JP2009295893A - Proximity effect correction method and electron beam drawing device using the method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a proximity effect correction method capable of reducing processing time for correction of a proximity effect while maintaining high drawing precision of a drawing pattern in a proximity effect correction method in fine pattern drawing using an electron beam drawing device of high acceleration voltage. <P>SOLUTION: The proximity effect correction method is characterized that the whole drawing pattern is divided into unit zones with predetermined size, pattern area density in each unit zone is calculated, the calculated pattern area density is smoothed between the adjacent unit zones to be considered as smoothed pattern area density, the whole drawing pattern is divided into a plurality of areas from an area in which change of the smoothed pattern area density is large to an area in which change of the smoothed pattern area density hardly exists in the adjacent unit zones, and different proximity effects are applied to each of the plurality of areas. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

  The present invention relates to drawing of a fine pattern, and in particular, in an electron beam drawing apparatus for manufacturing a semiconductor, a photomask for semiconductor, a template used for nanoimprint, an optical element, a biochip, and the like using an electron beam. The present invention relates to a proximity effect correction method and an electron beam drawing apparatus using the method.

  The pattern forming method by the electron beam drawing apparatus has a high fine pattern forming ability, but the so-called proximity effect that the incident electron beam for exposure is scattered in the resist and the substrate and a pattern as designed cannot be obtained. With problems. That is, the electron beam applied to the substrate on which the resist film for pattern formation is formed is scattered (forward scattered) at a minute angle with respect to the traveling direction in the resist and the substrate. In addition, a phenomenon is known in which light is scattered (backscattered) at a wide angle in the direction opposite to the incident direction in the substrate, and then returned to the resist film and exposed over a wide range. Due to the scattering of the electron beam, energy due to the electron beam is accumulated even in a region that is not originally intended for exposure, and the resist material is exposed to light, resulting in a phenomenon of proximity effect in which the pattern line width differs from a desired size. In recent years, the widening of the proximity effect due to the higher acceleration voltage of electron beam lithography systems aimed at improving the resolution of increasingly fine and complex patterns has become a serious obstacle in electron beam lithography, and the correction of this It is necessary.

  The phenomenon of the proximity effect can be analyzed by calculating an energy distribution in which incident electrons are scattered and accumulated. In order to correct this proximity effect, electron scattering is predicted from the density and pattern size of the drawing pattern, the correction amount is calculated from this prediction value, and the drawing pattern shape and exposure amount are set so that the accumulated energy distribution is uniform. A technique for obtaining a desired pattern dimension by changing the above is called proximity effect correction, and practically a technique for correcting by changing the exposure amount by controlling the exposure time when drawing a drawing pattern is the mainstream.

  Conventionally, various methods for correcting the proximity effect have been proposed. For example, (1) a representative point is provided at the corner of the drawing pattern or at the midpoint of each side, the exposure intensity at each representative point is calculated using an EID (Energy Intensity Distribution) function, and the exposure energy required for resist pattern formation A method for optimizing the exposure amount in each exposure shot so as to match the threshold value (sequential calculation method based on representative point evaluation), and (2) dividing the drawing pattern into unit partitions (called meshes) of a certain size, A method of performing one-by-one calculation by representative point evaluation (representative figure) by replacing one or a plurality of figures with one rectangular figure which is equal to the total area of the figure in the unit block and located at the area centroid (3) Divide the entire drawing pattern into unit blocks (mesh) of a predetermined size, calculate the area density of the drawing pattern in each mesh, and calculate the area density map. Calculating the accumulated energy by using a method of storing energy in each mesh to optimize the exposure amount to be constant (area density method) and the like.

  The size of mesh division by the representative figure method or the area density method (hereinafter referred to as mesh size) is generally about several μm. The higher the area density, the greater the influence of backscattering and the greater the stored energy due to the proximity effect. Therefore, the exposure amount of each exposure shot is decreased, and the exposure amount of each exposure shot is increased as the area density is decreased. In the area density method, a method of calculating by using an area density map obtained by smoothing (averaging) the pattern area density between adjacent meshes is generally used.

As a method for correcting the proximity effect, for example, Patent Document 1 (Japanese Patent Laid-Open No. 9-186058) discloses an electron that irradiates an electron beam photosensitive resist material on a substrate with an electron beam and draws a pattern on the resist material. Proximity effect correction that divides the pattern to be drawn into predetermined unit sections and corrects the amount of electron beam exposure to be irradiated to each unit section in consideration of accumulated energy caused by backscattering of electrons in the line lithography technique (A) a step of developing each unit partition into a bitmap and calculating a pattern area density in each unit partition; and (B) performing an averaging process on the pattern area density in each unit partition, A later step of calculating the pattern area density, (C) a step of calculating a gradient vector of the pattern area density after the averaging process, and (D) a larger of the calculated gradient vectors. A proximity effect correction comprising: extracting a unit partition whose length is equal to or greater than a predetermined value, and determining a unit partition having a large correction error in proximity effect correction and adjusting a size of the unit partition. The law is disclosed.

  Further, in Patent Document 2 (Japanese Patent Laid-Open No. 2004-48018), a calculation formula for calculating a correction amount for proximity effect correction from the pattern area density in the area density method is added to obtain a corrected area density, thereby increasing the accuracy. A method for improving the accuracy of proximity effect correction is disclosed.

In the calculation for proximity effect correction, among the sequential calculation method, the representative graphic method, and the area density method, the representative graphic method and the area density method are generally used practically. The reason is that the sequential calculation method is a calculation method that takes into account the drawing pattern shape, so that the calculation time increases in proportion to the increase in the number of patterns accompanying the miniaturization of the design rule, and the practical usability is low. is there. On the other hand, in the representative figure method and the area density method, the drawing pattern is used only to calculate the area density, so it is handled only by the center of gravity and area of the figure, and the calculation time for proximity effect correction is the number of patterns. The processing can be performed in a short time without depending on it. The representative figure method and the area density method can be set to a smaller unit partition (mesh) in order to capture the changes in the pattern density in more detail in accordance with the fine drawing pattern shape, and to calculate more accurately in order to calculate more accurately. A method to be used in combination with the law has been proposed.
JP-A-9-186058 Japanese Patent Laid-Open No. 2004-48018

  At present, in order to cope with processing of a finer pattern, in the next-generation type electron beam lithography apparatus, the acceleration voltage is increased from 50 kV to 100 kV. However, when the acceleration voltage of the electron beam lithography apparatus is increased in order to obtain a high resolution, the forward scattering diameter becomes small, while the back scattering diameter becomes large. In electron beam lithography with an acceleration voltage of 100 kV, the back scattering diameter is 50 kV. 3 to 4 times as large as 30 to 40 μm. Along with the widening of the range affected by backscattering, it is necessary to widen the storage area for calculation of stored energy and the smoothing range when performing proximity effect correction. As a result, when calculating the stored energy of the proximity effect correction method such as the representative figure method, the number of meshes to be captured increases, so that there is a problem that the calculation time for proximity effect correction increases, and the operation rate of the electron beam lithography system There was a problem of causing a drop in

  Therefore, the present invention has been made in view of the above problems. That is, the first object of the present invention is a proximity effect correction method in fine pattern drawing using an electron beam drawing apparatus with a high acceleration voltage, and processing for proximity effect correction while maintaining high drawing accuracy of a drawing pattern. Providing a proximity effect correction method that can reduce time.

  The second object of the present invention is to provide an electron beam drawing apparatus capable of improving the apparatus operating rate by reducing the processing time for proximity effect correction in fine pattern drawing using the electron beam drawing apparatus. That is.

There is periodicity in patterns of some products of semiconductors and semiconductor photomasks, nanoimprint tenn plates, optical elements, biochips, and the like. The present invention simplifies proximity effect correction of a drawing pattern having such periodicity.

  To this end, the invention according to claim 1 of the present invention is a proximity effect correction method for drawing a drawing pattern in which figures of the same shape are periodically arranged with an electron beam, and the entire drawing pattern has a predetermined size. The pattern area density in each unit section is obtained, and the obtained pattern area density is smoothed between adjacent unit sections to obtain a smoothed pattern area density, and the smoothed pattern area between adjacent unit sections is obtained. The entire drawing pattern is divided into a plurality of regions from a region where the density change is large to a region where there is almost no change in the smoothing pattern area density, and different proximity effect correction is applied to each of the plurality of regions. And

  According to a second aspect of the present invention, in the proximity effect correction method according to the first aspect, the proximity effect correction in which the convergence condition of the proximity effect correction calculation is strict is applied to a region where the change in the smoothing pattern area density is large. Applied to the region where there is almost no change in the smoothing pattern area density is to apply proximity effect correction that relaxes the convergence condition of proximity effect correction calculation, or proximity effect correction that assigns a fixed correction value. .

  The invention according to claim 3 is the proximity effect correction method according to claim 1 or 2, wherein the region where the change in the smoothing pattern area density is large is at least the region where the change in the smoothing pattern area density is large. And a region having a small change, and different proximity effect corrections are applied to each of the regions.

  According to a fourth aspect of the present invention, in the proximity effect correction method according to any one of the first to third aspects, the region where the change in the smoothing pattern area density is large is from the outer edge of the drawing pattern. The region having a predetermined width and having almost no change in the smoothing pattern area density is a region obtained by removing the region from the outer edge of the drawing pattern to the predetermined width from the drawing pattern.

  The invention according to claim 5 is the proximity effect correction method according to any one of claims 1 to 4, wherein the drawing pattern includes a region where a change in the smoothing pattern area density is large, and The boundary to be divided into regions having almost no change in the smoothing pattern area density is characterized by being designated by a user who performs proximity effect correction.

  The invention according to claim 6 is the proximity effect correction method according to any one of claims 1 to 4, wherein the drawing pattern includes a region where the change in the smoothing pattern area density is large; A boundary to be divided into regions having almost no change in the smoothing pattern area density is determined by determination based on proximity effect correction calculation.

  According to a seventh aspect of the present invention, in the proximity effect correction method according to the second aspect, a constant correction value assigned to a region in which the smoothing pattern area density hardly changes is a change in the smoothing pattern area density. This is a correction value for a region with a large change adjacent to a region where there is almost no change in the proximity effect correction calculation result for a region with a large or a correction value for a region with a large saturation change at the boundary with a region with little change. It is characterized by that.

  The invention according to claim 8 is the proximity effect correction method according to any one of claims 1 to 6, wherein the correction value of the region in which the smoothing pattern area density hardly changes is a proximity effect. It is characterized in that it is performed by designation of a user who performs correction.

The invention according to claim 9 is the proximity effect correction method according to any one of claims 1 to 8, wherein the proximity effect correction uses an area density map for proximity effect correction. It is characterized in that the mesh size of the area density map is set separately for a region where the change in smoothing pattern area density is large and a region where there is almost no change.

  Further, the invention according to claim 10 is the proximity effect correction method according to any one of claims 1 to 9, wherein the proximity effect correction is a region in which the change in the smoothing pattern area density is large. The number of correction operations for proximity effect correction is set separately for an area with little change.

  The invention according to claim 11 is the proximity effect correction method according to any one of claims 1 to 10, wherein the proximity effect correction uses a sequential calculation method for proximity effect correction. A representative point used for correction calculation is set separately for a region where the change in smoothing pattern area density is large and a region where there is almost no change.

  The invention according to claim 12 is the proximity effect correction calculation method according to any one of claims 1 to 11, wherein the drawing pattern can be expressed using line symmetry or rotational symmetry. Alternatively, a proximity effect correction value is calculated for one region divided at a symmetry point, and the calculation result is used as a proximity effect correction value for another region corresponding to the region via a symmetry axis or a symmetry point. To do.

  According to a thirteenth aspect of the present invention, in the proximity effect correcting method according to the twelfth aspect of the invention, the setting of the symmetry axis or the symmetry point is performed by a user who performs the proximity effect correction.

  According to a fourteenth aspect of the present invention, in the proximity effect correcting method according to the twelfth aspect, the setting of the symmetry axis or the symmetry point is determined by determination based on proximity effect correction calculation.

  The invention according to claim 15 is the proximity effect correction method according to any one of claims 1 to 14, wherein the figure having the same shape is a figure having a single shape made up of a plurality of figure groups. It is characterized by being.

  An invention according to claim 16 is an electron beam drawing apparatus using the proximity effect correction method according to any one of claims 1 to 15.

  According to the proximity effect correction method of the present invention, the proximity effect correction calculation with severe convergence conditions is performed in the region near the drawing pattern outer edge where the change in the smoothing pattern area density between adjacent unit sections of the drawing pattern is large, Proximity effect correction that relaxes the convergence condition of proximity effect correction calculation or proximity effect correction that assigns a fixed correction value in the area near the center of the drawing pattern where there is almost no change in the smoothing pattern area density between adjacent unit sections Applicable. This convergence condition is intimately related to the calculation time of proximity effect correction. According to the present invention, in a proximity effect correction method in fine pattern drawing using an electron beam drawing apparatus with a high acceleration voltage, a drawing pattern is divided into a plurality of regions, and different appropriate proximity effect correction parameters are set in each region. By applying, it is possible to reduce the calculation time for proximity effect correction while maintaining high drawing accuracy of the drawing pattern.

  According to the electron beam drawing apparatus of the present invention, it is possible to improve the apparatus operating rate by reducing the proximity effect correction processing time in fine pattern drawing.

  The proximity effect correction method of the present invention divides a drawing pattern in which figures having the same shape (hereinafter also referred to as a pattern) are periodically arranged into unit sections (also referred to as meshes) of a predetermined size, ) To obtain a pattern area density in each unit section, and smooth the obtained pattern area density between adjacent unit sections to obtain a smoothed pattern area density. From a region where the change in the smoothed pattern area density is large, a smoothed pattern is obtained. In this method, the drawing pattern (whole) is divided into a plurality of regions up to a region where there is almost no change in area density, and different proximity effect correction is applied to each of the plurality of regions. In the present invention, a region in which there is almost no change in the smoothing pattern area density includes a change that can be ignored even if there is no change.

As described above, the proximity effect correction method of the present invention applies the proximity effect correction in which the convergence condition of the proximity effect correction calculation is strict to the region where the change in the smoothed pattern area density is large, and the smoothing pattern area density Proximity effect correction in which the convergence condition of proximity effect correction calculation is relaxed or proximity effect correction to which a fixed correction value is assigned is applied to an area where there is almost no change.
Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(Embodiment 1)
FIG. 1 is a diagram illustrating an overview of a proximity effect correction method according to an example of an embodiment of the present invention. FIG. 1A shows a pattern to be drawn. As shown in a partially enlarged view of FIG. 1A, this drawing pattern is a rectangular pattern having a predetermined pattern size at a predetermined pattern pitch. This is a drawing pattern in which figures having the same shape are arranged periodically. As shown in FIG. 1, in the present invention, each figure is referred to as a “pattern”, and the entire figure to be drawn, which is a set of them, is indicated as a “drawing pattern”.

  Here, acceleration voltage and proximity effect correction in electron beam drawing will be described. 2A and 2B, the drawing patterns shown in FIG. 1A are drawn at an electron beam acceleration voltage of 50 kV (back scattering diameter: 10 μm) and 100 kV (back scattering diameter: 40 μm), respectively. It is the figure which performed proximity effect correction for the purpose, and mapped the proximity effect correction value. Regardless of the electron beam acceleration voltage, the proximity effect correction value is required when the size of the drawing pattern to be drawn is approximately the same as the backscattering diameter, and the size of the pattern to be drawn is three times the backscattering diameter. When the degree is reached, the proximity effect correction value at the center begins to be constant (region B in FIGS. 2A and 2B). When the size of the drawing pattern to be drawn is larger than about three times the backscattering diameter, the region A in FIG. 2 where the proximity effect correction value changes is constant regardless of the size of the drawing pattern to be drawn. Become. The size of the region A in FIG. 2 where the proximity effect correction value changes depends on the area density of the drawing pattern and the material constituting the substrate. If the pattern area density is low, the area A is small, whereas if the pattern area density is high, the area A is large. In addition, the region A increases as the backscattering increases due to the material constituting the substrate.

  The present invention is a proximity effect correction method when drawing a drawing pattern having periodicity to a region exceeding the region affected by backscattering of an electron beam as described above, and further, such a proximity effect correction method is provided. The adopted electron beam drawing apparatus is also included. The main drawing object by the electron beam of the electron beam drawing apparatus is a resist coated on the substrate.

  Examples of products having periodic patterns that are effective when the proximity effect correction calculation method of the present invention is adopted include, for example, some products of semiconductors and photomasks for semiconductors, optical elements such as microlens arrays, LEDs, and CMOS sensors, Examples include biochips such as immunoassay chips, storage products such as patterned media and memories, photonic crystals, gratings (light diffraction elements), and imprint templates for display antireflection plates.

  The proximity effect phenomenon can be analyzed by calculating the energy distribution accumulated by the electrons incident on the resist. For example, in the proximity effect correction calculation by the sequential calculation method, when the proximity effect correction calculation of a predetermined region is performed, the influence of the region around the region (the energy storage calculation capturing region) is calculated. Energy storage is the energy (accumulated energy) that the incident electrons store in the resist. The average energy stored in the resist by the electrons incident on one point at a certain depth from the resist surface. The distribution is represented by the sum of two Gaussian distributions called an EID (Energy Intensity Distribution) function, as shown in the following formula (1).

  Here, α, β, and η are constants determined by the depth from the resist surface, α is a forward scattering diameter (forward scattering spread), β is a back scattering diameter (back scattering spread), and η is backward. The scattering coefficient (backscattering energy intensity / forward scattering energy intensity: ratio to forward scattering) is shown. The first term of the EID function is the energy distribution accumulated when electrons entering the resist enter while being scattered forward, and the second term is the backscattered by the nuclei in the resist and the substrate. The energy distribution accumulated when scattering in the direction opposite to the incident direction.

  FIG. 3 is an example of a drawing pattern having periodicity due to repetition of the same figure, as shown in the enlarged view, and stored energy when this drawing pattern is drawn with an electron beam without proximity effect correction being performed. It is a simulation figure which shows distribution (plan view) and the energy storage amount of the drawing pattern center part x direction. The energy storage amount simply increases from the outer edge portion of the drawing pattern toward the center portion, and becomes a constant value at a predetermined location. The present invention performs proximity effect correction using the above-described characteristics of a drawing pattern having periodicity.

As described above, in the proximity effect correction method of the present invention, the proximity effect correction in which the convergence condition of the proximity effect correction calculation is strict is applied to the region where the change in the smoothed pattern area density is large, and the smoothed pattern area density In the region where there is almost no change, proximity effect correction is performed by relaxing the convergence condition of the proximity effect correction calculation, and the proximity effect correction method applied to these is not particularly limited, and is conventionally known. The proximity effect correction method is used, and examples thereof include a sequential calculation method, an area density method, and a representative figure method. In the claims and the description, it is expressed as “region with a large change in smoothing pattern area density”, “region with almost no change in smoothing pattern area density”, etc. Defining that the area density change is a region larger than the first predetermined value and that the latter is a region where the smoothing pattern area density change is smaller than the second predetermined value. Can do.
Hereinafter, an outline of each method will be described.

(Sequential calculation method)
An outline of the sequential calculation method will be described. The sequential calculation method focuses on the drawing pattern portion and determines the exposure amount for each drawing pattern so that the accumulated energy density (accumulated energy amount / pattern area) of the drawing pattern portion is constant for all patterns. Is. FIG. 4 is a diagram for explaining the concept of the sequential calculation method in the proximity effect correction calculation. FIG. 4A shows a drawing pattern for a general electronic device, which has no periodicity and is not a drawing pattern targeted by the present invention. FIG. 4B is a diagram showing a drawing pattern having the periodicity targeted by the present invention over the entire drawing pattern region. Here, the capture energy calculation capture area is designated as a rectangular area, but a circular area may also be designated.

FIG. 5 is a diagram for explaining the concept of the sequential calculation method in proximity effect correction. Here, the storage area of the stored energy calculation is specified as a rectangular area, but a circular area may be specified. As shown in FIG. 5, when there are m figures in the stored energy calculation capturing area in the proximity effect correction, the energy accumulated in the area ε _i in the kth drawing is the m figure's energy. When the exposure amount with respect to each is taken, it represents with Formula (2).

  In addition, the stored energy average value in the kth figure is expressed by Equation (3).

  If the exposure amount (j = 1, 2,..., M) is determined so that this energy average value is constant for all figures k = 1, 2,. Can be corrected. In this proximity effect correction method, a high-precision drawing accuracy of the drawing pattern can be obtained. On the other hand, when calculating the accumulated energy amount of each figure, two area integration calculations are required in Equation (3). Since an m × m determinant must be solved for m figures existing in a certain area, a very long calculation time is required. It is not possible to correct the proximity effect.

(Area density method)
Next, the area density method will be described with reference to FIG. FIG. 6 is a diagram for explaining the concept of the area density method in proximity effect correction. FIG. 6A is a diagram showing a drawing pattern for a general electronic device, and is not a drawing pattern targeted by the present invention. FIG. 6B is a diagram showing a drawing pattern having the periodicity targeted by the present invention over the entire surface of the pattern. In the area density method, the drawing pattern is divided into unit sections (mesh) that are sufficiently smaller than the backscattering diameter as shown in FIG. 6 (A) or FIG. 6 (B). The ratio of the pattern in each mesh is calculated as area density (pattern area / mesh area), and the accumulated energy in each mesh is calculated using the map (called area density map) to determine the corrected exposure amount. To go.

  In the area density method in proximity effect correction, a unit section (mesh) is defined, and a drawing pattern is divided by the unit section. Next, the area density of the drawing pattern in this unit section is calculated. And in order to improve that a correction calculation result changes with the position where a drawing pattern is divided | segmented within a unit division, the smoothing process (smoothing) which smoothes the pattern area density of adjacent unit divisions is performed. Based on the above processing, the dose for each unit section is adjusted. That is, in the proximity effect correction using the area density method, the area density in the unit section and the area density in other unit sections in the vicinity influence the calculation of the correction value regardless of the pattern edge of the drawing pattern. It will be. Here, the vicinity refers to a range that is affected by backscattering, and usually refers to a region having a radius around the correction target pattern that is about three times the backscattering diameter.

Since the area density method calculates as described above, in the drawing pattern having periodicity over the entire area, the influence of the backscattering is uniform in the unit section near the center. Is saturated and the correction by the proximity effect correction calculation is uniform. On the other hand, the unit section near the outer edge of the drawing pattern does not become like this because the area outside the drawing pattern exists in the vicinity. However, in the case of a drawing pattern with periodicity on the entire surface, the correction change due to the proximity effect correction between adjacent unit sections becomes smaller and saturates as it goes from the outer edge to the center of the drawing pattern. Or only figure periodicity remains. The proximity effect correction calculation method according to the present invention focuses on such points.

  In other words, in the present invention, in a portion where the proximity effect correction value changes greatly between adjacent unit sections (near the outer edge of the drawing pattern), the proximity effect correction calculation uses strict convergence conditions by precise calculation, and the adjacent unit sections In places where there is almost no change in the proximity effect correction value between them (near the center of the drawing pattern), proximity effect correction based on a convergence condition that relaxes the proximity effect correction calculation from the previous severe convergence condition, or proximity that assigns a fixed correction value The effect correction is applied.

  It should be noted that the strict convergence condition of the proximity effect correction referred to in this specification means that the mesh size of the area density map is sufficiently small (about 1/10 or less of the backscattering diameter β) in the area density method, or the proximity If the number of correction calculations for effect correction can be directly specified, this means increasing the number. The latter is also true for the sequential calculation method. In addition, to relax the proximity effect correction convergence condition, in the area density method, increase the mesh size of the area density map, or reduce the number of times when the proximity effect correction calculation count can be directly specified, or This refers to loosening the convergence condition for correction such as increasing the allowable error, and the latter is also applicable to the sequential calculation method.

  In addition, the setting of “mesh of area density map”, “number of operations until convergence, or error tolerance” as described above is a well-known matter in a computer program (software) for proximity effect correction calculation. Even in the present invention, based on such well-known matters, proximity effect correction calculation with severe convergence conditions is performed near the drawing pattern outer edge portion where the change in the smoothing pattern area density between adjacent unit sections is large, and adjacent to each other. Proximity effect correction in which the convergence condition of proximity effect correction calculation is relaxed or proximity effect correction to which a fixed correction value is assigned is applied in the vicinity of the drawing pattern center where there is almost no change in the smoothed pattern area density between unit sections.

  As described above, in the present invention, the calculation accuracy is maintained in a portion where the correction value greatly changes, and the convergence condition is relaxed in the portion where the correction value is predicted to be constant, thereby maintaining the drawing accuracy. The calculation time for proximity effect correction can be greatly shortened.

(Representative figure method)
Next, the representative graphic method will be described. As described above, it is known that the sequential calculation method requires enormous calculation because it requires area integration twice and solving a determinant when calculating the amount of stored energy. Therefore, a simplified version of the sequential calculation method has been proposed, and a representative figure method is one of such simplified sequential calculation methods. FIG. 7 is a diagram showing the concept of the representative graphic method. As shown in FIG. 7, the representative graphic method is a method of calculating a graphic pattern for each predetermined unit section (mesh) by approximating it to a representative graphic having the same area and center of gravity.

In the representative graphic method, the drawing pattern is divided into unit sections (mesh) sufficiently smaller than the backscattering diameter, as in FIG. 6 showing the area density method. The area density of the drawing pattern in each mesh is calculated, the accumulated energy in each mesh is calculated using the map, and the exposure amount for proximity effect correction is determined. In other words, in the proximity effect correction using the representative figure method, the area and centroid position of the pattern in the mesh and the area and centroid position of the pattern in other meshes in the vicinity are corrected regardless of the pattern edge of the drawing pattern. It will affect the calculation of the value. Since the calculation time for proximity effect correction using the representative figure method does not depend on the number of patterns included in the drawing area, the proximity effect correction time will increase significantly even if the drawing pattern is further refined. There is no.

  In the proximity effect correction using the representative graphic method, the above correction calculation is performed. Therefore, in a drawing pattern having periodicity, a region to be taken in to calculate the accumulated energy (distance from the correction target mesh = backscattering diameter × In the area that is further inward from the drawing pattern edge than the size of about 3 to 4 times, there is almost no change in the smoothed pattern area density of the captured area, so the influence of backscattering is also constant and close The correction value by effect correction can be made constant.

  On the other hand, in the region up to the same level as the region captured to calculate the stored energy from near the pattern edge where the area density changes, the value of the area density map included in the captured region in the stored energy calculation for each mesh is This does not happen because it changes. In the case of a drawing pattern having periodicity, the number of meshes in which there is a change in the pattern area density occupying the capture area at the time of stored energy calculation as the smoothing pattern area density changes from the outer edge to the center of the drawing pattern. In any case, all become only meshes with a constant pattern area density. The proximity effect correction calculation method according to the present invention focuses on such points. Here, the size of the region where the correction value by proximity effect correction changes is an example. For example, in a drawing pattern having a periodicity in which rectangular patterns of a predetermined size as shown in FIG. 1 are arranged at a pattern pitch twice as large as the pattern size, the backscattering diameter is 1.5 to 2 times from the vicinity of the drawing pattern edge. The correction value can be made constant in the area that is inside the degree pattern. This is because the area where the smoothing pattern area density changes and the area affected by backscattering differ depending on the area density of the drawing pattern. The smaller the drawing pattern area density, the more the area where the smoothing pattern area density changes and the backscattering This is because the influence range becomes small. Therefore, even in the case of a drawing pattern having a periodicity that is similarly composed of a rectangle, the size of the area where the smoothing pattern area density and the proximity effect correction correction value change depending on the pattern pitch and the material constituting the substrate. Note that is determined.

  In other words, in the present invention, in a region where a mesh having a large change in smoothed pattern area density between adjacent meshes is taken as an accumulated energy calculation region (near the outer edge of the drawing pattern), the mesh size is reduced or proximity effect is corrected in proximity effect correction. Increasing the number of correction calculations and increasing the mesh size or decreasing the number of proximity effect correction calculations as the change in the smoothing pattern area density between adjacent meshes decreases in the mesh captured as the stored energy calculation region.

  The setting of “mesh of area density map” and “number of proximity effect correction calculations” as described above is a well-known matter in the computer program (software) for proximity effect correction calculation, and this is also the case in the present invention. Based on this well-known matter, the mesh size of the area density map, which is the proximity effect correction, is reduced in the vicinity of the drawing pattern outer edge where there is a large change in the area density of the smoothed pattern between adjacent meshes in the area where stored energy is captured. Increase the number of proximity effect correction operations, increase the number of proximity effect correction operations, increase the mesh size of the area density map as the change in smoothing pattern area density between adjacent meshes decreases, It can be configured to perform proximity effect correction by, for example, reducing it.

As described above, in the present invention, in a portion where there is a mesh having a large change in area density of the smoothing pattern between adjacent meshes in the region where the stored energy is captured, the area density map is reduced in size or the number of proximity effect correction calculations. Increase the proximity effect correction calculation to increase the area density map mesh as the change in smoothing pattern area density between adjacent meshes in the area where stored energy is captured becomes smaller, or proximity By performing the proximity effect correction by reducing the number of effect correction calculations, it is possible to greatly shorten the proximity effect correction calculation while maintaining the drawing accuracy.

(region)
Next, the area | region in the proximity effect correction method of this invention is demonstrated. FIG. 1B shows an image of a region where the proximity effect correction of the drawing pattern of FIG. In the example shown in FIG. 1B, the first region and the second region are defined as two regions from the outer edge of the drawing pattern toward the center.

  In the first region, different correction values are assigned to each figure by proximity effect correction, and the correction values are saturated in the second region, and all the same correction values are taken.

  In the present invention, if the calculation is performed under a certain convergence condition for the proximity effect correction calculation in the first region, the proximity effect correction in the first region is calculated in the proximity effect correction calculation in the second region. The calculation is performed using the convergence condition relaxed from the calculation. In addition, the above-mentioned thing is used about the definition of a convergence condition. Hereinafter, the case of two regions will be described as an example.

  FIG. 8 shows an example of area division. FIG. 8B-1 shows a mesh defined in the first region, which is the outer edge of the drawing pattern in FIG. 8A. FIG. 8B-2 shows an area density graph before smoothing processing. FIG. 8B-3 is an area density graph subjected to the smoothing process, and a change in area density is recognized (in the present invention, the change in the smoothing pattern area density is large). FIG. 8C-1 shows a mesh defined in the second region inside the drawing pattern of FIG. 8A. FIG. 8C-2 shows an area density graph before smoothing processing. FIG. 8C-3 is an area density graph subjected to the smoothing process, and the change in the area density is negligibly small (in the present invention, the change in the smoothing pattern area density is hardly changed). ). FIG. 8D shows a region including a mesh having a large change in smoothing pattern area density between adjacent meshes as shown in FIG. 8B-3 in the storage energy calculation capturing region in the proximity effect correction calculation. A second region that is a region including a mesh that has almost no change in the smoothed pattern area density between adjacent meshes as shown in FIG. 8C-3 in the storage energy calculation capturing region in the proximity effect correction calculation. Is defined.

  Within the first region, proximity effect correction values that are different for each mesh are allocated by proximity effect correction, and within the second region, the correction values are saturated and the same correction value is obtained for all meshes.

  In the present invention, if the calculation is performed using a predetermined proximity effect correction parameter for correcting the proximity effect in the first region, the proximity effect correction in the second region is more effective than the proximity effect in the first region. Also, the calculation is performed by increasing the mesh size to be applied to the proximity effect correction, or by using the proximity effect correction parameter in which the number of proximity effect correction operations is reduced.

  Here, FIG. 9 shows the relationship with the accumulated energy calculation region when the proximity effect correction is performed in the first region and the second region. In FIG. 9A, the smoothing pattern area density between the first area including the mesh that has a large change in the smoothing pattern area density between the adjacent meshes in the storage energy calculation capturing area in the proximity effect correction calculation, and the adjacent meshes. In the case where the proximity effect correction of the second region, which is a region including only the mesh having almost no change in the storage effect calculation capturing region in the proximity effect correction calculation, is performed, the stored energy of the mesh in the first region adjacent to the second region Represents the capture area.

In the accumulated energy calculation, the region up to the region affected by the accumulated energy due to backscattering (radius: backscattering diameter × about 3 to 4) is set as the calculation region centering on the mesh to be obtained. Therefore, in the proximity effect correction calculation, as the accumulated energy calculation area of the first area, the hatched area shown in FIG. 9B to which the area affected by the accumulated energy due to backscattering is added is necessary. As the accumulated energy calculation area, the shaded area shown in FIG. 9C to which the area affected by the accumulated energy due to backscattering is added is necessary (hereinafter, the area added to each area is calculated. Called).
Here, it is assumed that the mesh size of the calculation margin is the same as the mesh size of the calculation area.

In the present embodiment, an example in which the drawing pattern is divided into two regions has been described. However, the number of regions can be arbitrarily determined. For example, the first region is further divided into two regions, an outer edge region (first region A) where the change in the proximity effect correction value is large and a region (first region B) where the change rate is small and gentle. It is also conceivable to apply different proximity effect corrections.
Here, the boundary between regions may be designated by a user who performs proximity effect correction, or may be determined by determination based on proximity effect correction calculation.

  In the above embodiment, in the proximity effect correction calculation in the second region in which there is almost no change in the smoothed area density map for which the convergence condition can be set most gently, the proximity effect correction is performed in the first region. As a result, when the correction value becomes constant at the boundary with the second region, the computer for proximity effect correction calculation uses the constant value (saturation value) as the correction value of the second region. It can be automatically assigned by a program (software), or the proximity effect correction user can determine the result of the first area, determine the correction value to the second area, and assign it. Alternatively, the correction value of the first region at the boundary adjacent to the second region can be assigned.

(Drawing pattern symmetry)
Next, a method for calculating the proximity effect correction when the drawing pattern has line symmetry or rotational symmetry will be described. FIG. 10A is a diagram for explaining the outline of the proximity effect correction calculation method when there is one symmetry axis in the drawing pattern, and FIG. 10B schematically shows the state of symmetry. It is. FIG. 11A is a diagram for explaining the outline of the proximity effect correction calculation method in the case where there are two symmetry axes in the drawing pattern, and FIG. 11B schematically shows the state of symmetry. It is. FIG. 12A is a diagram for explaining the outline of the proximity effect correction calculation method in the case where the drawing pattern has rotational symmetry and has two-fold symmetry, and FIG. 12B schematically shows the state of symmetry. It is shown in FIG. 13A is a diagram for explaining the outline of the proximity effect correction calculation method in the case where the drawing pattern has rotational symmetry and has four-fold symmetry, and FIG. 13B schematically illustrates the state of symmetry. It is shown in

FIG. 10 is a diagram for explaining a calculation method when there is line symmetry with the axis AA ′ as a symmetry axis. In such a case, the drawing pattern is divided into the region (I) and the region (II) by the axis AA ′. It should be noted that the “region” here is a different concept from the “region” described so far based on the area density in proximity effect correction. The “area” in the description of FIGS. 10, 11, 12, and 13 is an area related to the symmetry of the drawing pattern. In FIG. 10, when line symmetry exists with respect to the axis AA ′, the calculation is first performed in the region (I) and the calculation result is applied to the region (II). It is shortened.

FIG. 11 is a diagram for explaining a calculation method when line symmetry exists with respect to the axes AA ′ and BB ′. In such a case, the axes AA ′ and BB 'To divide the drawing pattern into four regions (I), (II), (III), and (IV), and first perform the above calculation in region (I). II), region (III), and region (IV) are applied to the calculation results to shorten the calculation time.

FIG. 12 is a diagram for explaining a calculation method when there is rotational symmetry with respect to the point X, which is the center of the drawing pattern, and two-fold symmetry. In such a case, drawing is performed by the axis AA ′. Dividing the pattern into region (I) and region (II), first performing the above calculation in region (I), and applying this calculation result to region (II) further reduces the calculation time. is there.

  FIG. 13 is a diagram for explaining a calculation method when there is rotational symmetry with respect to the point X, which is the center of the drawing pattern, and four rotational symmetry. In such a case, the axes AA ′ and B -B ′ divides the drawing pattern into region (I), region (II), region (III), and region (IV), and first performs the above calculation in region (I). The calculation time is shortened by applying the calculation result to the region (II), the region (III), and the region (IV).

  Here, with respect to the symmetry of such a drawing pattern, the setting of the symmetry axis or the symmetry point can be determined by determination by a computer program (software) for proximity effect correction calculation. Alternatively, it can be performed by designation of a user who performs proximity effect correction.

  Further, some pattern formats input to a computer program (software) for proximity effect correction calculation have a concept of “hierarchy” in order to reduce the amount of data. The basic concept of “hierarchy” is to base a group of rectangles having a fixed rule in a drawing pattern (eg, cell) and refer to the group of basic figures. Since a figure having symmetry can be expressed by using such a “hierarchy” structure, there is a method of determining the above-mentioned symmetry by “hierarchy”.

(Processing flow)
Next, an outline of the processing flow of the proximity effect correction calculation in the present invention will be described. FIG. 14 is a diagram showing an example of a processing flow in the proximity effect correction calculation according to the embodiment of the present invention.

  In step S100, pattern data to be drawn by the electron beam drawing apparatus is input to a computer program (software) for proximity effect correction calculation. The format of the pattern data input here is not limited as long as it can be recognized by later proximity effect correction software.

  In step S101, for example, for the computer program (software), for example, the drawing pattern outer edge portion (including a mesh having a large change in area density of the mesh adjacent to the accumulated energy calculation capturing region) region, and the drawing pattern center portion Designation of a region is executed (not including a mesh having a large change in area density of meshes adjacent to the storage energy calculation capturing region).

  In step S102, a calculation margin for calculating stored energy is added to each region specified in step S101. Here, the size of this calculation is the same size as the area taken in when calculating the stored energy by the proximity effect correction program.

  In step S103, proximity effect correction calculation is performed on the outer edge portion of the drawing pattern (including a mesh having a large change in area density of the mesh adjacent to the stored energy calculation capturing region).

In step S104, the proximity effect correction calculation is performed for the drawing pattern center portion (not including a mesh having a large change in the area density of the mesh adjacent to the accumulated energy calculation capturing region). Here, if the change in the area density of the mesh adjacent to the area where the stored energy calculation is captured in the central area of the drawing pattern is negligible, the proximity effect obtained in step S103 is not performed without performing the proximity effect correction. A saturation value of the correction value (a correction value of a mesh adjacent to the central area of the outer edge area of the drawing pattern) is assigned.

  In step S105, the correction value data excluding the calculation margins obtained in steps S103 and S104 is synthesized.

  In step S106, the drawing pattern with the proximity effect correction value obtained in this way is output in a format that can be interpreted by the electron beam drawing apparatus.

Here, when the proximity effect correction calculation is performed using a system capable of parallel processing, step S103 and step S104 can be performed simultaneously.
Whether or not the area designated in step S101 has sufficiently maintained the proximity effect correction accuracy is determined based on the respective correction values and drawing patterns of the overlapping areas near the drawing pattern outer edge and the drawing pattern center. It is possible to determine by comparing the calculation margins near the outer edge and the correction values of the overlapping areas with the vicinity of the drawing pattern center.

(Embodiment 2)
Note that the proximity effect correction method of the present invention can also be applied to cases where the proximity effect correction value differs for each predetermined position of one graphic, and FIG. 15 shows the proximity effect correction value for each predetermined position of one graphic. It is a figure which shows a different case. In the enlarged view of one figure in FIG. 15, the area indicated by “A” is the proximity effect correction value A, the area indicated by “B” is the proximity effect correction value “B”, and “C”. The indicated area indicates that the proximity effect correction value C is obtained. The present invention is not limited to one that takes one proximity effect correction value for one graphic, but can also be applied to cases where different proximity effect correction values are taken depending on the position in the graphic.

  That is, in the present invention, the mesh size of the area density map in the proximity effect correction is reduced in the vicinity of the drawing pattern outer edge area included in the stored energy calculation capturing area where the change in the smoothed pattern area density of the adjacent mesh is large. In the proximity effect correction calculation, the number of correction calculations is increased and proximity effect correction with severe convergence conditions is performed, and there is almost no change in the smoothed pattern area density of the adjacent mesh included in the stored energy calculation capture area. In the vicinity of the center of the pattern, it is possible to increase the mesh size of the area density map in the proximity effect correction, or to reduce the number of correction calculations in the proximity effect correction calculation so as to ease the convergence condition of the proximity effect correction.

(Embodiment 3)
Further, in the above embodiment, in the proximity effect correction of a drawing pattern having a periodic pattern composed of figures of the same shape, the change in the smoothed pattern area density of the adjacent mesh included in the accumulated energy calculation capture region As described above, the proximity effect correction method of the present invention is not limited to such a form, and can also be applied to a case where a plurality of different graphic groups are periodically arranged as basic units. However, the region of the basic unit pattern group serving as the basic unit must be sufficiently smaller than the backscattering diameter. The area of the basic unit pattern group depends on the area density, the number of patterns having different shapes included in the basic unit, and the like, but can be, for example, about 1/10 to 1/4 of the backscattering diameter. The drawing pattern of FIG. 16 is a case where two different figures form a basic pattern group as shown in the enlarged drawing pattern, and the basic pattern group is arranged with periodicity maintained. As shown in the proximity effect correction enlarged view of FIG. 16, among the different figures forming the basic pattern group, the larger pattern takes the proximity effect correction value A, and the smaller pattern takes the proximity effect correction value B. It is shown that. The present invention is not limited to a drawing pattern in which one figure is periodically arranged, but can also be applied to a drawing pattern in which a plurality of figure groups are basic units and the basic pattern groups are arranged with periodicity. is there.

  That is, according to the present invention, the mesh size of the area density map in the proximity effect correction is reduced in the vicinity of the drawing pattern outer edge area that is included in the stored energy calculation capturing area and has a large change in the smoothed pattern area density of the adjacent mesh. In the proximity effect correction calculation, the proximity effect correction is performed by increasing the number of correction calculations, and there is almost no change in the smoothed pattern area density of the adjacent mesh included in the stored energy calculation capture area. Then, it is possible to set so that the mesh size of the area density map in the proximity effect correction is increased or the number of correction calculations in the proximity effect correction calculation is reduced.

  As described above, according to the present invention, the area density map in the proximity effect correction is included in the vicinity of the drawing pattern outer edge portion where the change in the smoothed pattern area density of the adjacent mesh is large, which is included in the storage energy calculation capturing region in the proximity effect correction. The proximity effect correction is performed by reducing the mesh size or increasing the number of correction calculations in the proximity effect correction calculation, and there is almost no change in the smoothed pattern area density of the adjacent mesh included in the stored energy calculation capture area. Near the center of the drawing pattern, the mesh size of the area density map for proximity effect correction is increased or the number of correction calculations in proximity effect correction calculation is reduced to reduce the calculation time while maintaining high accuracy. It becomes possible to do.

(Verification result of proximity effect correction calculation method)
Next, verification results of the proximity effect correction method according to the embodiment of the present invention will be described. FIG. 17 is a diagram showing a drawing pattern used for verifying the proximity effect correction method of the present invention. Non-drawing rectangles are periodically arranged at a pitch twice the size of the non-drawing rectangle that is not irradiated with an electron beam.

  Using the drawing pattern as shown in FIG. 17, the calculation was performed by three methods as shown in FIG. Here, the calculation method of the proximity effect correction used in this verification was a combination of the representative figure method and the sequential calculation method. As parameters in the proximity effect correction calculation, a forward scattering diameter α = 0.002 μm, a back scattering diameter β = 40.0 μm, and a back scattering coefficient η = 0.4 were used.

  FIG. 18 is a diagram for explaining the verified proximity effect correction drawing pattern. FIG. 18A shows the case where the calculation is performed by the conventional normal proximity effect correction method over the entire drawing pattern region shown in FIG.

  FIG. 18B shows a case where the drawing effect shown in FIG. 17 is divided into a first area and a second area and the proximity effect correction method of the present invention is performed. In this verification, the smoothing pattern area density between the first area where the change in the smoothing pattern area density between the meshes adjacent to the stored energy calculation area in the proximity effect correction is large and the smoothing pattern area density between the meshes adjacent to the stored energy calculation area is It was decided to divide it into a second region that hardly changed. Further, the proximity effect correction calculation is performed only in the first region where the change in the area density between meshes adjacent to the stored energy calculation region in the proximity effect correction is large, and between the meshes adjacent to the stored energy calculation region. The value of the second region boundary obtained in the proximity effect correction calculation of the first region is assigned to the second region where there is almost no change in the area density. Here, the boundary between the first region and the second region is a square shape having a width W as illustrated. In the verification of the proximity effect correction, calculation was attempted for four types of 250 μm, 200 μm, 150 μm, and 100 μm as the width W. In addition, when calculating the stored energy in the proximity effect correction of the first region, as shown in FIG. 18 (B '), a region surrounded by a broken line having a width of 2W (stored energy calculation region) is used.

  FIG. 18C shows the drawing pattern shown in FIG. 17 divided into the first area and the second area, and among the drawing pattern areas divided by the axes AA ′ and BB ′, FIG. A case where the first quadrant of (C) (shaded hatching: calculation region) is calculated by the proximity effect correction method of the present invention is shown. Here, at the time of calculating the stored energy in the first quadrant, as shown in FIG. 18C ′, an area (stored energy calculation area) surrounded by a broken line having a width 2W and a width W calculated in the direction of the adjacent quadrant. We will use it. In the verification of the proximity effect correction method, calculation was attempted for two widths of 200 μm and 150 μm as the width W.

  FIG. 19 is a diagram illustrating a verification result. FIG. 19A is a diagram showing the results when the proximity effect correction calculation is performed with the drawing pattern divided into the first region and the second region, as shown in FIG. 18B. Here, in the proximity effect correction calculation, the proximity effect correction calculation is performed only on the first region where the change in area density between meshes adjacent to the stored energy calculation region in the proximity effect correction is large, and the mesh adjacent to the stored energy calculation region. The second region boundary value obtained by the proximity effect correction calculation of the first region was assigned to the second region in which there was almost no change in area density. In the figure, “calculation time ratio” and “efficiency” are calculated with the time required for the conventional proximity effect correction shown in FIG. 18A as 1, and the larger the number, the better the efficiency. In addition, the determination result of “×” in the figure indicates whether or not the correction value by the proximity effect correction calculation in FIG.

  As shown in FIG. 19A, as W is decreased, the calculation time is shortened and the calculation efficiency is improved. However, in order to obtain an effective proximity effect correction result, W is 150 μm, that is, the backscattering diameter is reduced. It is necessary to set it to a level smaller than four times. Here, the size of W depends on the ratio (area ratio) between the drawing region irradiated with the electron beam and the non-drawing region not irradiated with the electron beam, and the material constituting the substrate. In the shape as shown in FIG. 17, W becomes smaller when the rectangular pitch becomes smaller with the size of the non-drawing area unchanged, and W becomes larger when the rectangular pitch becomes larger.

  FIG. 19B shows the drawing pattern area divided into the first area and the second area as shown in FIG. 18C and further divided by the axes AA ′ and BB ′. It is a figure which shows the result when proximity effect correction calculation which calculates the 1st quadrant of FIG.18 (C) is performed. In the figure, “calculation time ratio” and “efficiency” are calculated assuming that the time required for the proximity effect correction in FIG. In addition, the determination result of “×” in the figure indicates whether or not the correction value by the proximity effect correction calculation in FIG. In the case of FIG. 19B, it is determined that an effective correction calculation result can be obtained with the width W of either 200 μm or 150 μm. In the case of FIG. 19B using line symmetry, the calculation time can be greatly reduced (one order) as compared with the calculation result of FIG. 19A, and the calculation efficiency is also improved. It can be seen that it can be improved at about one order level.

As described above, according to the present invention, the proximity effect correction parameter near the drawing pattern outer edge portion where the change in the smoothed pattern area density in the adjacent mesh set by the area density method is large in the storage area for calculating the stored energy in the proximity effect correction. If the mesh size is smaller or the proximity effect correction calculation count is increased and the proximity effect correction calculation is performed with high accuracy, the accumulated energy calculation capture area in the proximity effect correction is set to the adjacent mesh set by the area density method. In the vicinity of the drawing pattern center where there is almost no change in the smoothing pattern area density, the mesh size, which is a proximity effect correction parameter, is increased or the number of proximity effect correction calculations is reduced. Alternatively, the proximity effect correction parameter mesh size can be increased near the drawing pattern center where there is almost no change in area density between adjacent meshes set by the area density method in the storage energy calculation capture area in proximity effect correction. Even if the number of proximity effect correction calculations is reduced, it is possible to obtain a proximity effect correction result with the same accuracy as that near the outer edge of the drawing pattern.

  As described above, according to the proximity effect correction method of the present invention, an appropriate proximity effect correction parameter can be set according to a change in area density in an adjacent mesh set by the area density method in the storage energy calculation capturing region in proximity effect correction. By using, it is possible to shorten the calculation time while maintaining high accuracy.

(Electron beam drawing device)
According to the electron beam drawing apparatus of the present invention, at the time of pattern drawing by an electron beam, the fine effect pattern drawing is performed using the proximity effect correction method of the present invention, thereby reducing the processing time of proximity effect correction and The apparatus operating rate can be improved while maintaining high accuracy.

It is a figure explaining the outline | summary of the proximity effect correction method which concerns on embodiment of this invention. It is a figure explaining the relationship between a drawing pattern area | region and the correction result of proximity effect correction. It is a simulation figure which shows the stored energy distribution when the drawing pattern which has periodicity is drawn with an electron beam in the state which has not performed proximity effect correction, and energy storage amount. It is a figure explaining the concept of the sequential calculation method in proximity effect correction. It is a figure explaining the concept of the sequential calculation method in proximity effect correction. It is a figure explaining the concept of the area density method in proximity effect correction. It is a figure explaining the concept of the representative figure method in proximity effect correction. It is a figure which shows the mode of the smoothing of the area density map in a representative figure method. It is a figure which shows the relationship between the division area | region and stored energy calculation area | region in this invention. It is a figure explaining the outline | summary of the proximity effect correction method in case a drawing pattern is line symmetry of one symmetry axis. It is a figure explaining the outline | summary of the proximity effect correction method in case a drawing pattern is line symmetry of two symmetry axes. It is a figure explaining the outline | summary of the proximity effect correction method in case a drawing pattern has a rotational symmetry of 2 times symmetry. It is a figure explaining the outline | summary of the proximity effect correction method in case a drawing pattern has a 4-fold rotational symmetry. It is a figure which shows an example of the flow of the process in the proximity effect correction which concerns on embodiment of this invention. It is a figure which shows the case where a proximity effect correction value differs for every predetermined position of one figure. It is a figure which shows the case where a basic pattern is comprised by several figures. It is a figure which shows the drawing pattern used in order to verify the proximity effect correction method of this invention. It is a figure explaining the drawing pattern of verified proximity effect correction. It is a figure which shows a verification result.

Explanation of symbols

1 ... Drawing pattern

Claims (16)

A proximity effect correction method when drawing a drawing pattern in which figures of the same shape are periodically arranged with an electron beam,
Dividing the entire drawing pattern into unit sections of a predetermined size, obtaining a pattern area density in each unit section, smoothing the obtained pattern area density between adjacent unit sections to obtain a smoothed pattern area density,
Dividing the entire drawing pattern into a plurality of regions from a region where the change in smoothing pattern area density is large between adjacent unit sections to a region where there is almost no change in smoothing pattern area density,
A proximity effect correction method, wherein different proximity effect corrections are applied to each of the plurality of regions.   Proximity effect correction in which the convergence condition of proximity effect correction calculation is strict is applied to a region where the change in smoothing pattern area density is large, and proximity effect correction calculation is applied to a region where there is almost no change in smoothing pattern area density. The proximity effect correction method according to claim 1, wherein proximity effect correction in which the convergence condition is relaxed or proximity effect correction in which a constant correction value is assigned is applied.   The region having a large change in the smoothing pattern area density is composed of a plurality of regions including at least a region having a large change in the smoothing pattern area density and a region having a small change, and different proximity effect correction is applied to each of the regions. The proximity effect correction method according to claim 1 or 2.   The region where the change in the smoothing pattern area density is large is a region from the outer edge of the drawing pattern to a predetermined width, and the region where there is almost no change in the smoothing pattern area density is the outer edge of the drawing pattern from the drawing pattern. The proximity effect correction method according to claim 1, wherein the region is a region excluding a region from a predetermined width to a predetermined width.   A boundary for dividing the drawing pattern into a region where the change in the smoothed pattern area density is large and a region where the change in the smoothed pattern area density is hardly changed is specified by a user who performs proximity effect correction. The proximity effect correction method according to any one of claims 1 to 4.   A boundary for dividing the drawing pattern into a region where the change in the smoothed pattern area density is large and a region where the change in the smoothed pattern area density is almost unchanged is determined by determination based on proximity effect correction calculation. The proximity effect correction method according to any one of claims 1 to 4.   The proximity effect correction method according to claim 2, wherein a fixed correction value assigned to a region where there is almost no change in the smoothing pattern area density is a proximity effect correction calculation result of a region where the change in the smoothing pattern area density is large. A proximity effect correction method characterized in that the correction value is a correction value for a region having a large change adjacent to a region having almost no change, or a correction value for a region having a large change saturated at a boundary with a region having almost no change.   The proximity effect correction according to any one of claims 1 to 6, wherein the correction value of the region in which the smoothing pattern area density hardly changes is specified by a user who performs proximity effect correction. Method. In the proximity effect correction, when an area density map is used for the proximity effect correction, the mesh size of the area density map is separately set for a region where the change in the smoothed pattern area density is large and a region where there is almost no change. The proximity effect correction method according to claim 1, wherein the proximity effect correction method is set.   2. The number of correction operations for proximity effect correction is set separately for a region where the change in smoothing pattern area density is large and a region where there is almost no change in the proximity effect correction. The proximity effect correction method according to any one of claims 9 to 9.   In the proximity effect correction, when a sequential calculation method is used for the proximity effect correction, representative points used for the correction calculation are separately provided for the region where the smoothing pattern area density change is large and the region where there is almost no change. The proximity effect correction method according to claim 1, wherein the proximity effect correction method is set.   When the drawing pattern can be expressed using line symmetry or rotational symmetry, the proximity effect correction value is calculated for one region divided by the symmetry axis or symmetry point, and the calculation result is obtained via the symmetry axis or symmetry point. 12. The proximity effect correction calculation method according to claim 1, wherein the proximity effect correction calculation value is used as a proximity effect correction value for another region corresponding thereto.   The proximity effect correction method according to claim 12, wherein the setting of the symmetry axis or the symmetry point is performed by designation of a user who performs proximity effect correction.   13. The proximity effect correction method according to claim 12, wherein the setting of the symmetry axis or the symmetry point is determined by determination by proximity effect correction calculation.   The proximity effect correction method according to claim 1, wherein the figure having the same shape is a figure having a single shape including a plurality of graphic groups.   An electron beam lithography apparatus using the proximity effect correction method according to claim 1.

Images