JP2008191364A - Design method of mask pattern - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a design method for mask pattern in which an auxiliary pattern for improving a focal depth of an isolated pattern is arranged, with respect to a mask of performing exposure by using a light source having a direction dependence for resolution. <P>SOLUTION: The method of designing a mask pattern used upon exposure having the direction dependency for resolution by using an eight-point light source includes: a step of producing a dummy data 11 of a polygonal shape having two edges which form vortices in the neighborhood of respective vortices of an isolated rectangular pattern 1 and an edge inclined in 45° directions from the two edges; and a step of arranging the auxiliary pattern 2 which adjoins the vortices and has a prescribed positional relation with the dummy data 11. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a method for designing a mask pattern, and more particularly to a method for arranging an auxiliary pattern preferably used for exposure of a semiconductor wafer.

  In the exposure process of the current semiconductor device, a sufficient depth of focus can be obtained by applying an oblique incidence illumination method to a dense pattern such as a line and space pattern. For example, a periodic pattern such as a gate of a DRAM device or a wiring pattern can be stably formed even with a very small dimension. The oblique incidence illumination method is a method of illuminating the mask with oblique incidence by cutting a vertical incidence component of the mask illumination light. In the normal illumination method, three beams including the 0th-order diffracted light and ± 1st-order diffracted light from the mask pattern are collected by the projection lens to obtain an image of three-beam interference. One of the first-order diffracted lights is discarded, and an image is formed with two light beams including the zero-order light and one of the ± first-order diffracted lights to obtain an image of two-beam interference.

  When the imaging states of both the three-beam interference and the two-beam interference are compared with the best focus, in the two-speed interference, the contrast is lowered by discarding the other of the ± first-order diffracted lights. However, considering the incident angle on the semiconductor substrate, which is the imaging plane, the imaging of the two-beam interference is ½ of the three-beam interference. For this reason, there is less blurring of the image when the focus is shifted, and a light intensity distribution sufficient for forming a resist pattern can be obtained in a wide focus range. Restriction of the direction and angle of light for illuminating the mask is realized by arranging a metal diaphragm immediately after the shape of the secondary light source formed by the fly-eye lens. This is because the original light source is a mercury lamp or an excimer laser device, but if this light source is used as it is, the mask cannot be illuminated with uniform intensity, and there are several hundred fly eye lenses. A set of light sources is formed to illuminate the mask.

  When the illumination optical system is viewed from the mask side, only a collection of point light sources formed by fly-eye lenses can be seen, and the shape of the collection of point light sources determines the exposure characteristics. Therefore, an original light source such as a mercury lamp is called a primary light source, and a set of point light sources formed by fly-eye lenses is called a secondary light source (effective light source). The light at the center of the secondary light source enters the mask perpendicularly, and the light from the outside of the secondary light source enters the mask obliquely. Therefore, in the oblique incident illumination method, the center of the secondary light source is shielded, or a stop that opens only outside is used. Due to the aperture shape (secondary light source shape), the exposure characteristics on the wafer change, and a secondary light source having a shape as shown in FIG. 22 has been proposed.

  The illumination shown in FIG. 22A is called two-point illumination, and has an effect of improving the depth of focus of a one-directional pattern (here, the horizontal direction). Illumination shown in FIG. 4B is called four-point illumination, and has an effect of improving the depth of focus of the vertical and horizontal two-way patterns. In the four-point illumination, if the light source is in the direction of 45/135 degrees, the exposure characteristics of the vertical and horizontal patterns are the same, and the vertical and horizontal exposure characteristics are different when deviating from that. For example, if the pitch in the vertical direction is relatively dense and the pitch in the horizontal direction is relatively sparse, the vertical direction opens the outermost periphery of the secondary light source in the example shown in FIG. The secondary light source 3 is shaped so as to open the inside of the outermost periphery. The illumination shown in FIG. 3C is called annular illumination, and this illumination has a high general-purpose quality because the exposure characteristics have no pattern direction dependency. For this reason, in general, when annular illumination is applied and a sufficient depth of focus cannot be obtained, application of four-point illumination or two-point illumination with a limited pattern to be applied is considered.

  An illumination method that improves the exposure characteristics using polarized light instead of the shape of the secondary light source has also been proposed. For example, Patent Document 1 discloses a technique for improving exposure characteristics of a specific direction pattern by polarizing exposure light so that TM polarization is obtained in a pattern in a specific direction even when annular illumination is used. Yes.

  It is also known that the depth of focus can be further increased by using a halftone phase shift mask. The depth of focus refers to a focus range where an effective resist pattern can be obtained. The halftone phase shift mask is a pattern on the mask which is a light shielding area formed in a semitransparent area, leaks about 2 to 20% of light, and between the leaked light and the light in the surrounding transparent area, This is a phase shift mask with the phase inverted by 180 degrees. In the case of a line-and-space pattern in which diffracted light is generated, the balance between the 0th-order diffracted light and the + 1st-order (or -1st-order) diffracted light is improved by using a halftone mask and the oblique incidence illumination method. , Improve the contrast.

  However, an isolated pattern in which no diffracted light is generated has little effect by the above-described modified illumination method, and the depth of focus does not increase so much. On the other hand, lowering the NA or reducing σ is more effective for increasing the depth of focus of the isolated pattern. σ is the ratio of the size of the illumination lens to the size of the pupil plane of the light source. That is, σ = NA of the illumination lens / NA of the projection lens. When the size of the illumination lens is the same as the pupil plane of the projection lens, σ is 1. Lowering the NA of the illumination optical system means illuminating the mask only with light close to the vertical component. Even when a halftone phase shift mask is used, the depth of focus is improved with small σ illumination. Any condition for increasing the depth of focus of these isolated patterns results in lowering the resolution of the dense pattern. For this reason, it has been difficult to achieve both exposure characteristics with dense fine patterns and isolated patterns.

  In view of this, a technique called an auxiliary pattern, which uses a fine pattern that does not resolve itself, has been studied as a method for achieving both a dense pattern and an isolated pattern in depth of focus. For example, Patent Document 2 discloses the auxiliary pattern. In Patent Document 2, when the mask is illuminated with obliquely incident light, the depth of focus of the pattern is improved by arranging an auxiliary pattern having a dimension less than or equal to the limit resolution near the pattern in accordance with the angle and direction of the obliquely incident light. Adopt a method. By using the mask on which the auxiliary pattern is arranged under oblique incidence illumination conditions, the imaging state of two-beam interference is approached and the depth of focus is expanded.

In the arrangement of the auxiliary pattern, the position and size influence the depth of focus of the device pattern. The optimum value of the interval between the auxiliary pattern and the main pattern varies depending on the dimensions and optical conditions used, but there is an optimum value in a range of about 1.5 times the limit resolution of the optical conditions. Further, the larger the dimension of the auxiliary pattern, the greater the effect of expanding the focal depth of the main pattern. However, if the dimension is too large, the auxiliary pattern itself is transferred onto the semiconductor substrate. The size of the auxiliary pattern is set slightly smaller with a margin than the limit size that is not transferred onto the wafer. A rule-based method and a model-based method have been proposed as auxiliary pattern placement methods. The rule-based method is a method of creating a mask pattern design method table in advance according to the interval between a target pattern and an adjacent pattern. After arranging the auxiliary patterns according to the rules for all the target patterns, processing for correcting inconveniences of the auxiliary patterns (insufficient spacing between the auxiliary pattern and the main pattern or insufficient spacing between the auxiliary patterns) is performed. The model-based method is a method of generating an auxiliary pattern when a contrast or depth of focus simulation is performed and the value is insufficient.
JP-A-7-183201 JP-A-4-268714

  The rule-based method has an advantage that the auxiliary pattern is generated and confirmed at high speed. In this method, since the arrangement rule is already made, the result can be easily confirmed by using the design rule check (DRC). The model-based method has an advantage that auxiliary patterns can be arranged so that exposure characteristics can be sufficiently obtained, but there is a problem that it takes a long processing time because a simulation is required for generation and confirmation of the result.

  In general-purpose memories such as DRAM devices, miniaturization is important for cost reduction, and the memory cell array is reduced to a resolution limit pitch of an exposure device. Therefore, the illumination condition of the exposure apparatus has been set to oblique incidence illumination specialized for the memory cell array. However, in this case, there is a problem that the depth of focus of a pattern having a loose pitch becomes narrow except for a miniaturized memory cell array.

  In the present invention, a sufficiently large depth of focus can be obtained even in the case of using a direction-dependent oblique incidence illumination method or polarized illumination method for improving the depth of focus of the most densely packed pattern, or in a coarsely arranged pattern or an isolated pattern. It is an object of the present invention to provide a mask pattern design method improved as described above.

In order to achieve the above object, a mask pattern design method of the present invention is a mask pattern design method used for exposure using illumination conditions having direction dependency on resolution characteristics,
Generating dummy pattern data in the vicinity of each vertex of the rectangular pattern based on the direction dependency of the resolution characteristic of the illumination condition;
An auxiliary pattern having a predetermined positional relationship with the dummy pattern is disposed adjacent to the apex.

  In the mask pattern design method of the present invention, the step of arranging the auxiliary pattern may be generated based on a predetermined rule-based method or a model-based method.

  Further, the direction dependency of the resolution characteristic may depend at least on the polarization or shape of the light source.

  When the illumination condition is to increase the depth of focus in a direction having an inclination of 45 degrees with each side of the rectangular pattern, in the step of generating data of the dummy pattern, two sides forming a vertex Data of a dummy pattern having a side inclined in the direction of 45 degrees from the angle may be generated. Alternatively, if the illumination condition is to increase the depth of focus in a direction orthogonal to each side of the rectangular pattern and in a direction having a 45-degree inclination with each side, the dummy pattern data is generated. In this step, data of an octagonal dummy pattern having two sides forming a vertex and sides inclined in the direction of 45 degrees from the two sides may be generated.

  The rectangular pattern may be square or rectangular.

  When the rectangular pattern is square, the dummy pattern may be a regular octagonal dummy pattern.

  When the rectangular pattern is a rectangular pattern, both the rectangular rectangular pattern and the dummy pattern generated from the rectangular pattern are corrected so as to reduce the width and length. You may go.

  Data of a dummy pattern of the auxiliary pattern may be generated, and another auxiliary pattern may be arranged adjacent to the auxiliary pattern based on the data of the dummy pattern.

  According to the mask pattern design method of the present invention, for example, when an optimum illumination condition is adopted for improving the depth of focus of a densely arranged rectangular pattern, an isolated pattern or a coarsely arranged pattern formed on the same mask pattern is used. The pattern also has an effect that an auxiliary pattern for improving the depth of focus can be arranged.

  The mask pattern targeted by the mask pattern design method of the present invention is not particularly limited. For example, a mask pattern used for exposure of a DRAM device can be targeted.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. First, as a first embodiment of the present invention, a method for arranging an auxiliary pattern on a mask pattern for forming an isolated hole of, for example, 80 nm will be described. The mask is a halftone phase shift mask (transmittance = 6%). An exposure apparatus using this mask has a reduction ratio of 4 times, uses an ArF laser (wavelength λ = 193 nm), has a numerical aperture (NA) = 0.92, and is a step-and-scan exposure apparatus. It will be explained as being. Further, the illumination condition is 8-point illumination shown in FIG. Each light source 3 is arranged at 0/45/90/135 degrees and its opposite direction (hereinafter referred to as 0/45/90/135 degree direction including the opposite direction) when viewed from the center of the pupil plane of the projection lens. The distance from the center of the pupil plane is σc = 0.85 as a ratio to the pupil plane radius rd, and the radius is similarly σr = 0.15 as the ratio to the radius rd of the pupil plane. . Furthermore, each light source 3 is polarized in a direction orthogonal to the radial direction of the pupil plane. The following dimensions show values on the wafer unless otherwise specified, and the values on the mask are four times that.

  In the 8-point illumination shown in FIG. 1, four light sources 3 that are arranged in an X shape on the drawing have an effect of improving the depth of focus with respect to a vertical and horizontal (xy direction) repetitive pattern, and are arranged in a cross shape on the drawing. The other four light sources have a large depth-of-focus effect of repeating patterns in an oblique direction of 45/135 degrees when viewed from the x-axis direction. Therefore, in order to improve the focal depth of the isolated hole pattern, it is necessary to arrange the auxiliary pattern 2 in the direction of 45 degrees vertically and horizontally. Therefore, as shown in FIG. 2, octagonal auxiliary pattern generation dummy data 11 is generated in a square isolated rectangular pattern (hole pattern) 1. A bias was applied to the hole pattern, and the dimension W was set to 100 nm. The dimensions of each side of the dummy pattern data (hereinafter also referred to as dummy data) 11 were set to A = 42 nm and B = 41 nm. Here, a regular octagon was formed so as to ride on a 1 nm grid, and the dimension of the oblique side was rounded off to 1 nm or less. FIG. 3 shows the auxiliary pattern 2. The dimension w of the auxiliary pattern 2 is 80 nm, and the dummy data 21 for arrangement is also generated in the auxiliary pattern 2. The dimensions of each side of the dummy data 21 are approximately square with a = 34 nm and b = 33 nm.

  As shown in FIG. 4, the auxiliary patterns are arranged by arranging two auxiliary patterns having a center on a straight line extending in the direction perpendicular to each side from the midpoint of each of the eight sides of the dummy data 11 in the hole pattern 1. To do. In the drawing, only the auxiliary pattern corresponding to one side is shown. The interval s between the dummy data 11 of the hole pattern 1 and the dummy data 21 of the auxiliary pattern 2 is set to 80 nm, and the interval s between the two auxiliary patterns 2 is also set to 80 nm. Again, a 1 nm grid is used, the side angle is fixed at 45/135 degrees, and the arrangement of the auxiliary pattern is finely adjusted so that the vertex of the pattern is on the grid. FIG. 5 shows a mask pattern in which the placement of the auxiliary pattern 2 is completed for the isolated hole pattern 1. The auxiliary pattern 2 is arranged corresponding to the eight sides of the substantially regular octagonal dummy data 11 generated from the hole pattern 1. Dummy data 21 is also generated from each auxiliary pattern 2 and used to arrange adjacent auxiliary patterns.

  FIG. 6 shows the light intensity distribution on the wafer when the illumination light source shown in FIG. 1 is used and the mask shown in FIG. 5 is used. 6 shows the light intensity distribution on the X axis of FIG. 5, and the position of 500 nm on the X axis is the center of the isolated hole pattern 1. The defocus is 0 nm. The parameter representing the quality of the light intensity distribution includes contrast, but here, NILS (Normalized Image Log-Slope) is used. Conventionally, the steepness of the light intensity distribution, which is called ILS, is obtained by dividing the light intensity gradient by the light intensity. Multiplying this by the size of the pattern, so that even patterns with different dimensions can be compared with the same pattern meter is called NILS. The definition of ILS is ILS = (1 / I) * (σ (I) / σx) = σ [ln (I)] / σx, where I is light intensity and x is position. The definition of NILS is NILS = pattern dimension xILS.

  Here, if the value of NILS is 1 or more, it is assumed that the pattern is resolved on the wafer. FIG. 7 shows the relationship between NILS and defocus obtained by calculating the light intensity distribution with a focus. In FIG. 7, in addition to complete polarization (polarization degree 1), the cases of polarization degree 0.5 and 0 (non-polarization) are also shown. It is shown that since the auxiliary pattern is appropriately arranged, the steepness of the light intensity distribution is increased by the polarized light, which is improved. When the degree of polarization is 1, the depth of focus is as wide as 0.5 μm.

  As a comparative example, FIG. 8 shows a mask generated by a conventional rule-based method. Here, the isolated hole refers to a hole pattern sufficiently separated so that the auxiliary pattern of the isolated hole and the auxiliary pattern of another isolated pattern do not interfere with each other. In FIG. 8, the generation rule of the auxiliary pattern 2 to be arranged for such an isolated hole pattern 1 is that the first one is arranged 80 nm away from the side of the isolated hole pattern 1 and the second one is arranged 220 nm away. It was decided. The width of the auxiliary pattern 2 was 40 nm for the first pattern and 50 nm for the second pattern. The length of the auxiliary pattern 2 was made longer than that of the isolated hole pattern 1 so that auxiliary patterns adjacent to each other at the corner portion were 10 nm apart. FIG. 9 shows the light intensity distribution at the best focus for the mask of FIG. The maximum intensity of the hole pattern center (x = 500 nm) is about 0.4, similar to the maximum intensity shown in FIG. 6 for the present invention. However, the light intensity distribution is slightly less than the curve in FIG. FIG. 10 shows the relationship between defocus and NILS obtained by calculating the light intensity with the focus shifted for the mask of FIG. 8, as in FIG. In the conventional auxiliary pattern arrangement of FIG. 8, the depth of focus is a little less than 300 nm at the time of non-polarized light, and the effect of improving the depth of focus by polarized light cannot be obtained.

  From the above results, when the configuration of the first embodiment of the present invention is adopted, the illumination using the light source shown in FIG. 1 is assisted at positions in the eight directions from the center of the isolated pattern as shown in FIG. By arranging the pattern, it can be understood that the effect of improving the depth of focus can be obtained by exposure using polarized light.

  In FIG. 11, an octagonal dummy pattern is created from a square-shaped isolated hole pattern 1 having a side of 100 nm as a mask pattern according to the second embodiment of the present invention, and a rule base for the dummy data 11 is created. The mask pattern which generated the auxiliary pattern by the technique is shown. In this embodiment, the rectangular auxiliary pattern 2 in the oblique direction arranged in the 45 degree direction has an oblique side. With this mask, there is a problem that the mask drawing time increases due to the mask fabrication using the variable rectangle type mask drawing apparatus, but the advantage that the depth of focus equivalent to the previous embodiment can be ensured even by simple generation. have. FIG. 12 shows the light intensity distribution at the best focus of the mask of FIG. FIG. 13 shows the relationship between NILS and defocus obtained from the light intensity distribution. In the embodiment mask of FIG. 11, a depth of focus of slightly over 400 nm is obtained.

  Next, a description will be given of a method for arranging the auxiliary pattern in the case where a cross four-point illumination (no polarization) as shown in FIG. 14 is used as the illumination condition except for the illumination condition. In the four-point illumination, the distance from the pupil plane center to the center of each light source (σc) is 0.8, and the radius (σr) of each light source is 0.15 at a ratio where the radius of the pupil plane of the projection lens is 1. To do. The four-point illumination of the cross has an effect of improving the depth of focus of the repeated pattern in the 45/135 degree direction and an effect of increasing the limit resolution of the dense pattern in the vertical and horizontal directions. For this reason, the four-point illumination in the cross shape is selected when it is desired to improve the contrast of a narrow pitch pattern with a limit resolution, although the required depth of focus can be obtained only with respect to the vertical and horizontal patterns. For this reason, it is required to arrange an auxiliary pattern in the isolated pattern.

  FIG. 15 shows a state in which dummy data 11 for auxiliary pattern placement is generated for the isolated hole pattern 1. Here, since the auxiliary pattern is arranged in the 45/135 degree direction, the dummy data 11 is a square having sides in the 45/135 degree direction. The dimension of the hole pattern 1 was W = 90 nm, and the dimension of the dummy pattern was A = 90 nm.

  FIG. 16 shows a state in which the rectangular auxiliary pattern 2 is generated by the rule-based method with respect to the dummy data 11 of FIG. FIG. 17 shows the light intensity distribution at the best focus (distribution on the oblique dotted line in FIG. 16). FIG. 18 shows the relationship between NILS and defocus. By arranging the auxiliary pattern, a focal depth of 350 nm is obtained.

  As described above, by arranging the auxiliary pattern in the 45/135 degree direction with respect to the hole pattern 1 as shown in FIG. 16, the depth of focus of the isolated hole pattern when using the four-point illumination of the cross is increased. It can be secured sufficiently. As a comparative example, FIG. 20 shows the light intensity obtained by a mask pattern (FIG. 19) in which an auxiliary pattern is generated by a rule-based method from a conventional hole pattern with four-point illumination. FIG. 21 shows the relationship between NILS and defocus. When the auxiliary pattern 2 is generated from the hole pattern 1 by the conventional method, only a focal depth of less than 200 nm can be obtained.

  As a third embodiment of the present invention, a case where the light source of FIG. 1 is employed and an auxiliary pattern is arranged with respect to an isolated rectangular pattern will be described. FIG. 23 shows an example of generating dummy pattern data 11 and 41 from the square pattern 1 and the rectangular pattern 4. In any case, the length (a) of the cut side (short side) and the length of the oblique side (b) formed by cutting on the dummy data obtained by cutting at 45 degrees at the corner portion. And the dummy data are generated so that the lengths are substantially the same. In the square pattern 1, the dummy data 11 is regular octagonal data. In the rectangular pattern 4, the dummy data 41 is vertically long octagonal data as shown.

  24A to 24C show a procedure for generating the auxiliary pattern 5 to be applied to the rectangular pattern 4 described above. If the square pattern (hole pattern) and the rectangular pattern (slit pattern) are on the same mask and the width of the rectangular pattern is the same length as the four sides of the square pattern, The rectangular pattern has a larger width and a shorter length than the square pattern. Therefore, an OPC (Optical proximity correction) tool is applied to the dummy pattern 41 obtained in FIG. 23, and automatic correction is performed as shown in FIG. 24A to generate a dummy pattern 42 with OPC correction. In this automatic correction, the dummy pattern 41 generated previously is narrowed and lengthened.

  Next, as shown in FIG. 24B, the auxiliary pattern 5 is arranged at a distance s from the dummy pattern 42 subjected to OPC correction. In this example, the auxiliary patterns arranged in eight directions are connected together. Next, OPC correction is also applied to the rectangular pattern 4 to correct the rectangular pattern so that the width of the rectangular pattern is smaller than the width of the dummy pattern and the length is longer than the length of the dummy pattern. 43 is generated ((c) in the figure). In addition, as described in the square pattern, the auxiliary patterns can be arranged separately in eight directions. The reason why the auxiliary pattern 5 is arranged as shown in FIG. 24C is as follows.

  In the rectangular pattern 4, near the center of the long side, as shown in FIG. 25A, the auxiliary pattern can be arranged only in the horizontal direction (x-axis direction) indicated by the solid line, and the other direction indicated by the dotted line Then, since it is too close to the rectangular pattern 4, the auxiliary pattern cannot be arranged. Further, in the vicinity of the tip of the rectangular pattern 4, as shown in FIG. 25 (b), as shown by a solid line among the eight directions, it is in five directions of 0/45/90/135/180 degrees when viewed from the x-axis direction. Auxiliary patterns can be placed. Therefore, in the whole rectangular pattern, the central part of the long side in FIG. 25A and the vicinity of the tip in FIG. 25B are collected, and the auxiliary pattern 5 is arranged as shown in FIG.

  As a fourth embodiment, an example will be described in which illumination by a light source shown in FIG. 14 is used and auxiliary patterns are arranged with respect to the rectangular pattern shown in FIG. In this case, as shown in FIG. 26A, the auxiliary pattern is arranged only near the front end of the rectangular pattern in order to improve the depth of focus in the vertical and horizontal directions. In other words, in order to obtain a desired depth of focus, the auxiliary pattern is arranged only in two directions of 45/135 degrees when viewed from the x-axis direction among the eight directions at the tip of the rectangular pattern. Near the center of the long side of the rectangular pattern, other parts are adversely affected, and it is not preferable to arrange the auxiliary pattern. Accordingly, as shown in FIG. 26B, auxiliary patterns are arranged in the 45/135 degree direction.

A method for determining the distance s between the auxiliary pattern and the dummy pattern shown in FIG. 4 will be described. When the light source shown in FIG. 1 is used, the purpose of the auxiliary pattern to be arranged is to realize an imaging state of two-beam interference. Here, in the repetitive pattern of the pitch P, the condition for realizing the two-beam interference of the light in the range from the position of the radius σin to the position of the radius σout in proportion to the radius of the pupil plane is that λ is the wavelength and NA is the numerical aperture Then,
(Λ / (1 + σin)) · NA <P <(1 + σ) · NA (1)
It is. Here, in FIG.
σin = σc−σr = 0.8−0.15 = 0.65
σout = σc + σr = 0.8 + 0.15 = 0.95
It is.

When the wavelength λ = 193 nm and the numerical aperture NA = 0.92, the expression (1) is
127nm <P <215nm
It becomes. The optimum value of P is approximately at the center of the above range, and for example, the maximum value of the depth of focus is obtained at a pitch of 160 nm. It should be noted that the layout on the CAD and the design rule check for confirming the layout are easier to specify at the interval between the sides than the center of the pattern, and are specified at the interval between the sides.

The distance s between the isolated pattern and the auxiliary pattern is determined in the same manner as the pitch P of the repetitive pattern. When the auxiliary pattern is arranged in the horizontal direction of the long side of the rectangular pattern, the distance s between the isolated pattern and the auxiliary pattern is
s = P-{(W1 + W2) / 2}
Is obtained. The value of s is also used when arranging in the diagonal direction of the rectangle.

  As described above, the mask pattern designing method of the present invention can be suitably employed when using illumination conditions (polarized light and light source shape) whose direction characteristics are dependent on the resolution characteristics. In the auxiliary pattern arrangement, the auxiliary pattern arrangement direction is determined according to the condition of the light source to be used, and the auxiliary pattern is arranged by applying the rule-based method or the model-based method to the arrangement direction. Accordingly, the auxiliary pattern is arranged for the target isolated pattern or coarse arrangement pattern in accordance with the illumination condition used for improving the depth of the most densely packed pattern, and the focal depth of the target pattern is improved.

  In particular, in a semiconductor device using a minute dimension near the limit resolution of an exposure apparatus in a memory cell, such as a DRAM device, it is necessary to select an optimal illumination condition for the pattern in the cell. In this case, for a pattern other than the memory cell, measures are taken to ensure the depth of focus even under illumination conditions optimal for the memory cell. In the present invention, a sufficient depth of focus of an isolated pattern can be ensured even when illumination conditions having direction dependency on exposure characteristics are used due to the dense pattern in the memory cell.

  Although the present invention has been described based on the preferred embodiment, the mask pattern design method of the present invention is not limited to the configuration of the above embodiment, and various modifications can be made from the configuration of the above embodiment. Further, modifications and changes are also included in the scope of the present invention.

  The present invention can be applied to a mask pattern design method for a semiconductor device, and more particularly to a mask pattern design method for arranging an auxiliary pattern to improve the depth of focus.

The top view which shows the pupil plane of the light source which irradiates the mask designed with the mask design method of this invention. Top view of isolated pattern. The top view which shows the shape of the auxiliary | assistant pattern produced | generated from an isolated pattern. The top view which shows the example which has arrange | positioned the auxiliary pattern in one direction with respect to the isolated pattern by the method of 1st Embodiment. The top view of a mask pattern which illustrates arrangement | positioning of an isolated hole pattern and an auxiliary pattern. 6 is a graph showing a light intensity distribution when the light source of FIG. 1 and the mask of FIG. 5 are used. The graph which shows the relationship between NILS and defocus calculated | required from the light intensity distribution of FIG. The top view which shows the mask of a comparative example. The graph which shows the light intensity distribution in the best focus at the time of using the mask of the comparative example of FIG. 10 is a graph showing the relationship between NILS and defocus obtained from the light intensity distribution of FIG. 9. The top view of the mask obtained by the method of 2nd Embodiment. The graph which shows the light intensity distribution in the best focus at the time of using the mask of FIG. The graph which shows the relationship between NILS and defocus calculated | required from the light intensity distribution of FIG. The top view which shows the pupil plane of the light source which irradiates a mask. The top view which shows the pattern shape of the dummy data calculated | required from a rectangular pattern, when exposing using the light source of FIG. The top view of the mask pattern which has arrange | positioned the auxiliary pattern obtained from the dummy data of FIG. The graph which shows light intensity distribution at the time of using the light source of FIG. 14, and the mask pattern of FIG. The graph which shows the relationship between NILS and defocus calculated | required from the light intensity distribution of FIG. The top view of the mask pattern of a comparative example. The graph which shows the light intensity distribution at the time of using the light source of FIG. 14, and the mask pattern of FIG. The graph which shows the relationship between NILS and defocusing obtained from the light intensity distribution of FIG. (A)-(c) is a top view of the pupil surface which shows the example of a secondary light source. The top view which shows the example which produces | generates the data of a dummy pattern from the square pattern of square shape and a rectangular shape. (A)-(c) is a top view which shows the example which produces | generates an auxiliary pattern from a rectangular-shaped rectangular pattern. (A)-(c) is a top view which shows the example which arrange | positions an auxiliary | assistant pattern in the vicinity of the long side and vertex of a rectangular-shaped rectangular pattern, when using the light source of FIG. (A) And (b) is a top view which shows the example which arrange | positions the auxiliary pattern for a rectangular-shaped rectangular pattern, when using the light source of FIG.

Explanation of symbols

1: Square-shaped rectangular pattern 11: Dummy data 2: Auxiliary pattern 21: Auxiliary pattern dummy data 3: Light source 4: Rectangular rectangular pattern 41: Dummy pattern 42: Dummy pattern 43 after OPC correction 43: After OPC correction Rectangular pattern 5: auxiliary pattern

Claims (10)

A mask pattern design method used for exposure using an illumination condition having direction dependency on resolution characteristics,
Generating dummy pattern data in the vicinity of each vertex of the rectangular pattern based on the direction dependency of the resolution characteristic of the illumination condition;
A mask pattern design method comprising: arranging an auxiliary pattern having a predetermined positional relationship with the dummy pattern adjacent to the apex.   The mask pattern design method according to claim 1, wherein the step of arranging the auxiliary pattern generates an auxiliary pattern based on a predetermined rule-based method or a model-based method.   The mask pattern design method according to claim 1, wherein the direction dependency of the resolution characteristic depends on at least the polarization or shape of a light source. The illumination condition is to increase the depth of focus in a direction having an inclination of 45 degrees with each side of the rectangular pattern,
2. The mask pattern design method according to claim 1, wherein in the step of generating dummy pattern data, data of a dummy pattern having sides inclined in a 45-degree direction from two sides forming a vertex is generated. The illumination condition is to deepen the depth of focus in a direction orthogonal to each side of the rectangular pattern and in a direction having a 45 degree inclination with each side,
The step of generating the dummy pattern data generates octagonal dummy pattern data having two sides forming a vertex and sides inclined in a 45-degree direction from the two sides. The mask pattern design method described.   The method for designing a mask pattern according to claim 1, wherein the rectangular pattern is a square.   The mask pattern design method according to claim 6, wherein the dummy pattern is a regular octagonal dummy pattern.   The mask pattern design method according to claim 1, wherein the rectangular pattern is a rectangular pattern.   The mask pattern design method according to claim 7, wherein both the rectangular rectangular pattern and the dummy data pattern generated from the rectangular pattern are corrected to make the width narrower and the length longer.   The mask pattern design method according to claim 1, wherein dummy pattern data of the auxiliary pattern is generated, and another auxiliary pattern is arranged adjacent to the auxiliary pattern based on the dummy pattern data of the auxiliary pattern. .

Images