JP2005215016A - Data compression method in electron beam lithography, and electron beam lithography method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a data compression method in electron beam lithography and a method for electron beam lithography, by which the number of steps of referring to a basic figure can be decreased, when compressed data are restored and the drawing time can be shortened. <P>SOLUTION: The data is basically compressed by a bit map system, in which one bit map compressed data contains more figures and presence/absence of a figure is not used as layout information, but rather the kinds of repeated figures are used. That is, a drawing figure is divided by a pitch such that the figure data, including repetition, is most frequently used in a predetermined drawing region. The divided figure is extracted as a basic figure. For example, when a figure in a divided region is classified into 7 kinds of patterns, library data are created from the 7 kinds of patterns as the basic figures, and in which order the 7 kinds of patterns are laid out and their pitches are recorded as the layout information data. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention provides, for example, data in electron beam lithography that compresses mask pattern data supplied to an electron beam lithography apparatus when a mask for a stepper used in a semiconductor device manufacturing process is created using the electron beam lithography apparatus. The present invention relates to a compression method and an electron beam drawing method for drawing on a mask dry plate based on compressed mask pattern data.

  In a semiconductor device such as an LSI, an integrated circuit is formed on a semiconductor wafer. In this case, the circuit is formed in multiple layers on the wafer. At this time, a resist (photosensitive agent) is applied to the surface of each layer, the wafer material is introduced into the stepper, and a mask having openings corresponding to the circuit pattern. Then, the surface of the wafer material is irradiated with light such as ultraviolet rays through the substrate, and a desired circuit pattern is collectively exposed. The wafer material for which the exposure process has been completed is subjected to development, etching, vapor deposition, ion plating, and other processes to form a single layer circuit of the integrated circuit. Such a process is repeated many times to form a multilayer integrated circuit on the wafer.

  In the mask used for the stepper used in the manufacturing process of the integrated circuit described above, a metal that does not transmit light is vapor-deposited on the surface of the glass dry plate, and only the circuit pattern portion to be exposed is removed from the vapor-deposited metal. ing. What is required of such a mask is that the circuit pattern is removed with high precision. If the position and shape of the circuit pattern deviate from the design position and shape of the integrated circuit, the mask is exposed. In such a circuit, the connection relationship with the upper layer and lower layer circuits is lost, and the manufactured semiconductor device becomes defective.

  For this reason, the mask is created using an electron beam drawing apparatus capable of drawing a pattern with high accuracy. When a mask is manufactured using an electron beam drawing apparatus, first, a metal that does not transmit light is deposited on a glass dry plate. A resist (electron beam photosensitive agent) is applied to the surface of the mask material on which metal or the like is deposited, and then the mask material is introduced into an electron beam drawing apparatus. In the electron beam drawing apparatus, a circuit pattern is drawn on the surface of the mask material with an electron beam based on circuit pattern data from CAD, and the resist is exposed. The mask material in which the resist is exposed is subjected to development processing and etching processing to form a mask for forming a circuit of one layer among many layers of the semiconductor device having a multilayer structure. A large number of masks are thus required even when a single semiconductor device having a multilayer structure is manufactured.

  As described above, in the electron beam drawing apparatus, for example, a mask corresponding to circuit patterns of many layers of a semiconductor device having a multilayer structure is drawn, but many of the circuit patterns formed on many masks are It is a combination of rectangle, triangle, or trapezoid. These various shapes are usually converted from a CAD data format for LSI into a dedicated format that can be handled by an electron beam drawing apparatus, and stored in a memory as a data file on a computer.

  This data file entails enormous amounts of data simply by enumerating the positions and sizes of figures, and the memory capacity for storing the data file must be significantly increased, and it takes a long time to transfer the data. As a result, drawing throughput is adversely affected.

  Therefore, data is usually expressed efficiently, and a data file is configured as some compressed data. In LSI, repeated graphic shapes are used very frequently, and therefore data is generally compressed by separating the data into basic graphic positions and sizes and repeated information of the graphic. As one of compressed data expressions, conventionally, a compressed expression of data based on array information (also referred to as a bitmap) as shown in FIG. 1 has been used.

  In FIG. 1, (a) is a shape P1 to be drawn. The figure P1 having such a shape is separated into a simple repetitive shape part and a non-repetitive shape part P2 as shown in (b), and an irregular repetitive shape part P3 as shown in (c). Is created. FIG. 2A shows an irregular repetitive shape portion P3, and FIG. 2B shows a configuration of compressed data of the irregular repetitive shape portion P3.

  As shown in FIG. 2B, the compressed data includes the basic arrangement pitches Px and Py and the number of repetitions Nx and Ny (see FIG. 2) for the square figure q that repeats irregularly as shown in FIG. In the example shown in FIG. 4, the information is composed of bitmap information indicating Nx = 3, Ny = 3) and the presence / absence of a graphic as 0, 1, and data (library data) of the shape of the basic graphic q to be repeated.

FIG. 3 shows an example of compressed expression of the drawing shape P1. FIG. 3A shows a drawing shape P1, and FIG. 3B shows an example of compressed expression. First, the following two figures correspond to non-repetitive figure data.
(1), (2): Long rectangles (two) at both ends (right end and left end)
(3): The following two data correspond to the graphic data of a simple repetitive portion of a rectangle with a long lower end.
(4): Vertically short rectangles (six) expressed by arrangement pitch and number of arrangement.
(5): Short rectangles (9) in the horizontal direction, expressed by arrangement pitch and number of arrangement.
The following one data corresponds to the graphic data of the irregular repeated portion.
(6): Square (six), expressed by repetition pitch and arrangement information.

Further, the compressed expression data includes shape data (library data) of repeated figures. That is, the next graphic shape data is prepared in the library data.
(4D): vertically long rectangular shape data (5D): horizontal short rectangular shape data (6D): square shape data Example in which data is compressed and transferred to the electron beam drawing apparatus in this way Patent Document 1 can be referred to.
JP-A-5-234860

  In the bitmap data expression method used for compressing the graphic data of the graphic P1 shown in FIG. 1, the number of repeated graphic types is one. As a result, in order to express all of the drawing data, it is necessary to create data by separating regular repeating parts and irregular repeating parts. In particular, the shape of the peripheral portion is often slightly different from the repeated shape, and it is necessary to separately describe that it is only a data representation method that repeats the same shape. For example, the shape of the figure P1 shown in FIG. 1 is expressed as data as shown in FIG.

  The problem here is that it takes time to read the compressed data and return it to the data before compression. That is, when drawing is started, it is necessary to return the compressed data to a state in which the positions and arrangements of the figures in the original data format are arranged.

  In the case of FIG. 3, arrangement information is read in the order of (1) to (6), and (4) to (6) are referred to as “repeated figure” (= basic figure) by the number of figures to be arranged. To convert it into a simple position and figure shape data enumeration. FIG. 4 shows a representation of compressed data for the figure (a) to be drawn, and FIG. 4 (c) shows a list of positions and arrangements of the figure in the original data format. It shows the data returned to the set state.

  As is apparent from FIG. 4, the data of the non-repetitive figures (1), (2), and (3) has the same data format as the compressed data (that is, the compression process is not performed).

  On the other hand, the graphic data is compressed for the simple repeated portion graphic (4) and (5) data, and the position information in the XY direction, the number of the library data (identification symbol), the XY direction are used as the arrangement information. Pitch (length) and number data in the XY direction. The data of the corresponding graphic is read from the identification symbols (4) and (5) of the library data in such arrangement information, and the compressed data is converted into original data as shown in FIG. It returns to the state where the position and arrangement of the figure which is the data format are listed.

  Furthermore, for the data of the irregularly repeated portion graphic (6), the graphic data is compressed, and the position data in the XY direction, the number of the library data (identification symbol), the pitch in the XY direction (length) as the arrangement information ), XY direction number data, and bitmap data.

  The data of the corresponding figure is read out from the identification symbol (6) of the library data in such arrangement information, and the compressed data is read in the original data format as shown in FIG. It returns to the state where the position and arrangement of a certain figure are listed.

  By the way, basic graphic data as library data is usually stored in a storage area (on a memory or a disk) separate from the main body of the data, so that extra time is required for reference. As a result, the compressed data is expanded and returned to the original data format in which the figure positions and arrangements are arranged, and individual patterns are drawn by the electron beam on the mask plate surface, and all patterns are drawn. It takes a long time.

  The present invention has been made in view of the above problems, and can reduce the number of times of reference to a basic figure when decompressing compressed data, thereby reducing the drawing time and the data compression method and electron in electron beam drawing. The present invention relates to a beam drawing method.

  A data compression method in an electron beam drawing apparatus according to the invention of claim 1 relates to a method for compressing drawing pattern data of a predetermined drawing area, and the drawing area is set to a desired length (pitch) in the X direction and the Y direction. Virtually divide into a grid, and use the entire figure contained in the divided area as a basic figure, and assign the identification symbol according to the basic figure contained in that area in units of areas divided into a grid. The X direction that divides the drawing pattern data into a grid pattern by attaching the same identification symbol to the basic figure region and attaching another identification symbol to the different basic figure region Library data consisting of pitch data in the Y and Y directions, identification symbol data corresponding to the basic graphic to be drawn and drawn for each divided area, and basic graphic data corresponding to the identification symbol It converted to, and characterized in that so as to compress the drawing pattern data.

  According to a second aspect of the present invention, there is provided a data compression method in an electron beam drawing apparatus according to the first aspect of the present invention, wherein the length of graphic data drawn repeatedly in the drawing area is set to one pitch each in the X direction and Y direction. This is characterized in that the drawing pattern is virtually divided into a grid pattern at this pitch.

  According to a third aspect of the present invention, there is provided a data compression method for an electron beam lithography apparatus according to the first or second aspect of the invention, wherein a specific basic figure and a basic figure of each divided area are compared by pattern matching processing. The basic figure and the basic figure of each divided area are judged to be the same, and the area of the basic figure regarded as the same is given an identification symbol of the basic figure.

  According to a fourth aspect of the present invention, there is provided a data compression method in an electron beam drawing apparatus according to the first to third aspects of the present invention, wherein the library data includes basic shape position data in a divided region and graphic shape data such as a basic graphic size. It is characterized by being.

  An electron beam drawing method based on the invention of claim 5 is an electron beam drawing method in which a desired pattern is drawn by deflecting an electron beam applied to a drawing material based on drawing pattern data. , Pitch data in the X and Y directions for dividing the drawing area into a grid, identification symbol data corresponding to the basic figure to be drawn for each divided area, and basic figure data corresponding to the identification symbol Is a compressed data in which the same basic symbol area is given the same identification symbol, and at the time of pattern drawing, according to the identification symbol given to each divided area. Reads the data of the corresponding basic figure, converts it to uncompressed data, and draws the pattern of the drawing material based on the uncompressed data It is.

  According to a sixth aspect of the present invention, there is provided the electron beam writing method according to the fifth aspect of the invention, wherein the pitches in the X and Y directions and the basic figure identification symbols for each area divided by the pitch are stored in the arrangement information memory. Graphic data such as the position and size of the basic graphic corresponding to each identification symbol is stored in the basic graphic memory, and the identification symbol for each area in the arrangement information memory is read in order, The graphic data corresponding to the identification symbol is read out from the basic graphic memory.

  According to a first aspect of the present invention, there is provided a data compression method for an electron beam drawing apparatus, wherein a predetermined drawing area is virtually divided into a lattice shape at a desired length (pitch) in the X and Y directions, and the divided areas are divided. The whole figure contained in the basic figure is the basic figure, and when the identification symbol is attached according to the basic figure contained in the area unit divided into a grid, the same identification symbol is applied to the area of the same basic figure. By attaching different identification symbols to different basic figure areas, the drawing pattern data is divided into X- and Y-direction pitch data for dividing the drawing area into a grid pattern, and the divided areas. The data is converted into identification symbol data corresponding to the basic graphic to be drawn and library data composed of basic graphic data corresponding to the identification symbol, and the drawing pattern data is compressed.

  As a result, since it is possible to describe the type of basic figure rather than information indicating whether or not there is a figure in each area divided in a lattice shape, it is possible to describe a repeated figure with one compressed data expression. Accordingly, since the reference of the basic figure arranged at the same position from the library data can be performed once, the number of reference of the basic figure can be reduced. As a result, the drawing speed can be improved. Further, the compression method according to the first aspect can provide the following effects.

  That is, the normal drawing data is sent to the drawing device in order from the upper left to the lower right of the field, for example, according to the arrangement information (bitmap information). Therefore, since the deflection position of the electron beam is also controlled in order from the upper left to the lower right, the deflection control time is shortened and the drawing speed is improved. Conventionally, when there are a plurality of repeated figures, arrangement information is also prepared for the number of repeated figures. Then, after drawing one repetitive figure over the entire field, the next repetitive figure is drawn again from the upper left of the field based on the arrangement information. In order to finish drawing all the figures in one field In this case, the deflection of the electron beam is repeated in the field many times, the deflection control time becomes longer, and the writing speed becomes slower.

  According to a second aspect of the present invention, there is provided a data compression method in an electron beam drawing apparatus, wherein the length of graphic data drawn repeatedly in the drawing area is set to one pitch each, and a drawing pattern is virtually set at this pitch. Therefore, the number of basic figures to be referenced can be reduced and library data can be easily created.

  According to a third aspect of the present invention, there is provided a data compression method in an electron beam drawing apparatus, wherein a specific basic figure and a basic figure of each divided area are compared by pattern matching processing, and the specific basic figure and the basic of each divided area are compared. Judgment of identity with a figure, and the area of a basic figure that is regarded as the same is given an identification symbol for the basic figure, so that data compression work including creation of library data can be performed accurately and quickly. Can do.

  According to a fourth aspect of the present invention, there is provided a data compression method in an electron beam lithography apparatus, wherein the library data includes basic figure position data in a divided region and figure shape data such as a basic figure size.

  In the electron beam drawing method according to the invention of claim 5, the drawing pattern data is attached to the drawing data in the X direction and the Y direction for dividing the drawing area in a grid pattern, and the basic drawing is performed for each divided area. The compressed data includes the identification symbol data corresponding to the graphic and the library data in which the basic graphic data corresponding to the identification symbol is described. At the time of pattern drawing, the corresponding basic figure data is read according to the identification symbol assigned to each divided area, converted to uncompressed data, and pattern drawing of the drawing material is performed based on the uncompressed data .

  As a result, as in the first aspect of the invention, since it becomes possible to describe the type of basic figure rather than information indicating whether or not there is a figure in each area divided in a grid pattern, a repeated figure can be expressed as one compressed data. Can be described. Accordingly, since the reference of the basic figure arranged at the same position from the library data can be performed once, the number of reference of the basic figure can be reduced. As a result, the drawing speed can be improved. Further, in the invention of claim 5, the following effects can be obtained.

  That is, the normal drawing data is sent to the drawing device in order from the upper left to the lower right of the field, for example, according to the arrangement information (bitmap information). Therefore, since the deflection position of the electron beam is also controlled in order from the upper left to the lower right, the deflection control time is shortened and the drawing speed is improved.

  In the electron beam writing method according to the sixth aspect of the invention, the pitches in the X and Y directions and the identification symbols of the basic figures for each area divided by the pitch are stored in the arrangement information memory and correspond to each identification symbol. The figure data such as the position and size of the basic figure is stored in the basic figure memory, and the identification symbols for each area in the layout information memory are read in order, and the basic figure memory is read based on the identification symbols. The graphic data corresponding to the identification symbol is read out, and the same effect as in the fifth aspect of the invention can be obtained.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 5 shows an example of a configuration for performing data compression according to the present invention. M1 is a disk memory in which circuit patterns of LSI layers designed by CAD are stored. The circuit pattern data of each layer from the disk memory M1 is supplied to the computer C for each layer.

  The computer C has a function of converting CAD data into drawing data that can be handled by an electron beam drawing apparatus. In addition, the computer C has a function of compressing supplied pattern data. The pattern data subjected to data conversion and data compression processing by the computer C is supplied to and stored in the memory M2.

  Here, in this embodiment, it is based on performing data compression by the bitmap method, but in order to include more shapes in one bitmap compressed data, it is not presence or absence of figures as arrangement information, The type of repeated shape is used. The data compression method will be described more specifically with reference to the drawing figure P1 shown in FIG.

  First, as shown in FIG. 6A, the computer C divides a drawn figure at a pitch (X, Y) at which repeated graphic data is most frequently used. This divided figure is extracted and used as a basic figure. In this case, the divided basic figures are shown in FIG. 6B, and a total of seven basic figures (1) to (7) are extracted. That is, when divided by pitches X and Y, it can be divided as four repetitive data in the X direction and four in the Y direction, and each divided area is classified into seven types of figures. These seven types of figures are extracted as basic figures, and in what order the seven types are arranged, and the pitches thereof are arranged as arrangement information data as shown in FIG.

  The drawing data compressed in this way is supplied from the computer C to the disk memory M2 and stored therein. Note that the classification of figures and the determination of the pitch length are executed by the computer C using a technique such as pattern matching processing.

  FIG. 8 shows an example of a variable area electron beam drawing apparatus for drawing a desired mask pattern based on the graphic data compressed by the system shown in FIG. Reference numeral 1 denotes an electron gun that generates an electron beam EB. The electron beam EB generated from the electron gun 1 is irradiated onto the first shaping aperture 3 through the irradiation lens 2.

  The aperture image of the first shaping aperture is formed on the second shaping aperture 5 by the shaping lens 4, and the position of the image formation can be changed by the shaping deflector 6. The image formed by the second shaping aperture 5 is irradiated onto the drawing material 9 through the reduction lens 7 and the objective lens 8. The irradiation position on the drawing material 9 can be changed by the positioning deflector 10.

  Reference numeral 11 denotes a computer. The computer 11 decompresses the compressed pattern data from the pattern data memory 12, converts the data into a simple position and figure shape data list, and transfers the data to the data transfer circuit 13. The data transfer circuit 13 includes a processing unit for scaling the transferred drawing data. Pattern data from the data transfer circuit 13 is generated from a control circuit 14 for controlling the shaping deflector 6, a control circuit 15 for controlling the positioning deflector 10, a control circuit 16 for controlling excitation of the objective lens 8, and the electron gun 1. It is supplied to a blanker control circuit 18 for controlling a blanker (blanking electrode) 17 for blanking the electron beam.

  Further, the computer 11 controls the drive circuit 21 of the stage 20 on which the material 9 is placed in order to move the material 9 for each field. Although not shown, a laser interferometer is provided in the stage portion to measure the amount of movement of the stage. The operation of such a configuration will be described next.

  First, a basic drawing operation will be described. The pattern data memory 12 stores compressed pattern data that is the same as the drawing data stored in the disk memory M2 of FIG. The stored pattern data is sequentially read into the computer 11 in field units. FIG. 9 is a block diagram for explaining a part of the internal memory and calculation function of the computer 11 used in the apparatus of FIG.

  In FIG. 9, the arrangement information memory 30 in the computer 11 stores bitmap unit pitch X and Y data and basic figure arrangement data. The basic figure memory 31 in the computer 11 stores the figure position and figure size of each extracted basic figure. Further, the computer 11 has a calculation unit 32 for expanding the compressed pattern data based on the arrangement information memory 30 and the basic graphic memory 31 and converting it into a sequence of original simple position and graphic shape data. The graphic data developed by the calculation unit 32 is supplied to the data transfer circuit 13. Based on the data from the data transfer circuit 13, the deflection control circuit 14 controls the shaping deflector 6, and the control circuit 15 controls the positioning deflector 10.

  As a result, the cross section of the electron beam is formed into a unit pattern shape by the shaping deflector 6 based on each pattern data, and the unit pattern is sequentially shot on the material 9 to perform pattern drawing of a desired shape. At this time, blanking of the electron beam is executed in synchronization with a shot of the electron beam on the material 9 by a blanking signal from the blanker control circuit 18 to the blanker 17.

  Further, when drawing on different regions on the material 9, the stage 20 is moved by a predetermined distance according to a command from the computer 11 to the stage drive circuit 21. Although the movement distance of the stage 20 is not shown, it is monitored by a laser length measuring device, and the position of the stage is accurately controlled based on the length measurement result from the length measuring device.

  Although one embodiment of the present invention has been described above, the present invention is not limited to this embodiment. For example, data may be compressed in units of fields in the mask drawing area, may be compressed for each of a plurality of fields, and further, data compression may be performed for the entire mask pattern.

  The present invention is used in the industry for manufacturing semiconductor devices.

It is a figure for demonstrating the conventional data compression method. 1 is a diagram showing an electron beam inspection system based on a scanning electron microscope which is the best mode for carrying out the present invention. It is a figure for demonstrating the expression method of an irregular repeating part. It is a figure which shows the example of the compression expression on the conventional data. FIG. 6 shows a compressed representation and a returned original uncompressed representation. It is a figure which shows the basic composition which performs the compression of the data based on this invention. It is a figure which shows the pattern which should be drawn, and the basic figure repeatedly drawn when it divides | segments by pitch X and Y. It is a figure which shows the arrangement | positioning information data at the time of dividing a drawing area | region like FIG. It is a figure which shows an example of the electron beam drawing apparatus which expand | extracts compression data and draws a desired pattern on to-be-drawn material. It is a block diagram of the internal memory and arithmetic function of a computer used for the apparatus of FIG.

Explanation of symbols

M1, M2 Disc memory C Computer 1 Electron gun 2 Irradiation lens 3 First shaping aperture 4 Molding lens 5 Second shaping aperture 6 Molding deflector 7 Reduction lens 8 Objective lens 9 Drawing material 10 Positioning deflector 11 Control CP
12 pattern data memory 13 data transfer circuit 14 shaping deflector control circuit 15 positioning deflector drive circuit 16 objective lens control circuit 17 blanker 18 blanker control circuit 20 stage 21 stage drive circuit

Claims (6)

  The present invention relates to a method for compressing drawing pattern data of a predetermined drawing area, and the drawing area is virtually divided into a grid pattern at desired lengths (pitch) in the X direction and the Y direction, and the figure included in the divided area When the whole figure is a basic figure and an identification symbol is attached according to the basic figure contained in that area in units divided into a grid, the same identification symbol is assigned to the area of the same basic figure. By attaching a different identification symbol to the basic figure area, the drawing pattern data is assigned to each divided area and the pitch data in the X and Y directions for dividing the drawing area into a grid pattern. For electron beam drawing that converts the identification symbol data corresponding to the basic figure to be drawn and the library data consisting of the basic figure data corresponding to the identification symbol, and compresses the drawing pattern data. Kicking data compression method.   2. The electron beam according to claim 1, wherein in the drawing area, the length of the graphic data repeatedly drawn in the X direction and Y direction is set to one pitch, and the drawing pattern is virtually divided into a grid pattern at this pitch. Data compression method for drawing.   A specific basic figure and the basic figure of each divided area are compared by pattern matching processing, the identity of the specific basic figure and the basic figure of each divided area is judged, and the basic figure considered to be the same 3. The data compression method in electron beam drawing according to claim 1, wherein an identification symbol of the basic figure is attached to the area.   4. The data compression method for electron beam drawing according to claim 1, wherein the library data includes figure shape data such as basic figure position data in a divided area and a basic figure size.   In an electron beam drawing method in which an electron beam applied to a material to be drawn is deflected based on drawing pattern data to draw a desired pattern, the drawing pattern data divides the drawing region into a lattice shape and the Y direction It includes direction pitch data, identification symbol data corresponding to the basic graphic to be drawn and drawn for each divided area, and library data in which basic graphic data corresponding to the identification symbol is described. The basic figure area is compressed data with the same identification symbol attached, and at the time of pattern drawing, the corresponding basic figure data is read out according to the identification symbol assigned to each divided area, and is converted into uncompressed data. An electron beam drawing method in which a pattern is drawn on a drawing material based on non-compressed data after conversion. The pitches in the X and Y directions and the basic symbol identification symbols for each area divided by the pitch are stored in the arrangement information memory, and the basic figure data such as the position and size of the basic graphic corresponding to each identification symbol is basic. 6. The identification symbol stored in the graphic memory, for each area in the arrangement information memory, is read in order, and graphic data corresponding to the identification symbol is read from the basic graphic memory based on the identification symbol. The electron beam drawing method as described.

Images