Opus correction method

Abstract

The application discloses an OPC correction method, comprising the following steps: step one, providing an initial target layer and setting a minimum resolution size of a mask plate; step two, selecting a first pattern which will violate mask rule checking in subsequent MBOPC correction from the initial target layer; step three, segmenting each side of the first pattern into multiple segments by segmenting the first pattern, the segments including corner segments and intermediate segments, and the initial target layer after segmenting is a second target layer; step four, performing MBOPC correction based on the second target layer and obtaining a mask plate layer, the MBOPC correction including multiple iteration loop operations, in each iteration loop operation, the corner segments and the intermediate segments are corrected separately, the corner segments are used to control the size of the top corner of the first pattern in the mask plate layer, and the intermediate segments are used to control the area of the first pattern in the mask plate layer. The application can realize the regulation and control of the size of the pattern in the mask plate layer, and can make the exposure pattern meet the target value and the production demand without violating the MRC.

Description

TECHNICAL FIELD

[0001] The present application relates to a semiconductor integrated circuit manufacturing method, in particular to an optical proximity correction (OPC) method. BACKGROUND

[0002] With the decrease of semiconductor manufacturing process nodes, especially for the process nodes below 14nm, the critical dimension of the pattern is close to the lithography limit, and the pattern structure is relatively more complex, so the mask correction is facing great challenges. In the real exposure process, due to the interference and diffraction effect of light, the exposure pattern will appear distortion, including angle rounding, line shortening, etc. Optical proximity correction (OPC) has become an indispensable process link. The process of OPC correction is essentially an iterative process. Through model simulation, the predicted contour pattern is infinitely close to the target value. In the correction process, when the mask pattern size such as critical dimension (CD) or space (Space) is close to the set mask rule check (MRC) rule (Rule), i.e. the minimum resolution size of the mask template, the edge will not move, avoiding the violation of MRC. In the correction process, the MRC rule is usually set artificially in advance, including the minimum CD and Space of the mask resolution. The set MRC rule is generally determined by the process node and the process capability of the mask shop (Mask Shop), and the process capability of the mask shop (Mask Shop) is the minimum CD and / or Space of the mask manufacturing resolution.

[0003] For lower process nodes, in addition to the change of the conventional exposure condition, the error caused by the mask manufacturing difference is particularly important. If there is a MRC problem, the real exposure pattern has the risk of opening, and the actual process window is small. The via layer plays a connecting role in the actual process, for example, it is a channel connecting the metal layer. The staggered structure is very common in the via layer, and the Space between the opposite angles is usually very small, and the MRC problem is often faced in the OPC correction.

[0004] The existing OPC correction method will be described in combination with the drawings as follows:

[0005] As shown in FIG. 1, it is a flow chart of the existing OPC correction method; the existing OPC correction method includes the following steps: Figure 1

[0006] Step S101, providing an initial layout, Figure 1 wherein Drawn represents the initial layout, which is formed by graphically drawing the circuit.

[0007] ​Step S102, rule-based OPC is performed on the initial layout to obtain a target layer, Figure 1 The target layer is also denoted as target.

[0008] As Figure 2A shown in FIG. 1 is a schematic diagram of the pattern structure of the target layer in the prior art OPC correction method. The target layer 201 includes a plurality of patterns 202. Figure 2A The patterns 202 in the target layer 201 are square, and the array structure formed by the patterns 202 is a dense staggered array structure. In the dense staggered array structure, the diagonals of the patterns 202 are aligned and periodically arranged. The minimum distance between the patterns 202 is the diagonal-to-diagonal distance d201, i.e., the distance between the adjacent corners of two adjacent patterns 202. The pitch of the patterns 202 is the sum of the diagonal length of the patterns and the minimum distance. Generally, the patterns in the via layer are square and arranged in the dense staggered array structure.

[0009] Step S103, model-based OPC correction is performed on the target layer to obtain a mask layer, Figure 1 The mask layer is also denoted as mask. The model-based OPC correction includes a plurality of iterative loop operations.

[0010] As Figure 2B shown in FIG. 2 is a schematic diagram of the pattern structure of the mask layer formed by the prior art OPC correction method. The patterns 204 in the mask layer 203 correspond to the patterns 202 in the target layer 201. The patterns 204 are obtained by performing a plurality of iterative loop operations on the patterns 202. Figure 2A

[0011] Step S104, MRC detection is performed, Figure 1 The MRC detection is also denoted as MRC. The MRC detection needs to ensure that Figure 2B the critical dimension and the diagonal-to-diagonal distance of the patterns 204 in the mask layer 203 satisfy the minimum resolution dimension of the mask, i.e., the minimum resolution value of the critical dimension and the minimum resolution value of the distance.

[0012] Step S104 can be inserted into each iterative loop operation of the MBOPC. When the critical dimension and the diagonal-to-diagonal distance of the patterns 204 approach the minimum resolution value of the critical dimension and the minimum resolution value of the distance, the edges will not be moved any more to avoid violating the MRC. As Figure 2A shown in FIG. 3, in the dense staggered array structure, the diagonal-to-diagonal distance d201 of the patterns 202 is small, so in the MBOPC correction, Figure 2B ​In the prior art, the corner-to-corner distance of the pattern 204 is close to the minimum resolution of the mask plate, which causes the edges of the pattern 204 to stop moving and the iteration loop to stop, and the area of the pattern 204 is insufficient. When the area of the pattern 204 is insufficient, the area of the exposure pattern corresponding to the pattern 204 formed by exposure using the mask plate layer 203 is also reduced. Figure 2B In the prior art, the contour pattern 205 is a simulation pattern of the exposure pattern of the pattern 204, and it can be seen that the area of the contour pattern 205 is small, and Figure 2A In the prior art, the target pattern, i.e., the pattern 202, is greatly different from the target value, and off-target defects occur. SUMMARY

[0013] The technical problem to be solved by the present application is to provide an OPC correction method that can solve the problem of violating the MRC in OPC correction and prevent the OPC correction pattern from deviating from the target value.

[0014] To solve the above technical problem, the present application provides an OPC correction method, characterized in that it comprises the following steps:

[0015] Step 1: providing an initial target layer and setting a minimum resolution size of a mask plate.

[0016] Step 2: selecting a first pattern that will violate the mask rule check in subsequent model-based OPC (MBOPC) correction from the initial target layer according to the minimum resolution size of the mask plate.

[0017] Step 3: performing a split processing on the first pattern, the split processing divides each edge of the first pattern into a plurality of segments, each segment of each edge includes an angle segment and an intermediate segment, one vertex of the angle segment is a vertex of the edge, and the other vertex of the angle segment is a vertex of the adjacent intermediate segment; and the initial target layer after the split processing is a second target layer.

[0018] Step 4: performing model-based OPC correction based on the second target layer and obtaining a mask plate layer, the model-based OPC correction comprises a plurality of iteration loop operations, in each iteration loop operation, the angle segment and the intermediate segment of each edge of the first pattern are corrected separately, the size of the vertex of the first pattern in the mask plate layer is controlled by the angle segment, and the area of the first pattern in the mask plate layer is controlled by the intermediate segment.

[0019] Further improvement is that after step 4, it further comprises:

[0020] Step five, performing a mask rule check on the mask plate layer.

[0021] Further improvement is that in step one, the initial target layer is obtained by rule based OPC correction on the initial layout.

[0022] Further improvement is that the mask plate minimum resolution size includes a critical dimension minimum resolution value and a pitch minimum resolution value.

[0023] Further improvement is that in step five, the mask rule check passes when the critical dimension of the first pattern in the mask plate layer is greater than the critical dimension minimum resolution value and the pitch is greater than the pitch minimum resolution value.

[0024] Further improvement is that the first pattern includes a square hole pattern.

[0025] Further improvement is that the square hole pattern includes a via layer pattern.

[0026] Further improvement is that in the initial target layer, the array structure formed by the arrangement of the first patterns is a dense staggered arrangement structure, in which the diagonal lines of the first patterns are aligned and periodically arranged, the minimum pitch of the first patterns is the distance between the adjacent top corners of two adjacent first patterns, and the step of the first patterns is the sum of the length of the diagonal line of the first patterns and the minimum pitch.

[0027] Further improvement is that the mask plate minimum resolution size is determined by a process node and a mask plate manufacturing capability.

[0028] Further improvement is that the process node is below 14nm, and both the critical dimension minimum resolution value and the pitch minimum resolution value are 18nm or both are 12nm.

[0029] Further improvement is that the minimum pitch of the first patterns is less than 20nm, and the step of the first patterns is less than 115nm.

[0030] Further improvement is that the target value of the side length of the first patterns is 68nm, and the corner segment size in the segment processing is 5-20nm.

[0031] Further improvement is that in step three, the segment processing divides each side of the first patterns into three segments, including two corner segments and one middle segment.

[0032] In prior art, the edges of the first pattern are not segmented before MBOPC, so when the corner-to-corner distance of the first pattern is stopped in iteration loop operation in MBOPC due to violation of MRC, the area of the first pattern remains small, and finally the simulation contour of the exposure pattern after MBOPC deviates from the target value (off target); the present application provides an initial target layer, and does not directly perform model-based OPC correction based on the initial target layer, but according to the features of the patterns in the initial target layer, the first pattern that will violate MRC in model-based OPC correction is selected in advance, then the selected patterns are segmented and a second target layer is formed, and then MBOPC is performed, after the segmentation of the edges of the first pattern, the corner segments and the middle segments can be corrected separately in each iteration loop operation in MBOPC, so that the top corner size and the middle region size of the first pattern can be adjusted separately, and through the adjustment of the top corner size of the first pattern, the corner-to-corner distance of the first pattern can be adjusted without violating MRC; through the adjustment of the size of the middle region, the area of the first pattern can be adjusted, so that the area of the first pattern can be increased when the corner-to-corner distance of the first pattern meets the requirements of MRC, and the increase of the area of the first pattern after MBOPC can make the contour on target, so that the pattern after real exposure is also on target.

[0033] The present application is particularly suitable for OPC correction of square patterns with Dense Stagger Pattern structure, such as via layer patterns, and can realize the size of square patterns on the mask layer, and can make the simulation contour of the exposure pattern not only on target, but also meet the requirements of process window (PW) and mask error enhancement factor (MEEF) under process variation (pvband) for mass production without violating the mask MRC. BRIEF DESCRIPTION OF DRAWINGS

[0034] The present application will be further described in detail below in combination with the drawings and specific embodiments:

[0035] Figure 1 is a flowchart of the prior OPC correction method;

[0036] Figure 2A is a schematic diagram of the pattern structure of the target layer in the prior OPC correction method;

[0037] Figure 2B is a schematic diagram of the pattern structure of the mask layer formed by the prior OPC correction method;

[0038] Figure 3 is a flowchart of the OPC correction method of the embodiment of the present application;

[0039] Figure 4A is a schematic diagram of the pattern structure of the initial target layer in the OPC correction method of the embodiment of the present application;

[0040] Figure 4B is a schematic diagram of the pattern structure of the second target layer in the OPC correction method of the embodiment of the present application;

[0041] Figure 4C is a schematic diagram of the pattern structure of the mask layer formed by the OPC correction method of the embodiment of the present application;

[0042] Figure 4D is a schematic diagram of the profile pattern formed by the exposure pattern simulation using the mask layer of Figure 4C ;

[0043] Figure 5A is a simulation diagram of the mask layer and the profile pattern formed by the prior OPC correction method;

[0044] Figure 5B is a simulation diagram of the mask layer and the profile pattern formed by the OPC correction method of the embodiment of the present application when the corner segmentation size is 5 nm;

[0045] Figure 5C is a simulation diagram of the mask layer and the profile pattern formed by the OPC correction method of the embodiment of the present application when the corner segmentation size is 10 nm;

[0046] Figure 5D is a simulation diagram of the mask layer and the profile pattern formed by the OPC correction method of the embodiment of the present application when the corner segmentation size is 15 nm;

[0047] Figure 5E is a simulation diagram of the mask layer and the profile pattern formed by the OPC correction method of the embodiment of the present application when the corner segmentation size is 20 nm. DETAILED DESCRIPTION

[0048] As shown in Figure 3 , it is a flow chart of the OPC correction method of the embodiment of the present application; as shown in Figure 4A , it is a schematic diagram of the pattern structure of the initial target layer 301 in the OPC correction method of the embodiment of the present application; as shown in Figure 4B , it is a schematic diagram of the pattern structure of the second target layer 301a in the OPC correction method of the embodiment of the present application; as shown in Figure 4C , it is a schematic diagram of the pattern structure of the mask layer 304 formed by the OPC correction method of the embodiment of the present application; as shown in Figure 4D , it is a schematic diagram of the profile pattern formed by the exposure pattern simulation using the mask layer of Figure 4CFig. 1 is a schematic diagram of a profile pattern formed by exposing a mask plate layer 304 of the present application according to an embodiment of the present application; the OPC correction method of the present application embodiment includes the following steps:

[0049] Step one, as shown in Fig. 1, an initial target layer 301 is provided and a mask plate minimum resolution size is set. Figure 4A

[0050] In the present application embodiment, the initial target layer 301 is obtained by performing rule-based OPC correction on an initial layout.

[0051] The mask plate minimum resolution size includes a critical dimension minimum resolution value and a pitch minimum resolution value.

[0052] The mask plate minimum resolution size is determined by a process node and a mask plate manufacturing capability.

[0053] In some embodiments, the process node is below 14 nm, and both the critical dimension minimum resolution value and the pitch minimum resolution value are 18 nm or both are 12 nm.

[0054] Step two, according to the mask plate minimum resolution size, a first pattern 302 in the initial target layer 301 that will violate mask rule checking in subsequent model-based OPC correction is selected. Figure 4A Only the first pattern 302 in the initial target layer 301 is shown in Fig. 1. Other types of patterns in the initial target layer 301 are not shown in Fig. 1. Figure 4A

[0055] In the present application embodiment, the first pattern 302 includes a square hole pattern.

[0056] In some preferred embodiments, the square hole pattern includes a via layer pattern.

[0057] In the initial target layer 301, an array structure formed by arranging each first pattern 302 is a dense staggered arrangement structure, in which the diagonal lines of each first pattern 302 are aligned and periodically arranged, the minimum pitch of the first pattern 302 is the distance between adjacent top corners of two adjacent first patterns 302, i.e. the corner-to-corner pitch d301, and the step of the first pattern 302 is the sum of the length of the diagonal line of the first pattern 302 and the minimum pitch.

[0058] In some embodiments, the minimum pitch of the first pattern 302 is less than 20 nm, and the step of the first pattern 302 is less than 115 nm.

[0059] Step three, as shown in Fig. 2, a model-based OPC correction is performed on the first pattern 302 in the initial target layer 301 to obtain a first pattern 302' in a target layer 301'. Figure 4B ​​As shown, the first graph 302 is subjected to segmentation processing, which divides each edge of the first graph 302 into a plurality of segments, each segment of each edge including an angle segment 303a and an intermediate segment 303b, one vertex of the angle segment 303a being a vertex of the edge, and the other vertex of the angle segment 303a being a vertex of the adjacent intermediate segment 303b. Figure 4B The broken line between the angle segment 303a and the intermediate segment 303b in the figure is only for visually distinguishing the angle segment 303a and the intermediate segment 303b, and in fact the angle segment 303a and the intermediate segment 303b are connected together.

[0060] The initial target layer 301 after the segmentation processing is a second target layer 301a.

[0061] In the embodiment of the application, the segmentation processing divides each edge of the first graph 302 into three segments, including two angle segments 303a and one intermediate segment 303b.

[0062] In some embodiments, the target value of the edge length of the first graph 302 is 68 nm, and the size of the angle segment 303a in the segmentation processing is 5-20 nm.

[0063] Step four, as shown in Figure 4C Based on the second target layer 301a, model-based OPC correction is performed to obtain a mask plate layer 304, the model-based OPC correction including a plurality of iterative loop operations, in each of the iterative loop operations, the angle segment 303a and the intermediate segment 303b of each edge of the first graph 302 are corrected separately, the size of the top corner of the first graph 305 in the mask plate layer 304 is controlled by the angle segment 303a, and the area of the first graph 305 in the mask plate layer 304 is controlled by the intermediate segment 303b. Figure 4C In the figure, the first graph is represented by a mark 305, Figure 4C In the figure, the first graph 305 is formed by Figure 4A In the figure, the first graph 302 is formed by MBOPC correction. As shown in Figure 4C As shown in the figure, the separate correction of the angle segment 303a and the intermediate segment 303b of each edge of the first graph 302 can avoid the need for the entire movement of each edge of the first graph 302 in the MBOPC correction process in the prior art, and instead enables segmented movement, so that the size of the movement of each edge of the first graph 302 at the angle segment 303a and the size of the movement at the intermediate segment 303b can be different. Figure 4ABy comparing the first graphic 302 located in the initial target layer 301, it can be seen that... Figure 4C After being corrected by MBOPC, the edges of the first graphic 305 are no longer flat, and a convex graphic 305a will be formed at the corresponding position of the middle segment 303b. The convex graphic 305a will increase the area of ​​the first graphic 305.

[0064] Step 5: Perform a mask rule check on the mask template layer 304.

[0065] In step five, the mask rule check passes when the key dimension of the first graphic 305 in the mask layer 304 is greater than the minimum resolution value of the key dimension and the spacing is greater than the minimum resolution value of the spacing.

[0066] like Figure 4C As shown, the key dimension of the first graphic 305 is the minimum width, and the spacing of the first graphic 305 is the diagonal spacing d302.

[0067] The mask rule check is inserted into each iterative loop operation of the model-based OPC correction. Even when the diagonal spacing d302 is close to the minimum resolution value, the edge at the intermediate segment 303b can still be corrected, i.e., moved. This avoids the defect in existing methods where the entire edge cannot be moved when the diagonal spacing d302 is close to the minimum resolution value. Moving the edge at the intermediate segment 303b increases the area of ​​the first shape 305. Therefore, this embodiment of the invention can increase the area of ​​the first shape 305 while ensuring that the diagonal spacing d302 does not violate the MRC.

[0068] Increasing the area of ​​the first pattern 305 increases the area of ​​the corresponding exposure pattern, thereby making the exposure pattern meet the target value, i.e., on-target. Figure 4D As shown, the contour pattern 306 is a schematic diagram of a simulation of the exposure pattern corresponding to the first pattern 305, and Figure 2B Compared to the outline pattern 205, the outline pattern 306 obtained in this embodiment of the invention has a larger area and is closer to the corresponding target pattern, thus eliminating... Figure 2B The corresponding off-target defect.

[0069] In existing technologies, because the edges of the first graphic 302 are not segmented before MBOPC, when the diagonal spacing of the first graphic 302 is violated by the MRC and the iterative loop stops during MBOPC, the area of ​​the first graphic 302 remains small. This ultimately causes the simulated outline of the first graphic 302 after MBOPC, i.e., the exposure graphic, to deviate from the target value. In this embodiment of the invention, after providing the initial target layer 301, it does not directly perform model-based OPC correction based on the initial target layer 301. Instead, it pre-selects first graphics 302 that will violate the MRC in model-based OPC correction based on the characteristics of the graphics in the initial target layer 301. Then, it segments these selected graphics to form the second target layer 30. 1a. After performing MBOPC, the edges of the first graphic 302 are segmented. The corner segments 303a and the middle segments 303b can be corrected separately in each iteration of MBOPC. This allows for separate adjustment of the top corner size and the size of the middle area of ​​the first graphic 302. By adjusting the top corner size of the first graphic 302, the diagonal spacing of the first graphic 302 can be made to meet the MRC. By adjusting the size of the middle area, the area of ​​the first graphic 302 can be adjusted. This increases the area of ​​the first graphic 302 while ensuring that the diagonal spacing of the first graphic 302 meets the MRC requirements. The increased area of ​​the first graphic 302 after MBOPC makes the outline graphic conform to the target value, so that the graphic after actual exposure will also be on target.

[0070] The embodiments of the present invention are particularly applicable to OPC correction of square patterns with dense, interlaced structures, such as through-hole layer patterns. They can realize the size of the square pattern in the mask layer 304, and without violating the mask MRC, the simulated Contour corresponding to the exposed pattern is not only on-target, but also meets the mass production requirements for pvband and meef under PW conditions.

[0071] like Figure 5A The image shown is a simulation diagram of a mask layer and a contour graphic formed using the existing OPC correction method; the mask layer 401a includes a mask graphic 403a that has been corrected by MBOPC on the target graphic 402, and the target graphic 402 corresponds to... Figure 2A In the figure 202, the mask figure 403a corresponds to Figure 2B The figure 204 in the figure. The existing method is also equivalent to the method described in the embodiment of the present invention, where the size of the corner segment is 0nm, i.e., Split 0nm. The contour figure 404a is a simulation diagram of the exposure figure corresponding to the mask figure 403a. It can be seen that the contour figure 404a and the target figure 402 are quite different and are in an off-target state.

[0072] Figure 5Bis a simulation diagram of a mask plate layer and a contour pattern formed by the method of the embodiment of the present application when the angle segmentation size is 5nm in the OPC correction method. The mask plate layer 401b includes a mask pattern 403b obtained by performing MBOPC correction on a target pattern 402 corresponding to the pattern 302 in Figure 4A . Here, the target pattern 402 is the same as the target pattern 402 in Figure 5A , that is, different MBOPC correction is performed on the corresponding target pattern 402 to compare the correction results. The contour pattern 404b is a simulation diagram of an exposure pattern corresponding to the mask pattern 403b; Figure 5B . The size of the angle segmentation in the simulation diagram is 5nm, that is, Split 5nm. It can be seen that the contour pattern 404b is close to the target pattern 402 and is in an on-target state.

[0073] Figure 5C is a simulation diagram of a mask plate layer and a contour pattern formed by the method of the embodiment of the present application when the angle segmentation size is 10nm in the OPC correction method. The mask plate layer 401c includes a mask pattern 403c obtained by performing MBOPC correction on a target pattern 402 corresponding to the pattern 302 in Figure 4A . The contour pattern 404c is a simulation diagram of an exposure pattern corresponding to the mask pattern 403c; Figure 5C . The size of the angle segmentation in the simulation diagram is 10nm, that is, Split 10nm. It can be seen that the contour pattern 404c is close to the target pattern 402 and is in an on-target state.

[0074] Figure 5D is a simulation diagram of a mask plate layer and a contour pattern formed by the method of the embodiment of the present application when the angle segmentation size is 15nm in the OPC correction method. The mask plate layer 401d includes a mask pattern 403d obtained by performing MBOPC correction on a target pattern 402 corresponding to the pattern 302 in Figure 4A . The contour pattern 404d is a simulation diagram of an exposure pattern corresponding to the mask pattern 403d; Figure 5D . The size of the angle segmentation in the simulation diagram is 15nm, that is, Split 15nm. It can be seen that the contour pattern 404d is close to the target pattern 402 and is in an on-target state.

[0075] Figure 5E is a simulation diagram of a mask plate layer and a contour pattern formed by the method of the embodiment of the present application when the angle segmentation size is 20nm in the OPC correction method. The mask plate layer 401e includes a mask pattern 403e obtained by performing MBOPC correction on a target pattern 402 corresponding to the pattern 302 in Figure 4A . The contour pattern 404e is a simulation diagram of an exposure pattern corresponding to the mask pattern 403e;Figure 5E The size of the angular segment in the Split 20nm. It can be seen that the contour pattern 404e and the target pattern 402 are close, in an on-target state.

[0076] Figures 5A to 5E The corresponding test data is shown in Table 1.

[0077] Table 1

[0078]

[0079] The size of the angular segment in the Split 20nm. It can be seen that the contour pattern 404e and the target pattern 402 are close, in an on-target state.

[0080] In the test results, Target represents the critical dimension of the target pattern 402;

[0081] Mask Space represents the pitch of the mask pattern;

[0082] Mask CD represents the critical dimension of the mask pattern;

[0083] Mask Area represents the area of the mask pattern.

[0084] Contour represents the critical dimension of the contour pattern.

[0085] PW_minCD represents the minimum critical dimension of the process window.

[0086] Pvband represents the process variation band.

[0087] Meef represents the mask error enhancement factor.

[0088] It can be seen that, under the condition that the Mask Space does not violate the MRC, the Mask CD corresponding to the embodiment of the present application will increase, the Mask Area will increase, the critical dimension of the Contour will increase, the PW_minCD will increase, the Pvband will decrease, and the Meef will decrease, which all indicate that the OPC correction result of the method of the embodiment of the present application is better.

[0089] The above has described the present application in detail through specific embodiments, but these do not constitute a limitation on the present application. Those skilled in the art can also make many modifications and improvements without departing from the principles of the present application, and these should also be considered as the protection scope of the present application.

Claims

1. An OPC correction method characterized by, The method comprises the following steps: Step one, providing an initial target layer and setting a mask plate minimum resolution size; Step two, selecting a first pattern which will violate mask rule check in subsequent model-based OPC correction from the initial target layer according to the mask plate minimum resolution size; Step three, segmenting the first pattern, the segmentation divides each side of the first pattern into multiple segments, each segment of each side comprises an angle segment and an intermediate segment, one vertex of the angle segment is a vertex of the side, and the other vertex of the angle segment is a vertex of the adjacent intermediate segment; the initial target layer after the segmentation is a second target layer; Step four, performing model-based OPC correction based on the second target layer and obtaining a mask plate layer, the model-based OPC correction comprises multiple iteration loop operations, in each iteration loop operation, the angle segment and the intermediate segment of each side of the first pattern are corrected separately, the corner size of the first pattern in the mask plate layer is controlled by the angle segment, and the area of the first pattern in the mask plate layer is controlled by the intermediate segment.

2. The OPC correction method of claim 1, wherein: After step four, the method further comprises: Step five, performing mask rule check on the mask plate layer.

3. The OPC correction method of claim 1, wherein: In step one, the initial target layer is obtained by performing rule-based OPC correction on an initial layout.

4. The OPC correction method of claim 2, wherein: The mask plate minimum resolution size comprises a critical dimension minimum resolution value and a pitch minimum resolution value.

5. The OPC correction method of claim 4, wherein: In step five, the mask rule check passes when the critical dimension of the first pattern in the mask plate layer is greater than the critical dimension minimum resolution value and the pitch is greater than the pitch minimum resolution value.

6. The OPC correction method of claim 4, wherein: The first pattern comprises a square hole pattern.

7. The OPC correction method of claim 6, wherein: The square hole pattern comprises a via layer pattern.

8. The OPC correction method of claim 7, wherein: In the initial target layer, an array structure formed by the first patterns is a dense staggered arrangement structure, in the dense staggered arrangement structure, diagonal lines of the first patterns are aligned and periodically arranged, a minimum pitch of the first pattern is a distance between adjacent top corners of two adjacent first patterns, and a step of the first pattern is a sum of a length of the diagonal line of the first pattern and the minimum pitch.

9. The OPC correction method of claim 8, wherein: The mask plate minimum resolution size is determined by a process node and a mask plate manufacturing capability.

10. The OPC correction method of claim 9, wherein: The process node is below 14 nm, and the critical dimension minimum resolution value and the pitch minimum resolution value are both 18 nm or both 12 nm.

11. The OPC correction method of claim 10, wherein: The minimum pitch of the first pattern is less than 20 nm, and the step of the first pattern is less than 115 nm.

12. The OPC correction method of claim 10, wherein: A target value of a side length of the first pattern is 68 nm, and an angle segment size in the segmentation is 5-20 nm.

13. The OPC correction method of claim 8, wherein: In step three, the segmentation divides each side of the first pattern into three segments, the three segments comprise two angle segments and one intermediate segment.

Images