceeds to step S34. If there are a plurality of unprocessed selection pattern convex vertices, for example, the convex vertices of the processing target candidates are selected according to the order recorded in the layout data 20d (see FIG. 9) (for example, selected in the order of early order). . The vertices selected as processing target candidates are hereinafter referred to as own vertex candidates.

-Step S34-
Next, the CPU 4 has a convex vertex that has not been processed among the vertices of the non-selected pattern (for example, the second convex vertex 48b to the ninth convex vertex 48i in FIG. 5; hereinafter referred to as the non-selected pattern convex vertex). It is judged whether to do (S34). If there is no unprocessed non-selected pattern convex vertex, the process returns to step S32.

  If there is an unprocessed unselected pattern convex vertex, one convex vertex (for example, the second convex vertex 48b) is selected from the unprocessed unselected pattern convex vertices, and the process proceeds to step S36. The selected non-selected pattern convex vertex is hereinafter referred to as a partner vertex candidate. If there are a plurality of unprocessed unselected pattern convex vertices, the other vertex candidate is selected according to the order recorded in the layout data 20d as in step S32.

  When there are a large number of non-selected pattern convex vertices, the vertices selected as the opponent vertex candidates may be limited to the vertices facing the own vertex candidate.

-Step S36-
Next, the CPU 4 measures an interval cl between the own vertex candidate (for example, the first convex vertex 48a) and the counterpart vertex candidate (for example, the second convex vertex 48b) by, for example, an intersection search using the Pythagorean theorem and a unit vector. (S36).

-Step S38-
Next, the CPU 4 determines whether the measured interval cl is narrower than the first reference value el (S36). If the measured interval cl is greater than or equal to the first reference value el, the CPU 4 returns to step S34. When the measured interval cl is narrower than the first reference value el, the CPU 4 proceeds to the next step S40.

  When the interval between the convex vertices of adjacent patterns is narrow, the convex vertices may be connected in the image of the reticle inspection apparatus even if they are separated on the reticle. Further, the convex vertices may be connected to each other on the pattern obtained by the reticle simulation. In such a case, an error occurs in the reticle inspection.

  Such an error does not occur when the interval between the convex vertices is wide. The first reference value el is an interval that hardly causes the error. The first reference value el is preferably the narrowest interval among the intervals that do not cause the error. The first reference value el is, for example, 20 to 50 nm.

  When the CPU 4 detects an interval narrower than the first reference value el in step S36, it generates / updates error detection information (data). The error detection information is used in steps S40 to S46 described later.

  FIG. 12 is a table showing the error detection information 50. The error detection information 50 includes, for example, the number of times that an interval narrower than the first reference value el has been detected (interval error number 52).

  The error detection information 50 further includes the number 54 of the selected pattern (for example, the first pattern 24b) having the own vertex candidate whose interval cl is narrower than the first reference value el, and the own vertex candidate whose interval cl is narrower than the first reference value el. Number 56.

  Reference numeral 54 denotes the order of the pattern data 32d and 33d in the layout data 20d (see FIG. 9) (hereinafter the same). The number 56 is the order of the vertex coordinates 36 in the layout data 20d (the same applies hereinafter).

  The error detection information 50 further includes the number 58 of the non-selected pattern including the other vertex candidate whose interval cl with the own vertex candidate is narrower than the first reference value el, and the interval cl with the own vertex candidate is the first reference value el. It has a narrower partner vertex candidate number 60.

  As described above, the CPU 4 performs steps S36 to S38 between the convex vertex (first vertex) included in the first pattern 24b subjected to OPC processing and the convex vertex (second vertex) included in the second pattern 25b subjected to OPC processing. It is inspected whether the interval is narrower than the first reference value el.

-Step S40-
Next, the CPU 4 determines whether or not there is an abnormal value or less in the measured interval cl (S40). If there is an abnormal value or less, the CPU 4 proceeds to step S48. If there is no abnormal value below, the CPU 4 proceeds to step S42.

  The abnormal value is an interval at which there is a problem in the OPC process (S8) (see FIG. 2). The abnormal value is an interval (for example, 10 nm) narrower than the first reference value el.

  The own vertex candidate selected in steps S36 to S40 is hereinafter referred to as an own vertex. The partner vertex candidate selected in steps S36 to S40 is hereinafter referred to as a partner vertex.

-Step S48-
If there is a measured interval cl that is less than the abnormal value, the CPU 4 outputs an error warning and ends the program.

-Step S42-
The CPU 4 determines whether or not two sides (edges) that are in contact with the other vertex 74 (for example, the second point 48b) can be seen from the own vertex 72 (for example, the first convex vertex 48a in FIG. 5). A determination method by the CPU 4 will be described later. When two sides of the opponent vertex 74 are visible from the own vertex 72, the CPU 4 proceeds to step S44.

  The case where the two sides of the opponent vertex 74 can be seen from the own vertex 72 means that the straight line passing through the own vertex (first vertex) 72 and the opponent vertex (second vertex) 74 is between the two sides of the inner angle having the opponent vertex 74 as the vertex. Is to go through. For example, in the example shown in FIG. 5, a straight line 104 passing through the own vertex 72 and the other vertex 74 passes between the side 48 b 48 c of the inner angle 38 that appoints the other vertex 74 and the sides 48 b and 48 i.

  When one of the two sides of the opponent vertex 74 cannot be seen from the own vertex 72, the CPU 4 proceeds to step S46. The case where one of the two sides of the opponent vertex 74 cannot be seen from the own vertex 72 means that the straight line 104 passing through the own vertex 72 and the opponent vertex 74 does not pass between the two sides of the inner angle 38 having the opponent vertex 74 as the vertex. is there.

-Step S44-
When two sides of the opponent vertex 74 can be seen from the own vertex 72, the CPU 4 selects the first vertex portion including the own vertex (for example, the first convex vertex 48a) by, for example, SUB (subtract) processing (for example, the first pattern). 24a). The CPU 4 further deletes the second vertex portion including the counterpart vertex 74 (for example, the second convex vertex 48b) from the non-selected pattern (for example, the second pattern 25b) by, for example, the SUB process.

  FIG. 13 shows an example of the first vertex portion 62 and the second vertex portion 63. As shown in FIG. 13, the first vertex portion 62 has a step-like first outer peripheral line 64 that faces the own vertex 72 (for example, the first convex vertex 48a). The second vertex portion 63 has a step-like second outer peripheral line 65 that faces the counterpart vertex 74 (for example, the second convex vertex 48b).

  FIG. 14 shows a first pattern 24c from which the first vertex portion 62 has been deleted and a second pattern 25c from which the second vertex portion 63 has been deleted. As shown in FIG. 14, the interval c1 between the first pattern 24c and the second pattern 25c is widened by deleting the first vertex portion 62 and the second vertex portion 63.

  Specifically, for example, the CPU 4 first generates correction pattern data 66. FIG. 15 is a table showing an example of the correction pattern data 66. The correction pattern data 66 includes pattern data 68 of the first vertex portion 62 and pattern data 70 of the second vertex portion 63. The structure of the correction pattern data 66 is substantially the same as the structure of the layout data 20b (see FIG. 4) except that it does not have 30 patterns.

  Next, the CPU 4 changes the first pattern data 32d of the layout data 20d (see FIG. 9) subjected to the OPC process. FIG. 16 shows an example of the changed first pattern data 32e. The vertex coordinates 36 of the corrected first pattern 24c (see FIG. 14) increase as the number of steps (edges) increases. In FIG. 16, the flag 46 is omitted.

  The CPU 4 also changes the second pattern data 33d of the layout data 20d (see FIG. 9) subjected to the OPC process. FIG. 16 shows an example of the changed second pattern data 33e.

-Pattern structure of the vertex-
FIG. 17 is a diagram for explaining an example of the pattern structure of the first vertex portion 62 and the second vertex portion 63. As shown in FIG. 17A, the first outer peripheral line 64 (see FIG. 13) of the first vertex portion 62 has both ends (point A1, point A1) arranged at a certain distance from the counterpart vertex (second vertex) 74. B1). The first outer peripheral line 64 of the first apex portion 62 is further provided with a concave apex (point C1) disposed at the predetermined distance from the counterpart vertex (second apex) 74, and further away from the counterpart vertex 74 than the predetermined distance. It has the convex vertex (point D1, point E1) arranged. The fixed distance is (cl−el) / 2 + el (= average value of cl and el). A circle 51 in FIG. 17A is a circle centered on the opponent vertex 74 and having a radius of the certain distance.

  As shown in FIG. 17A, the point A <b> 1 is a point on one of the two sides that are in contact with the own vertex 72. The point B1 is a point on the other side of the two sides that are in contact with the own vertex 72. The point C1 is a point on a line passing through the midpoint between the points A1 and B1 and the other vertex 74 (for example, the second convex vertex 48b). The angles A1D1C1 and C1E1B1 are, for example, 90 °. The angle D1C1E1 is 270 °.

  As shown in FIG. 17B, the second outer peripheral line 65 (see FIG. 13) of the second vertex portion 63 has both ends (points A2, A2) arranged at the predetermined distance from the own vertex (first vertex) 72, as shown in FIG. Point B2). The second outer peripheral line 65 of the second vertex portion 63 further includes a concave vertex (point C2) arranged at a certain distance from the own vertex (first vertex) 72, and a distance farther than the certain distance from the own vertex 72. It has convex vertices (point D2, point E2) arranged. A circle 53 in FIG. 17B is a circle centered on its own vertex 72 and having a radius of the certain distance.

  As shown in FIG. 17B, the point A <b> 2 is a point on one of the two sides that are in contact with the opponent vertex 74. The point B <b> 2 is a point on the other side of the two sides that are in contact with the opponent vertex 74. The point C2 is a point on a line passing through the midpoint of the points A2 and B2 and the own vertex 72 (for example, the first convex vertex 48a).

  As shown in FIG. 17B, the second vertex portion 63 has a structure obtained by rotating the first vertex portion 62 by 180 °. The point A2 of the second vertex portion 63 corresponds to the point A1 of the first vertex portion 62. The points B2, C2, D2, and E2 of the second vertex portion 63 correspond to the points B1, C1, D1, and E1 of the first vertex portion 62, respectively. Angle A2D2C2 and angle C2E2B2 are 90 °. The angle D2C2E2 is 270 °.

  The distance between the point C1 and the point C2 is approximately el [≈ {(cl-el) / 2 + el} × 2-cl]. The distance between the points A1 and B2 is also substantially el. The distance between the points B1 and A2 is also approximately el.

  As described above, el is a distance at which adjacent convex vertices do not cause an error in reticle inspection. Therefore, according to step S44 (see FIG. 11), an error in the reticle inspection due to the close proximity of the convex vertices (where the other two sides can be seen) is less likely to occur.

―Resist pattern―
FIG. 18 is a diagram illustrating resist patterns corresponding to the first patterns 24b and 24c. FIG. 18 shows a resist pattern 78a (specifically, a rounded corner adjacent to the position of the first convex vertex 48a) corresponding to the first pattern 24b (see FIG. 5) subjected to the OPC process.

  FIG. 18 further shows a resist pattern 78b corresponding to the first pattern 24c (see FIG. 14) in which the OPC process is performed and the first vertex portion 62 is deleted. The step 80 of the first pattern 24c is 10 nm. The resist patterns 78a and 78b in FIG. 18 are obtained by lithography simulation.

  Due to the optical proximity effect, the outer peripheral line of the resist pattern 78a recedes from the first convex vertex 48a. The outer peripheral line of the resist pattern 78b corresponding to the first pattern 24c from which the first apex portion 62 is deleted is hardly retreated with respect to the outer peripheral line of the resist pattern 78a. Therefore, even if the first vertex portion 62 is deleted, the resist pattern hardly changes.

  However, if the step 80 of the first pattern 24c, that is, the step of the first apex portion 62 becomes too large, the retreat amount of the resist pattern 78b also increases. Therefore, in the first embodiment, the vertex portions 62 and 63 are deleted from both the own vertex 72 (for example, the first convex vertex 48a) and the counterpart vertex 74 (for example, the second convex vertex 48b). Thereby, even if the level difference between the first vertex portion 62 and the second vertex portion 63 is small, the distance between both ends (point A1, point B1) of the first pattern 24a and both ends (point A2, point B2) of the second turn 24b. Becomes wider. Further, the distance between the convex vertex (point C1) of the first pattern 24a and the convex vertex (point C2) of the second turn 24b is also widened. On the other hand, the receding amount of the resist pattern 78b hardly changes.

-Step S46-
Step S46 is similar to step S44. Therefore, the description of the parts common to step S44 is omitted or simplified. FIG. 19 is a diagram illustrating step S46.

  As described above, step 46 is executed when one of the two sides (edges) in contact with the opponent vertex 74 from the own vertex 72 is not visible. The CPU 4 deletes the first vertex portion 84 (see FIG. 19A) including the own vertex 72 from the pattern (for example, the first pattern 24b) including the own vertex 72 by, for example, SUB processing. The CPU 4 further deletes the second vertex portion 85 (see FIG. 19B) including the opponent vertex 74 from the pattern including the opponent vertex 74 (for example, the second pattern 25b).

  As shown in FIG. 19A, the first vertex portion 84 has an L-shaped third outer peripheral line 164 facing the own vertex 72 (first vertex). The first vertex portion 84 is rectangular.

  The first vertex portion 84 has a self vertex 72 (point B3) and a point A3 on the side that touches the self vertex 72 and is visible from the counterpart vertex 72. The first vertex portion 84 further has a point C3 on a line that is in contact with the opponent vertex 74 and that is not visible from the own vertex 72, and extends on the side of the own vertex 72.

  The distance between the partner vertex 74 and the point A3 is (cl-el) / 2 + el. The distance between the partner vertex 74 and the point C3 is also (cl-el) / 2 + el.

  The first vertex portion 84 further has a point D3 and a point E3. An angle A3D3C3 having the point D3 as a vertex is 90 °. The distance between the partner vertex 74 and the point D3 is greater than (cl-el) / 2 + el. An angle C3E3B3 having the vertex at the point E3 is 90 °. An angle E3B3A3 having the point B3 as a vertex is 90 °. The distance between the partner vertex 74 and the point E3 is greater than (cl-el) / 2 + el. The point C3 is a point on the side D3E3.

  As shown in FIG. 19B, the second vertex portion 85 has a structure obtained by rotating the first vertex portion 84 by 180 °. The point A4 of the second vertex portion 85 corresponds to the point A3 of the first vertex portion 84. The points B4, C4, D4, and E4 of the second vertex portion 85 correspond to the points B3, C3, D3, and E3 of the first vertex portion 84, respectively.

  As shown in FIG. 19B, the second vertex portion 85 has an L-shaped fourth outer peripheral line 165 that faces the counterpart vertex 74 (second vertex). The second vertex portion 85 is a rectangle.

  The distance between the side A3D3 of the first vertex portion 84 and the side A4D4 of the second vertex portion 85 is substantially el [= {(cl−el) / 2 + el} × 2−cl]. Further, the distance between the side D3E3 of the first vertex portion 84 and the side D4E4 of the second vertex portion 85 is also substantially el. Therefore, according to step S46, the error in the reticle inspection due to the close proximity of the convex vertices (where one of the opponent's sides cannot be seen) is reduced.

  As described above, the CPU 4 includes the own vertex 72 (first vertex) when the interval cl between the own vertex 72 (first vertex) and the opponent vertex 74 (second vertex) is smaller than the reference value el in steps S42 to S46. The first convex vertex portions 62 and 84 are deleted from the first pattern 24b (see FIG. 5). Alternatively, the CPU 4 deletes the second vertex portions 63 and 85 including the counterpart vertex 74 (second vertex) from the second pattern 25b.

-Correction of concave vertex (S24)-
The concave vertex correction step (S24) (see FIG. 7) is similar to the convex vertex correction (S22). Therefore, the description of the part common to the correction of the convex vertex part (S22) is omitted or simplified.

  20 and 21 are flowcharts of steps for correcting the concave apex portion. 22-24 is a figure explaining the correction | amendment step (S24) of a concave vertex part.

  As shown in FIG. 22A, a case is considered where the pattern 83a before the OPC process has adjacent concave vertices 86. The region including the concave vertex 86 is cut as shown in FIG. 22B by the OPC process (S8) (see FIG. 2). For this reason, the pattern width W is narrowed, and an error in reticle inspection is likely to occur.

  In the concave vertex correction step (S24), the concave vertex 88 that narrows the pattern width W is detected and corrected.

-Step S50-
First, the CPU 4 determines whether or not there is a pattern having an unprocessed concave vertex (S50). If there is no pattern having an unprocessed concave vertex, the CPU 4 ends the concave vertex correction step (S24). “Processing” in the step of correcting the concave apex portion (S24) refers to steps S54 to S64.

  If there is a pattern having an unprocessed concave vertex, the CPU 4 selects a pattern to be processed (pattern 83b) from patterns having an unprocessed concave vertex, and proceeds to the next determination step (S52).

-Step S52-
Next, the CPU 4 determines whether or not an unprocessed concave vertex exists in the selected pattern (S52). If there is no unprocessed concave vertex, the process returns to step S50.

  When there is an unprocessed concave vertex, the CPU 4 selects one concave vertex (hereinafter referred to as a self-vertex candidate) from among the unprocessed concave vertices. The CPU 4 further selects another concave vertex (hereinafter referred to as an opponent vertex candidate) from the unprocessed pattern concave vertices, and proceeds to step S54.

-Step 54-
Next, the CPU 4 measures the interval CL between the own vertex candidate and the opponent vertex candidate.

-Step S56-
Next, the CPU 4 determines whether the measured interval CL is narrower than the second reference value EL (S56). When the measured interval CL is narrower than the second reference value EL, the CPU 4 proceeds to the next step S58. If the measured interval CL is greater than or equal to the second reference value EL, the CPU 4 returns to step S52.

  If the distance between adjacent concave vertices is narrow, a single pattern may be separated into two patterns on the reticle inspection apparatus image (inspection image). Furthermore, the pattern obtained by the reticle simulation may be separated into two patterns. In such a case, an error occurs in the reticle inspection.

  Such an error does not occur when the interval between the concave vertices is wide. The second reference value EL is an interval that hardly causes the error. The second reference value EL is preferably the narrowest interval among the intervals that do not cause the error. The second reference value EL is, for example, 20 to 50 nm. The second reference value El may be the same as the first reference value el.

  As described above, in steps S54 to S56, the CPU 4 determines that the interval between the concave vertex (first vertex) included in the OPC-processed pattern 83b and the concave vertex (second vertex) included in the OPC-processed pattern 83b is the second. Inspect whether it is narrower than the reference value EL.

-Step S58, Step S60, and Step S66-
Step S58, step S60, and step S66 are substantially the same as step S40, step S42, and step S48 (see FIG. 11) of convex vertex correction (S22) (see FIG. 7), respectively. Therefore, description of these steps is omitted.

  The own vertex candidate selected in steps S54 to S60 is hereinafter referred to as an own vertex. The partner vertex candidate selected in steps S54 to S60 is hereinafter referred to as a partner vertex.

-Step S62-
Step S62 is performed when the distance CL between the own vertex 72 and the opponent vertex 74 is smaller than the reference value EL and two sides of the opponent vertex 74 can be seen from the own vertex 72. “The two sides of the opponent vertex 74 are visible from the own vertex 72” means that a straight line passing through the own vertex 72 and the opponent vertex 74 passes between two sides of the inner angle having the opponent vertex 74 as a vertex.

  For example, as shown in FIG. 23C, the CPU 4 converts the step-like first supplementary pattern 90 in contact with two sides (edges of the pattern 83 b) extending from the own vertex (first vertex) 72 into an OPC-processed pattern 83 b. For example, it is added by OR processing. The CPU 4 further adds a step-like second supplementary pattern 91 in contact with two sides (edges of the pattern 83b) extending from the counterpart vertex (second vertex) 74 to the OPC-processed pattern 83b by, for example, OR processing. As a result, the width of the pattern 83 c is increased in the vicinity of the own vertex 72 and the counterpart vertex 74.

  FIG. 23A shows the structure of the first supplement pattern 90. As shown in FIG. 23A, the structure of the first supplement pattern 90 is substantially the same as the structure of the step-like first apex portion 62 (see FIG. 17A).

  FIG. 23B shows the structure of the second supplement pattern 91. As shown in FIG. 23B, the structure of the second replenishment pattern 91 is substantially the same as the structure of the step-like second apex portion 63 (see FIG. 17B).

-Step S64-
Step S64 is executed when the distance CL between the own vertex 72 and the opponent vertex 74 is smaller than the second reference value EL and one of the two sides contacting the opponent vertex 74 from the own vertex 72 is not seen. “I can't see one of the two sides that touch the opponent's vertex 74 from the own vertex 72” means that the straight line passing through the own vertex 72 and the opponent's vertex 74 does not pass between the two sides of the inner corner with the opponent's vertex 74 as the vertex. is there.

  In this case, for example, as shown in FIG. 24C, the CPU 4 performs an OPC process on a pattern 83b that is a rectangular first supplementary pattern 92 that touches two sides (edges of the pattern 83b) extending from its own vertex (first vertex) 72. For example, by OR processing. Further, the CPU 4 adds a rectangular second supplement pattern 93 in contact with two sides (edges of the pattern 83b) extending from the counterpart vertex (second vertex) 74 to the OPC-processed pattern 83b by OR processing. As a result, the width of the pattern 83 c is increased in the vicinity of the own vertex 72 and the counterpart vertex 74.

  FIG. 24A shows the structure of the first supplement pattern 92. As shown in FIG. 24A, the structure of the first replenishment pattern 92 is substantially the same as the structure of the rectangular first apex portion 84 (see FIG. 19A).

  FIG. 24B shows the structure of the second supplement pattern 93. As shown in FIG. 24B, the structure of the second supplement pattern 93 is substantially the same as the structure of the rectangular second apex portion 85 (see FIG. 19B).

  By step S62 and step S64, the pattern width narrowed by the OPC process is widened, so that an error in reticle inspection is less likely to occur.

  As described above, when the interval CL between the own vertex 72 and the opponent vertex 74 is narrower than the second reference value EL in steps S60 to S64, the CPU 4 makes contact with the two sides extending from the own vertex 72 (first vertex). The supplement patterns 90 and 92 are added to the OPC corrected pattern 83b. Further, the CPU 4 adds the second supplementary patterns 91 and 93 in contact with the two sides extending from the counterpart vertex 74 (second vertex) to the OPC corrected pattern 83b.

  According to the above example, when the straight line 104 passes over one of the two sides instead of between the two sides of the inner angle 38 having the counterpart vertex 74 as a vertex, the first vertex portion 62 defines the third outer peripheral line 164. And the second vertex portion 63 has a fourth outer peripheral line 165.

  However, when the straight line 104 passes over one of the two sides, the first vertex portion 62 may have the first outer peripheral line 64 and the second vertex portion 63 may have the second outer peripheral line 65.

  In any case, when the straight line 104 does not pass between the two sides and on the two sides, the first vertex portion 62 has the third outer peripheral line 164 and the second vertex portion 63 has the second 4 outer peripheral lines 165 are provided.

(Iii) Figure detection / judgment method-convex vertex detection / concave vertex detection-
FIG. 25 is a diagram for explaining a convex vertex detection method and a concave vertex detection method.

  Each vertex coordinate 36 of the pattern data 32b to 32d, 33b to 33d (see FIG. 6) is the order in which each vertex passes through the outer periphery of the pattern 95 corresponding to the pattern data 32 and 33, for example, in the counterclockwise direction 97. It is recorded in.

  In this case, as shown in FIG. 25, with respect to a side (edge) extending from the (i−1) -th vertex 94 toward the i-th vertex 96, the i-th vertex 96 toward the i + 1-th vertex 98. If the extending side is located on the right side, the i-th vertex 96 is a concave vertex.

  On the other hand, with respect to the side extending from the i-th vertex 96 toward the i + 1st vertex 98, the side extending from the i + 1st vertex 98 toward the i + 2nd vertex 100 is bent to the left side. In this case, i + 1 is the 98th vertex convex vertex.

  CPU4 detects a convex vertex and a concave vertex based on such a law.

―Judgment method of projection relationship―
FIG. 26 is a diagram for explaining a method of determining whether or not the sides 102 and 103 in contact with the opponent vertex 74 can be seen from the own vertex 72 (projection relationship). First, the side (edge) 102 positioned in the clockwise direction around the partner vertex 74 will be described.

  As shown in FIG. 26A, when the side 102 is positioned on the left side of the straight line 104 extending from the own vertex 72 toward the other vertex 74, the side 102 cannot be seen from the own vertex 72. On the other hand, as shown in FIGS. 26B to 26D, when the side 102 is located on the right side of the straight line 104, the side 102 can be seen from the own vertex 72.

  Whether the side 102 is located on the right side or the left side of the straight line 104 can be determined based on whether the straight line 104 has the same inclination, the side 102 has the same inclination, and whether the direction of the straight line 104 and the side 102 is the same. “The direction of the straight line 104 and the side 102 is the same” means that the angle formed by the straight line 104 and the side 102 is less than 90 °. Alternatively, it can be determined from the outer product of the straight line 104 and the side 102 whether the side 102 is located on the right side or the left side of the straight line 104.

  For the side 103 positioned in the counterclockwise direction with the counterpart vertex 74 as the center, the positional relationship between the straight line 104 and the side 103 is reversed. That is, when the side 103 is located on the right side of the straight line 104, the side 103 cannot be seen from the own vertex 72. On the other hand, when the side 102 is located on the left side of the straight line 104, the side 102 can be seen from the own vertex 72.

  Based on such a rule, the CPU 4 recognizes a side that can be seen from its own vertex and a side that cannot be seen in steps S42 and S60. When the straight line 104 passes over the side 103 (or the side 104) as shown in FIG. 26C, the first vertex portion and the second vertex portion may be stepped or rectangular as described above. There may be.

―Detection method of opposing vertices―
FIG. 27 is a diagram for explaining a method of detecting the opponent vertex 74 that faces the own vertex 72. Here, consider a case where the own vertex 72 and the counterpart vertex 74 are convex vertices.

  As shown in FIG. 27A, the angle of the angle formed by the line segment 106 connecting the own vertex 72 and the opponent vertex 74 and the one side 107 contacting the opponent vertex 74 and the other of the line segment 106 and the other vertex 74 contacting each other. When the sum of the angles 110 formed by the side 109 is 270 °, the own vertex 72 and the opposite vertex 74 face each other.

  On the other hand, as shown in FIG. 27 (b), the line segment 106 that connects the own vertex 72 and the partner vertex 74 and the angle 108 formed by one side 107 that contacts the partner vertex 74, and the line segment 106 contacts the partner vertex 74. When the sum of the angles 110 formed by the other side 109 is other than 270 °, the own vertex 72 and the opponent vertex 74 are not opposed to each other.

  Based on such a rule, the CPU 4 may limit the partner vertex candidates to vertices facing the own vertex 72 in steps S34 and S52.

(Embodiment 2)
The second embodiment is similar to the first embodiment. Therefore, description of portions common to Embodiment 1 is omitted or simplified.

  FIG. 28 is a diagram for explaining the second embodiment. FIG. 28 shows a partially enlarged view of the first pattern 24b (see FIG. 5) subjected to the OPC process and a partially enlarged view of the resist pattern 78a corresponding to the first pattern 24b.

  FIG. 28 further shows a partially enlarged view of the first pattern 24c (see FIG. 14) in which the OPC process is performed and the first vertex portion 62 (see FIG. 13) is deleted. FIG. 28 further shows a partially enlarged view of the resist pattern 78b corresponding to the first pattern 24c (a rounded corner adjacent to the first convex vertex 48a). The step 80 of the first pattern 24c is 50 nm.

  FIG. 18 of the first embodiment shows an example where the step 80 of the first pattern 24c is 10 nm. In the example shown in FIG. 18, the retraction amount of the resist pattern 78b from the resist pattern 78a is only about 5 nm.

  On the other hand, in the example shown in FIG. 28, the retraction amount (deviation amount) 120 of the resist pattern 78b from the resist pattern 78a is 54 nm. This is because the step of the first pattern 24c in FIG. 28 is as large as 50 nm.

  Therefore, for example, in step S44 (see FIG. 11), the CPU 4 prevents each side of the first outer peripheral line 64 of the first vertex portion 62 (see FIG. 13) from exceeding a certain length (for example, 10 to 50 nm). The number of steps of the first outer peripheral line 64 is adjusted.

  Further, for example, in step S44, the CPU 4 keeps the second outer peripheral line 65 so that each side of the second outer peripheral line 65 of the second vertex portion 63 (see FIG. 13) does not become a certain length (for example, 10 to 50 nm) or more. Adjust the number of steps.

  FIG. 28 shows the first pattern 124 and the resist pattern 78c corresponding to the first pattern 124 when the number of steps of the first apex portion 62 is four and the step is 25 nm. As shown in FIG. 28, the amount of retraction is reduced by reducing the step. In the example shown in FIG. 28, the receding amount decreases from 54 nm to 46 nm.

  FIG. 29 is a diagram showing the relationship between the step of the first pattern 24c from which the first vertex portion 62 has been deleted and the retraction amount. There are two stages. The horizontal axis is a step. The vertical axis represents the retraction amount. As shown in FIG. 29, when the step of the first pattern 24c is 10 nm or more (particularly 30 nm or more), the retraction amount 120 increases rapidly. Accordingly, the step of the first pattern 24c from which the first vertex portion 62 is deleted is preferably 10 nm to 30 nm. The same applies to the second pattern 25c from which the second vertex portion 63 has been deleted.

(Embodiment 3)
The third embodiment is similar to the first embodiment. Therefore, description of portions common to Embodiment 1 is omitted or simplified.

  FIG. 30 is a diagram for explaining the third embodiment. FIG. 30 shows the first pattern 24c of the first embodiment in which the first vertex portion 62 is deleted. FIG. 30 further shows the second pattern 25c of the first embodiment in which the second vertex portion 63 is deleted. Points A1 to C2 have been described with reference to FIG.

  In the first embodiment, as described with reference to FIG. 17, the distance between the own vertex 72 and the point A2, the distance between the own vertex 72 and the point B2, and the distance between the own vertex 72 and the point C2 are all the same. For this reason, as shown in FIG. 30B, the point A2 and the point B2 protrude from the point C2 toward the own vertex 72 side. Similarly, the points A1 and AB protrude from the point C1 toward the counterpart vertex 74.

  An arc 134 in FIG. 30B is a part of a circle centered on its own vertex 72. The distance between the points A1 and B2 and the distance between the points B1 and A2 are slightly shorter than the distance between the points C1 and C2 due to the protrusion. The straight line 130 is a straight line passing through the points A2 and B2.

  Therefore, in the third embodiment, as shown in FIG. 30 (c), the points A2 and B2 are set such that the distance between the points A1 and B2 and the distance between the points B1 and A2 are the same as the distance between the points C1 and C2. Is moved away from its own vertex 72. Similarly, the points A1 and B1 are moved away from the opponent vertex 74.

  The straight line 132 is a straight line that passes through the point A 2 away from the own vertex 72 and the point B 2 away from the own vertex 72. The points A2 and B2 are moved away from the own vertex 72 so that the straight line 132 is parallel to the straight line 130, for example. Similarly, the points A1 and B2 are moved away from the opponent vertex 74.

  Therefore, the distance between the concave vertex (point C1) on the first outer peripheral line 64 of the first vertex portion 62 and the concave vertex (point C2) on the second outer peripheral line 65 closest to the concave vertex, The distance between one end (point A1) and the end point (point B2) of the second outer peripheral line 65 closest to the one end (of the second vertex portion 63) becomes equal. Further, the distance between one end (point A1) of the first outer peripheral line 64 and the end point (point B2) of the second outer peripheral line 65 closest to the one end, the other end (point B1) of the first outer peripheral line 64, and the other end The distance between the end points (point A2) of the nearest second outer peripheral line 65 becomes equal.

According to the third embodiment, the interval between the points A1 and B2 and the interval between the points B1 and A2 are widened, so that an error in the reticle inspection is further less likely to occur.
(Embodiment 4)
The fourth embodiment is similar to the first embodiment. Therefore, description of portions common to Embodiment 1 is omitted or simplified.

  FIG. 31 is a diagram for explaining the fourth embodiment. FIG. 31 shows the structure of the first vertex portion 162 and the second vertex portion 163 of the fourth embodiment. Except for points C1 and C2, points A1 to E2 have been described with reference to FIG.

  As described with reference to FIG. 17A, in the first embodiment, the point C1 is a point on a line passing through the midpoint between the points A1 and B1 and the other vertex 74 (for example, the second convex vertex 48b). is there. The same applies to the point C2.

  On the other hand, in the fourth embodiment, as shown in FIG. 31A, the point C1 is a point on a line passing through the opponent vertex 74 and the own vertex 72 (for example, the first convex vertex 48a). Similarly, the point C2 is a point on a line passing through the own vertex 72 and the opponent vertex 74, as shown in FIG. Therefore, the interval between the points C1 and C2 is exactly e1, which is wider than that of the first embodiment. Therefore, an error in reticle inspection is less likely to occur than in the first embodiment.

(Embodiment 5)
The fifth embodiment is similar to the first embodiment. Therefore, description of portions common to Embodiment 1 is omitted or simplified.

  FIG. 32 is a diagram for explaining the fifth embodiment. FIG. 32 shows a first pattern 224 from which the first vertex portion is deleted and a second pattern 225 from which the second vertex portion is deleted. The broken lines in FIG. 32 indicate the first vertex portion 62 and the second vertex portion 63 of the first embodiment.

  The first vertex portion 262 of the fifth embodiment includes a side connecting the point A1 and the point C1 of the first vertex portion 62 of the first embodiment, a side connecting the point C1 and the point B1, and a point B1 and the own vertex 72. It is a polygon having a side to be connected and a side to connect the own vertex 72 and the point A1. Similarly, the second vertex portion 263 of the fifth embodiment includes the side connecting the point A2 and the point C2 of the second vertex portion 63 of the first embodiment, the side connecting the point C2 and the point B2, and the point B2 and the opponent. It is a polygon having a side connecting the vertices 74 and a side connecting the opponent vertex 74 and the point A2.

  For this reason, the outer peripheral line of the 1st pattern 224 from which the 1st vertex part 262 was deleted has a smooth shape without a level | step difference. Similarly, the outer peripheral line of the second pattern 225 from which the second vertex portion 263 has been deleted has a smooth shape having no step.

  As shown in the first to fifth embodiments, the shape and size of the first vertex portion and the second vertex portion are not limited to a certain shape and size. However, it is preferable that the shapes and sizes of the first vertex portion and the second vertex portion satisfy the following conditions.

  FIG. 33 is a diagram for explaining selection criteria for the shape and size of the first vertex portion 62 and the shape and size of the second vertex portion 63 according to the first embodiment.

  First, consider the formation of a resist pattern corresponding to the OPC-processed first pattern 24b (see FIG. 5) and a resist pattern corresponding to the OPC-processed second pattern 25b.

  First, a reticle having a first pattern 24b subjected to OPC processing and a second pattern 25b subjected to OPC processing is formed. A plurality of first resist patterns are formed by photolithography using the reticle as a mask.

  FIG. 33A shows a third resist pattern 78a (specifically, a rounded corner adjacent to the position 138 of the first convex vertex 48a) corresponding to the first pattern 24b among the plurality of first resist patterns. Has been. FIG. 33A further shows a fourth resist pattern 79a (specifically, a rounded corner adjacent to the position 139 of the second convex vertex 48b) corresponding to the second pattern 25b among the plurality of first resist patterns. It is shown.

  Next, a reticle having a first pattern 24c from which the first vertex portion 62 has been deleted and a second pattern 25c from which the second vertex portion 63 has been deleted is formed. A plurality of second resist patterns are formed by photolithography using the reticle as a mask.

  FIG. 33B shows a fifth resist pattern 78b (specifically, a rounded corner adjacent to the position 138 of the first convex vertex 48a) corresponding to the first pattern 24c and a plurality of second resist patterns. A sixth resist pattern 79b (specifically, a rounded corner adjacent to the position 139 of the second convex vertex 48b) corresponding to the second pattern 25c among the second resist patterns is shown.

  As described with reference to FIG. 18, the fifth resist pattern 78b recedes inward from the position of the outer peripheral line 200a (see FIG. 33A) of the third resist pattern 78a. Similarly, the sixth resist pattern 79b recedes inward from the position of the outer peripheral line 201a of the fourth resist pattern 79a. In the first embodiment, the shapes and sizes of the first vertex portion 62 and the second vertex portion 63 are selected so that the amount of retreat at this time becomes an allowable value (for example, 10 to 50 nm) or less.

  The shape and size are similarly selected for the concave vertex portion of the first embodiment and the first vertex portion and the second vertex portion of the second to fifth embodiments.

  Incidentally, the position of the outer peripheral line 200a of the third resist pattern 78a (see FIG. 33A) can be represented by a position 136a where the radius of curvature of the outer peripheral line 200a is minimum. The receding amount of the fifth resist pattern 78b (see FIG. 33B) is, for example, the position 136b where the radius of curvature of the outer peripheral line 200b of the fifth resist pattern 78b is minimized and the outer peripheral line 200a of the third resist pattern 78a (see FIG. 33 (a)) is a distance from the position 136a at which the radius of curvature is minimized.

  Similarly, the retraction amount of the sixth resist pattern 79b (see FIG. 33B) is, for example, the position 137b where the radius of curvature of the outer peripheral line 201b of the sixth resist pattern 79b is minimized and the outer peripheral line 201a of the fourth resist pattern 79a. The distance from the position 137a (see FIG. 33 (a)) at which the radius of curvature is minimum.

  In the description of the first to fifth embodiments, there are two patterns included in the layout pattern 22a. However, the layout pattern 22a may include three or more patterns.

  Regarding the above first to fifth embodiments, the following additional notes are disclosed.

(Appendix 1)
A method of forming a reticle pattern,
A first step of OPC processing the first pattern and the second pattern;
The interval between the first vertex of the first interior angle less than 180 ° included in the first pattern subjected to the OPC process and the second vertex of the second interior angle less than 180 ° included in the second pattern subjected to the OPC process is a reference value. A second step to check for narrowerness,
When the interval is narrower than the reference value, the first vertex portion including the first vertex is deleted from the first pattern, and the second vertex portion including the second vertex is deleted from the second pattern. A method for forming a reticle pattern comprising a third step.

(Appendix 2)
In the method of forming a reticle pattern according to appendix 1,
When a straight line passing through the first vertex and the second vertex passes between two sides of the second interior angle, the first vertex portion has a step-like first outer peripheral line facing the first vertex. And the second vertex portion has a stepped second outer peripheral line facing the second vertex,
When the straight line does not pass between the two sides of the second interior angle and on the two sides, the first vertex portion has an L-shaped third outer peripheral line facing the first vertex, and The second vertex portion has an L-shaped fourth outer peripheral line facing the second vertex,
When the straight line passes over one of the two sides of the second interior angle, the first vertex portion has a first outer peripheral line and the second vertex portion has a second outer peripheral line, or the first vertex portion The method for forming a reticle pattern, comprising: having the third outer peripheral line, and the second vertex portion having the fourth outer peripheral line.

(Appendix 3)
In the method of forming a reticle pattern according to appendix 2,
The first outer peripheral line of the first apex portion has both ends arranged at a certain distance from the second apex, and apexes having an interior angle arranged at the certain distance from the second apex greater than 180 °, A vertex having an interior angle of less than 180 ° disposed farther than the distance from the second vertex,
The second outer peripheral line of the second apex portion includes both ends arranged at the fixed distance from the first apex, and apexes having an interior angle arranged at the constant distance from the first apex greater than 180 °. A vertex having an interior angle of less than 180 ° disposed at a distance greater than the certain distance from the first vertex,
The method of forming a reticle pattern, wherein the fixed distance is an average value of the reference value and the interval.

(Appendix 4)
In the method of forming a reticle pattern according to appendix 2,
A distance between a third vertex having an inner angle on the first outer peripheral line of the first vertex portion larger than 180 ° and a vertex on the second outer peripheral line having an inner angle closest to the third vertex larger than 180 °; and the first outer periphery. The distance between one end of the line and the end point of the second outer peripheral line closest to the one end and the distance between the other end of the first outer peripheral line and the end point of the second outer peripheral line closest to the other end have a common value. A method for forming a reticle pattern, comprising:

(Appendix 5)
In the method for forming a reticle pattern according to appendix 2 or 3,
In the third step, the number of steps of the first outer peripheral line and the second outer peripheral line is adjusted so that each side of the first outer peripheral line and the second outer peripheral line does not exceed a certain length. A method of forming a characteristic reticle pattern.

(Appendix 6)
In the method for forming a reticle pattern according to any one of appendices 1 to 5,
The shape and size of the first vertex part are the first pattern subjected to the OPC process before the first vertex part is deleted and the second pattern subjected to the OPC process before the second vertex part is deleted. The first pattern and the second vertex portion in which the first vertex portion is deleted from the position of the outer peripheral line of the resist pattern corresponding to the first pattern among the plurality of first resist patterns formed by the reticle having Among the second resist patterns formed by the reticle having the second pattern from which the pattern is deleted, the receding amount by which the resist pattern corresponding to the first pattern recedes toward the inside of the resist pattern is less than the allowable value. Selected
The shape and size of the second apex portion are changed from the position of the outer peripheral line of the resist pattern corresponding to the second pattern of the plurality of first resist patterns to the second pattern of the plurality of second resist patterns. A method for forming a reticle pattern, characterized in that the corresponding resist pattern is selected such that the amount of retraction of the corresponding resist pattern toward the inside of the resist pattern is equal to or less than the allowable value.

(Appendix 7)
A reticle pattern forming method for forming pattern data of the pattern corrected by OPC from layout data having pattern data of a pattern included in a layout pattern,
A first step of OPC processing the pattern;
A second step of checking whether a distance between a first vertex having an inner angle greater than 180 ° included in the OPC-processed pattern and a second vertex having an inner angle greater than 180 ° included in the pattern is smaller than a reference value;
When the interval is smaller than the reference value, a supplement pattern that touches two sides of the pattern extending from the first vertex is added to the pattern, and a supplement pattern that touches two sides of the pattern extending from the second vertex is added. A method of forming a reticle pattern comprising the step of adding to the pattern.

20a to 20d ... layout data 22 ... layout patterns 24a to 24c ... first patterns 25a to 25c ... second patterns 32a to 32e, 33a to 33e ... pattern data 38, 42 ... Inner angle 40: convex vertex 44 ... concave vertex 62, 84, 162, 262 ... first vertex portion 63, 85, 163, 263 ... second vertex portion 64 ... first outer peripheral line 65 ... second outer peripheral line 72 ... own vertex 74 ... mating vertex 86 ... concave vertex 88 ... convex vertex 90,91 ... stepped replenishment patterns 92,93 ... rectangular Replenishment pattern

Claims (5)

A method of forming a reticle pattern,
A first step of OPC processing the first pattern and the second pattern;
The interval between the first vertex of the first interior angle less than 180 ° included in the first pattern subjected to the OPC process and the second vertex of the second interior angle less than 180 ° included in the second pattern subjected to the OPC process is a reference value. A second step to check for narrowerness,
When the interval is narrower than the reference value, the first vertex portion including the first vertex is deleted from the first pattern, and the second vertex portion including the second vertex is deleted from the second pattern. A method for forming a reticle pattern comprising a third step. The method of forming a reticle pattern according to claim 1,
When a straight line passing through the first vertex and the second vertex passes between two sides of the second interior angle, the first vertex portion has a step-like first outer peripheral line facing the first vertex. And the second vertex portion has a stepped second outer peripheral line facing the second vertex,
When the straight line does not pass between the two sides of the second interior angle and on the two sides, the first vertex portion has an L-shaped third outer peripheral line facing the first vertex, and The second vertex portion has an L-shaped fourth outer peripheral line facing the second vertex,
When the straight line passes over one of the two sides of the second interior angle, the first vertex portion has a first outer peripheral line and the second vertex portion has a second outer peripheral line, or the first vertex portion The method for forming a reticle pattern, comprising: having the third outer peripheral line, and the second vertex portion having the fourth outer peripheral line. The method of forming a reticle pattern according to claim 2,
The first outer peripheral line of the first apex portion has both ends arranged at a certain distance from the second apex, and apexes having an interior angle arranged at the certain distance from the second apex greater than 180 °, A vertex having an interior angle of less than 180 ° disposed farther than the distance from the second vertex,
The second outer peripheral line of the second apex portion includes both ends arranged at the fixed distance from the first apex, and apexes having an interior angle arranged at the constant distance from the first apex greater than 180 °. A vertex having an interior angle of less than 180 ° disposed at a distance greater than the certain distance from the first vertex,
The method of forming a reticle pattern, wherein the fixed distance is an average value of the reference value and the interval. In the method of forming a reticle pattern according to claim 2 or 3,
In the third step, the number of steps of the first outer peripheral line and the second outer peripheral line is adjusted so that each side of the first outer peripheral line and the second outer peripheral line does not exceed a certain length. A method of forming a characteristic reticle pattern. A reticle pattern forming method for forming pattern data of the pattern corrected by OPC from layout data having pattern data of a pattern included in a layout pattern,
A first step of OPC processing the pattern;
A second step of checking whether a distance between a first vertex having an inner angle greater than 180 ° included in the OPC-processed pattern and a second vertex having an inner angle greater than 180 ° included in the pattern is smaller than a reference value;
When the interval is smaller than the reference value, a supplement pattern that touches two sides of the pattern extending from the first vertex is added to the pattern, and a supplement pattern that touches two sides of the pattern extending from the second vertex is added. A method for forming a reticle pattern, comprising: a third step of adding to the pattern.

Images