JPS61230321A - Drawing method with charged particle rays - Google Patents

Abstract

PURPOSE:To enable a high-precision drawing by reading out from storage means the data for a small area which is divided including the additional area, and comparing the size of width direction of the respective patterns in the additional area and the original area with a reference width. CONSTITUTION:A CPU 2 sequentially reads out the data for one small area from a disk memory 3 and stores it in a memory 4. A high-speed data transmission control mechanism 5 compares the size of width direction of the respective patterns in the lug area and the original area with a micro pattern size by means of the pattern data in the small area. A trimming process is performed after performing data selection, the remaining data is converted to a drawing signal, and the signals of position, time and size are respectively sent to a deflector for beam position control 15, a deflector for blacking 8, and a deflector for beam size control 12, thereby drawing a pattern on the position corresponding to an area on the material placed in the means.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Field of Application] The present invention relates to a charged particle beam drawing method for drawing figures with high precision.

[Prior Art 1] Recently, charged particle beam lithography methods, typified by electron beam lithography methods, have been attracting attention as a method for manufacturing LSI devices and VLSI devices. In particular, taking the electron beam writing method as an example, two slits are placed on the electron beam path, a lens and a deflector are arranged between the two slits, and the deflection signal supplied to the deflector is controlled. An electron beam with a predetermined cross section is taken out from the slit placed below,
A variable area electron beam drawing method in which a figure is drawn by scanning the electron beam on a material is expected to be a high-speed, high-precision drawing method.

In such a drawing method, figures are drawn in the following manner. An area on the material consisting of magnetic storage means in which data necessary for drawing is stored (see A in Figure 3(a))
The minute figure drawn in (Figure 3 (a) 1) 1.1) 2
．． 1)3. D4. . . . )), and read out the graphic data for a minute area (see Fig. 3 (b)).
) of a I, az, a3. a4, ・
. . al . . .)) and stored in the storage means.

Now, in this case, the figures to be drawn often span different minute areas. For example, a case will be explained in which a figure p1 continuous to the minute heads 1aa+ and al is drawn. Figure 4(a) shows the tiny head bi! This is an enlarged view of at and 81. First, data for one minute area, ie, data for the minute head area a1, is read out from the storage means. Then, the data is converted into writing signals (shot position, shot time, and shot beam size signals) and introduced into the writing means, and a pattern is placed at a position corresponding to the area bltat on the material placed in the means. Make it look like you are drawing 11++.

Next, the data of the minute head area a1 is read out, and similarly, the data is converted into a writing signal (shot position, shot time, and shot beam nizing signal), introduced into the writing means, and placed in the means. The figure pt is drawn by placing a pattern p12 at a position corresponding to the area a1 on the material.

In this way, in the past, as shown in FIG. 4(b), data included in different minute regions were processed as separate data during drawing.

[Problems to be Solved by the Invention] However, if the pattern in any of the minute areas is extremely minute, taking the above-mentioned FIG. 4 as an example, the minute area a1
If the pattern p12 is, for example, an ultra-fine size pattern of 0.1 μm or less, it is difficult to accurately draw it using current drawing technology because it is ultra-fine size pattern data.

The present invention aims to solve such problems.

[Means for Solving the Problems] The charged particle beam drawing method of the present invention has data necessary for drawing divided into minute regions and stored in a storage means, and the data for the minute regions are stored in the storage means. A method in which the cross-sectional size of the charged particle beam is set to a predetermined size by reading it out from the charged particle beam and introducing it into a drawing means, and the charged particle beam is irradiated onto a predetermined position on a material to draw a predetermined figure. When dividing the data necessary for the drawing into each minute area, an additional area having an ultra-minute reference width SO is allowed to be added from the boundary of the minute area to the next minute area, and the data is stored in the storage means. The data for the minute area divided including the additional area is read out from the storage means, and the size in the width direction of each pattern in the additional area and the original area is compared with the Ji41 width So. If the size of the pattern in the width direction in the area is smaller than the standard width, the pattern data in the additional area is left as the pattern data in the original area, otherwise it is discarded, and the pattern data in the original area is If the size in the width direction is smaller than the reference width So, the data of the pattern of the waist region is discarded, and if not, the data is introduced into the drawing means so as to remain as the pattern data of the original region. It is.

[The source of invention]! l] First, the principle of the present invention will be explained. For convenience of explanation, a case will be described in which the figure 1)+ is drawn as shown in FIG. 4(a).

First, a figure (pl in Fig. 3(a), pl,
p3. p4.・・・・・・・・・Refer to S) is read out 1, and the figure data is read out for a minute area (a in FIG. 3(b)).
1. a2, a3, aa, -...
・Al...) When dividing into equal parts, see Figure 1 (a >
As shown by the broken line in the enlarged view of the minute area al+ai, an additional area of an appropriate size (hereinafter referred to as the ear area, area a
, for OI, e2. e3,04, territory
km tweets are es, ei, ey, ee)
It is classified by adding and stored in the storage device. Figure 1(b)
As shown in , the width So of the pinna (hereinafter referred to as ultra-small size) is set to, for example, 0.1 μm. Then, the data of the minute head area at with the eye area, that is, the pattern in the H area and the minute area a1 (pattern +1+++ear area pattern C++'), as shown in FIG. 1(b').
) and read out the data for the eye area pattern +1++
Compare whether the size S1 in the width direction of - is larger or smaller than the ultra-small size So. According to the comparison, if the size SL in the width direction of the pattern p11' in the eye area is smaller than the ultra-small size SO, data outside the original small area a1,
In other words, even if the data is for the eye area, the data is transferred to the minute area a.
1 data, and remove the other eye area data (referred to as ear removal processing). Then, the data of the area (see FIG. 1(C)) where the ear-picking process has been performed is converted into a drawing signal.
shot position.

shot time and shot beam size signal)
and is introduced into a drawing means, so that a pattern is drawn at a position corresponding to the area a1 on the material placed within the means. Next, in the same manner, the data of the minute area a1 with the eye area, that is, the data of the pattern (pattern pI2+pattern DI2' of the ear area) in the eye area and the minute area aI, are read out, and the pattern in the minute area a1 is read out. Compare whether the four noises Sz in the width direction of D+z are larger or smaller than the ultra-small size So. According to the comparison, if the size S2 in the width direction of the pattern m in the micro region a1 is smaller than the ultra-micro size So, even if the data is within the original micro head 1 dal,
The data of the area a1 is removed together with the data of the eye area without leaving this data as the data of the minute area al. In this way, figure 1) t is drawn. In this way, when a figure straddles an adjacent micro area, even if the pattern in the micro area is ultra-small, the ultra-fine pattern data is larger than the ultra-small size in the adjacent micro area. Since the data is processed as a pattern, the figure can be accurately drawn.

Note that if the size S2 in the width direction of the pattern 1) 12 in the minute area al is larger than the ultra-minute size So, the pattern D in the eye area is
Since the size St in the width direction of 1t- is naturally larger than the ultra-micro size So, the pattern of the ■ area I)+
+ - data are discarded, and when performing ear cutting processing for the micro head 111a+, the size S2 in the width direction of the pattern I)+z in the micro area a1 is naturally larger than the ultra micro size So. The data is left as the data of the minute area a.

Then, by drawing the pattern D++ and 1) tz in separate shots, the figure p1 is drawn.

At this time, since the size of the pattern p12 is larger than the ultra-fine pattern size, it is drawn with high precision.

[Example] FIG. 2 is a schematic diagram of an electron beam lithography SN apparatus shown as an applied example of a device for drawing figures based on the above-mentioned principle.

In the figure, 1 is a magnetic storage means in which data necessary for drawing is stored, and 2 is a magnetic storage means for reading graphic data to be drawn on an area on the material from the storage means. A central control unit (hereinafter referred to as CPU) that attaches an auricle to the auricle and divides it into minute regions, stores the divided data in a disk memory 3, and further stores data for one minute region of the data in a memory 4. ). 5 is the CPU2
A high-speed data transmission control mechanism (hereinafter referred to as NSC) that operates as a central control van mechanism for drawing operations according to the instructions of
Data is read from the memory 4, and the size of the pattern in the width direction of the eye area and the original area is compared with the ultra-fine pattern size, and the ear removal process is performed using the data for the one minute area. After performing this sequentially, writing signals, that is, a shot time signal, a beam size signal, and a shot position signal are created. Reference numeral 6 denotes a drawing means, and a blanking deflector 8 is sequentially installed under the electron gun 7. Focusing lens with aperture 9
゜Slit with square or rectangular hole 10. lens 1
1. Beam size control deflector 12. The slit 10
A slit similar to 13. Projection lens 14, beam position control deflector 15. Material 16 is placed. and,
The H8C5 supplies a shot time signal to the blanking deflector 8 via a pulse generator 17, and a beam size signal to the beam size control deflector 12 via a DA converter 18 for beam position control. DA converter 19
are supplied to the beam position control deflector 15 through the respective beam position control deflectors 15.

In such a device, the CPU 2 has a magnetic storage means 1.
The data of the figure to be drawn in the area formed on the material 16 is read out from the figure, and the figure data is divided into even minute parts with ears and stored in the disk memory 3. Further, the CPLI sequentially reads data for one minute area from the disk memory and stores it in the memory 4. Then, using the pattern data in the small area, HS C5 compares the size in the width direction of each pattern in the ear area and the original bear area with the ultra-fine pattern size, as explained in the principle. After discarding the data, perform the ear-picking process, and use the remaining data as a drawing signal @ (shot position).

shot time and shot beam size signal)
Convert the shot position signal to deflection for beam position control! 15, a deflector 8 for blanking the shot time signal;
Then, the beam size signal is transmitted to the beam size control deflector 12.
and draw a pattern at a position corresponding to the area on the material placed in the means. Thereafter, in the same manner, the I''SC5 stores the data of the minute area in the memory 4.
The size in the width direction of each pattern in the ear region and the original (original) region is compared with the ultra-fine pattern size, data is discarded, and then ear removal processing is performed. The data is converted into a drawing signal and the shot position signal is sent to the beam position control deflector 15, the shot time signal is sent to the blanking deflector 8, and the beam size signal is sent to the beam size control deflector 12. A pattern is drawn at a position corresponding to the area on the material placed inside. If a shot is made based on the data remaining in each of the minute areas al and al in this way, the figure p1 is drawn.

[Effects of the Invention] According to the present invention, even when the pattern size extending over adjacent micro-areas created by dividing graphic data is extremely small, the data can be treated as a pattern with a size larger than ultra-micro in the adjacent micro-areas. Because it is processed, it is possible to draw accurately. In addition, in such a case, the processing time can be reduced compared to the conventional case where pattern data of each minute area is processed separately and then combined later.

[Brief explanation of the drawing]

Figure 1 is used to explain the invention in detail, Figure 2 is used to explain the invention in detail.
The figure is a schematic diagram of an electron beam lithography system shown as an application example of the present invention, FIG. 3 shows the area formed on the material, and FIG.
The figure is used to explain the conventional drawing method.

Claims (1)

[Claims] The data necessary for drawing is divided into minute regions and stored in a storage means, and the data for the minute regions are read out from the storage means and introduced into the drawing means, thereby determining the size of the cross section of the charged particle beam. In a method in which the charged particle beam is irradiated onto a predetermined position on a material to draw a predetermined figure, when the data necessary for drawing is divided into minute regions, From the boundary to the adjacent minute area, an additional area having an ultra-minute reference width S_0 is allowed to be extra, and is stored in a storage means, and the data for the divided minute area including the additional area is stored from the storage means. Read and compare the size in the width direction of each pattern in the additional area and the original area with the reference width S_0, and if the size in the width direction of the pattern in the additional area is smaller than the reference width, the size in the width direction of the pattern in the additional area is The pattern data within is left as the pattern data of the original area, otherwise it is discarded, and if the size of the pattern in the original area in the width direction is smaller than the standard width S_0, the pattern data of the original area is A charged particle beam drawing method in which the data is introduced into a drawing means in such a way that the data is discarded, and if not, it is left as pattern data of the original area.