JPS6239535B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a data processing method in an electron beam exposure system.

It is well known that electron beam exposure systems are used in the manufacture of ICs (Integrated Cicuit). This electron beam exposure system generally includes an electron beam emitting unit that exposes a resist coated on an IC wafer along a predetermined pattern, and an electron beam emitting unit that instructs data regarding the predetermined pattern to be scanned. It is broadly divided into the central processing section. Among these, the present invention particularly refers to the latter central processing section. This central processing section is further divided into a central processing unit (CPU) and a pattern generator. The central processing unit supplies data instructing a predetermined pattern to a pattern generator, and the pattern generator drives and controls the electron beam emitting section in accordance with the supplied data.

Incidentally, in recent years, ICs have become more dense, and the amount of drawing by the electron beam emitting section and the amount of data supplied from the central processing unit to the pattern generator have become enormous. In this case, the electron beam emitting section has the ability to follow the enormous amount of drawing. However, the processing capacity of the central processing unit for a certain amount of data is quite saturated, and an increase in the amount of data directly leads to an increase in processing time, making it impossible to improve the processing capacity of the entire system. .

In order to solve this problem, a method has been adopted in which a buffer memory is placed between the central processing unit and the pattern generator. However, in order to allow the exchange of huge amounts of data, a large capacity buffer memory is required, resulting in a disadvantage that system performance deteriorates.

FIG. 1 is a schematic diagram showing an outline of a general electron beam exposure system. In this figure, 11 indicates the area of the central processing section, and 12 indicates the area of the electron beam emitting section. The electron beam emitting unit 12 includes an electron gun 12-1, a condenser lens 12-2, a reduction lens 12-3,
Consisting of a projection lens 12-4 and a deflector 12-5,
The semiconductor substrate 13 coated with resist is drawn and exposed with an electron beam. On the other hand, the central processing unit 11 mainly includes a central processing unit (CPU) 14, a buffer memory 15, and a pattern generator 16, and the output of the pattern generator 16 is provided with a D/A converter 17, a deflection amplifier 18, and the like. The output from the deflection amplifier 18 directs the electron beam along a predetermined pattern to
and scan in the Y direction. Further, the output directly supplied from the pattern generator 16 to the electron gun 12-1 is a so-called blanking output, which blanks out the electron beam when the electron beam scans a portion that should not be exposed.

The central processing unit 14 temporarily stores data regarding a predetermined pattern to be drawn and exposed in the buffer memory 15, and the pattern generator 16 raises the end signal every time one unit of drawing is completed and reads the next data from the buffer memory 15. .

In order to make the following explanation easier to understand, the semiconductor substrate 1
An example of a predetermined pattern to be drawn on 3 is shown in FIG. In this figure, it is assumed that, for example, several thousand convex patterns are arranged on the semiconductor substrate 13. 21 is the first type of convex pattern;
2 is a second type of convex pattern, and many other types of patterns are also arranged. however,
In IC memories and the like, many patterns of the same type will be arranged in a matrix. The first convex pattern 21 has a rectangular pattern A and a rectangular pattern B as one unit, and the second convex pattern 22 has a rectangular pattern P and a rectangular pattern Q, respectively.
and as one unit. When the central processing unit 14 (FIG. 1) specifies these patterns, conventionally, for example, a data structure shown in FIG. 3 has been adopted. The data structure shown in this figure is shown by taking one of the convex patterns 21 as an example, but as shown, one word of data 31 to 38 is required.
These are indicated by symbols such as AX ₁ , AY ₁ . . . BX ₂ , BY ₂ , etc., and their meanings are as follows.

AX ₁ : X coordinate of the starting point of rectangular pattern A (however, the starting point is the lower left corner of the rectangle, the same applies hereinafter) AY ₁ : Y coordinate of the starting point of rectangular pattern A AX ₂ : X of rectangular pattern A Length of one side in the Y direction AY ₂ : Length of one side in the Y direction of rectangular pattern A BX ₁ : X coordinate of the starting point of rectangular pattern B BY ₁ : Y coordinate of the starting point of rectangular pattern B BX ₂ : X of rectangular pattern B Length of one side in direction BY ₂ : Length of one side of rectangular pattern B in Y direction In other words, (X1, Y1) is the position information of the rectangular pattern, and (X2, Y2) is its size information. As described above, according to the conventional data processing method, eight words of data 31 to 38 are required to draw one convex pattern. After all, in order to obtain one convex pattern, the central processing unit 1 must
4 had to send eight words of data to the pattern generator 16, which not only increased the capacity of the buffer memory 15 but also increased the data processing time.

SUMMARY OF THE INVENTION It is therefore an object of the present invention to propose a data processing method in an electron beam exposure system, which makes it possible to eliminate the above-mentioned drawbacks.

In accordance with the above object, the present invention classifies data that is frequently used repeatedly out of data to be handled within a central processing unit into a plurality of groups and stores them in a pattern generator;
The present invention is characterized in that when the central processing unit supplies data to the pattern generator, the data is configured by designating the number of the group containing the data.

The present invention will be explained below with reference to the drawings.

The present invention proposes the following data processing method.
The amount of data specifying a pattern is reduced to about half of the conventional amount, and it is also possible to include other necessary information. The basic idea of this method is that at least data that is frequently used repeatedly, such as the size information A(X2, Y2), B(X2, Y2)...
...P(X2, Y2) and Q(X2, Y2) are classified into their respective groups in advance, and the data is designated by the group number instead of detailed bit data. This is diagrammatically shown in FIG. 4. In FIG. 4, Ga, Gb...Gp, and Gq represent the aforementioned groups. The contents of each group are size information A(X2, Y2), B(X2, Y2)...P(X2,
Y2), Q(X2, Y2). These group information are
It is stored in advance in the pattern generator 16 of FIG. Specifically, the groups Ga, Gb...Gp,
A register group corresponding to each Gq is provided in the pattern generator 16. On the other hand, the central processing unit 14 (FIG. 1) processes data 51, 52, 53, and 54 having the data structure shown in FIG.
shall be sent. Since each of the data 51 to 54 is one word, it is possible to express one unit of convex pattern from four words, which is half of the conventional method. Each word of data 51 to 54
are divided into a control block C and a data block D. In the control block C, one of the numbers Ga, Gb...Gp, Gq, etc. of the group (G) is written, and in the data block D, the starting point coordinates ( Write X1, Y1). Therefore, the structure of the data specifying one of the convex patterns 21 in FIG. 2 is as shown in FIG.
It becomes 4. That is, the control block C of each data specifies size information by a group number, and its specific contents are retrieved from the corresponding register in the pattern generator 16 (FIG. 4). Further, data block D specifies X or Y position information.

This data structure was possible because
This is also due to the fact that the X or Y position coordinates are usually expressed as 12-bit data. I mean,
Since the data handled by the central processing unit 14 generally has a structure of 1 word: 16 bits, 16-2=4 bits are so-called surplus bits. It can be said that the present invention makes effective use of this surplus bit as a control block.

A further advantage of the present invention is that each control block for data 52 and 54 becomes empty, as shown in FIG. This empty block was created because the size information was no longer specified as X2 and Y2, but was now displayed simply as a group number. As a result, it becomes possible to store other control information in these empty blocks. For example, use as control information for electron beam exposure is one example. It is generally known that when a high-density fine pattern is exposed, a so-called proximity effect occurs. This means that if two adjacent patterns are strongly exposed, patterns in between them that should not be exposed will also be exposed. In such a case, exposure amount suppression information can be placed on the empty block in order to eliminate the proximity effect. Specifically, the clock speed of the exposure clock is switched to the low side. If there is no concern about such a proximity effect, this may be specified by specifying the high side or middle side, or the high side or middle range of the continuously variable range.

As described above, according to the present invention, a data processing method is realized which can increase the processing capacity by approximately twice that of the conventional method and can also handle other control information, such as exposure control information, at the same time.

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram showing an overview of a general electron beam exposure system, FIG. 2 is a plan view illustrating a predetermined pattern to be drawn on the semiconductor substrate 13 of FIG. 1, and FIG. FIG. 4 is a diagram for explaining the data structure of data, FIG. 4 is a diagram for explaining the meaning of groups used in the present invention, and FIG. 5 is a diagram showing the data structure of data used in the present invention. In the figure, 11 is a central processing unit, 12 is an electron beam emitting unit, 13 is a semiconductor substrate, 14 is a central processing unit, 16 is a pattern generator, 21 and 22 are exposure patterns shown as examples, and 51 to 54 are each shown as an example. The data according to the invention, Ga to Gp, are each a group.

Claims (1)

[Scope of Claims] 1. An electron beam emitting unit that draws and exposes a predetermined pattern on a semiconductor substrate, a central processing unit that generates data regarding the predetermined pattern, and a central processing unit that receives data regarding the predetermined pattern from the central processing unit and In a data processing method in an electron beam exposure system comprising a central processing section having a pattern generator for driving and controlling an electron beam emitting section, Pairs of length data in the Y direction are stored in the pattern generator in advance in correspondence with group numbers specifying each of the data pairs, and the central processing unit stores the predetermined data as data related to the predetermined pattern. A data processing method in an electron beam exposure system, wherein the group number corresponding to a rectangular pattern constituting a pattern and position information of the rectangular pattern are supplied to the pattern generator.