JPS59189627A - Charged beam exposing apparatus - Google Patents

Abstract

PURPOSE:To execute high precision positioning with high productivity by providing an array slit, a blanking electrode, a pattern data generating circuit and an electro-optical system which focuses a reduced image of the one of the beams deflected or not deflected by the blanking electrode to the surface of sample. CONSTITUTION:An array slit 28a is illuminated by the beam having the uniform intensity through a condenser lens 27 and the NXN multi-beam is generated from an aperture 31. A blanker apparatus 28b has the blanker 14 corresponding to the aperture 31 of array slit 28a and each blanker 41 is provided with the blanking electrodes 43, 44 supported by an insulating substrate 42. The beam having passed the aperture 31 enters the one blanker 41. An interval of blanking electrodes 43, 44 is longer than the length of a side of aperture 31 and thereby the beam passes between the blanking electrodes 43, 44 without collision with them. When a voltage is applied to the blanking electrodes 43, 44, the beam having passed the corresponding aperture 31 can be turned ON and OFF selectively.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a charged beam exposure apparatus used for forming fine patterns of semiconductor integrated circuits, etc., and for use in the Sogelafye process.

[Background of the invention]

Conventionally, exposure devices using charged beams include ■ spot beam method, ■ shaped beam method, ■ multi-beam method,
■Mask transfer methods Devices using various methods have been developed. The spot beam method (2) has the disadvantage that the exposure time is long because the pattern is drawn by scanning the sample surface with a finely focused beam. ■The shaped beam method is 1,
This method forms a fixed or variable rectangular beam and draws a pattern for each rectangle, but since exposure is done with a single beam, productivity is not sufficient as the pattern becomes finer. . ■The multi-beam method generates multiple beams simultaneously to draw a pattern.
Drawing time is shorter than using a single beam as in methods (2) and (2), and higher productivity can be expected. The mask transfer method described in (2) is a method in which the image of the mask on which the pattern to be transferred is formed in advance is formed as it is or is reduced in size and formed on the sample surface for exposure.The mask is a thin metal plate with beam transmission holes. There is also one that uses a photo emitter. However, in this mask transfer method, since it is necessary to prepare a mask pattern in advance, it is not possible to generate an arbitrary pattern, and it is also difficult to produce a mask, making it difficult to obtain a highly accurate mask. Furthermore, in overlay exposure, highly accurate positioning becomes a problem. That is, it is difficult to align the mask and the sample, and it is not possible to correct distortions of the electron optical system, distortions of the chip, etc., which is a practical problem.

FIG. 1 is a diagram showing a conventional multi-beam type charged beam exposure apparatus. In this device, a beam emitted from an electron gun 1 is passed through a blanker 5 to a shaping aperture 2.
An image of the shaping aperture 2 is formed on the surface of the sample 4 using an aperture lens 3, and the beam is scanned using a deflector 6. In this case, since the deflector 6 is common to each aperture lens 3, the same pattern can be drawn at the same time. That is, since one chip corresponds to each aperture lens 3, chips equal to the number of aperture lenses 3 can be written in - times of drawing.

However, in this device, either the deflector 6 or each aperture lens 3
Since the distortion of the image caused by each aperture lens 3 cannot be corrected, it is difficult to perform highly accurate positioning, which is important in fine pattern formation, which is a problem.

[Purpose of the invention]

The present invention has been made in order to solve the above-described drawbacks of the conventional charged beam exposure apparatus, and an object of the present invention is to provide a charged beam exposure apparatus that has high productivity and can perform highly accurate positioning. With the goal.

[Summary of the invention]

To achieve this objective, the present invention includes an array slit having two-dimensionally arranged apertures, blanking electrodes corresponding to the apertures, and pattern data generation for applying a voltage to each of the blanking electrodes. a circuit, and an electro-optical system for forming a reduced image of either the beam undeflected by the blanking electrode or the beam deflected by the blanking electrode on the sample surface; The width of the grating section and the aberration of the double electron optical system are set so that the width and the aberration of the electron optical system are approximately the same.

According to the present invention, since multiple beams are used, productivity is improved compared to a charged beam exposure apparatus using a single beam. Furthermore, since each chip is exposed in relatively small areas, distortion correction, position alignment, etc. can be easily performed with high precision. Therefore, the charged beam exposure apparatus according to the present invention is suitable for use as an exposure apparatus for mass production of semiconductor integrated circuits having a large number of fine patterns.

[Embodiments of the invention]

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 2 shows the overall configuration of a charged beam exposure apparatus according to the present invention. In the figure, 21 is an electron gun, 22 is a Wehnelt, 23 is an anode, 24 is a Fran force 25 is a condenser lens for focusing the beam at the position of a blanking aperture 26, and 28a is an array lens having two-dimensionally arranged apertures.・N-128b is a blanker device having blankers corresponding to each opening thereof, and each blanker is provided with a flanking electrode. 27 is a condenser lens for illuminating the array pick-up lens 28a; 29.
211 is a reduction lens that reduces the image of the array slit 28a, 210 is a blanking aperture that cuts the beam deflected by the blanking electrode of the blanker device 28b, 212 is an objective lens, 213 is a deflector, 214 is a sample, and 215 is a Mark detector, 216, sample stage, 2
17 is a blanking control circuit, 218 is a pattern data generation circuit that applies voltage to each of the blanking electrodes of the blanker device 28b, 219 is a deflection control circuit, 220 is a mark detection circuit, 221 is a stage control circuit, and 222 is a computer. The interface 223 is a computer.

Next, the specific shape of the array slit 28a is
As shown in the figure. 31 is an opening, and 32 is a lattice portion between the openings 31. The aperture 31 is square, the pin-notches of the aperture 31 are equally spaced, and the arrangement is NxN. The array slit 28a is illuminated with a beam of uniform intensity and intensity by the condenser lens 27, and N×N multi-beams are generated from each aperture 31. A blanker device 28b is arranged directly below the array slit 28a, and the relationship therebetween is shown in FIG. The blanker device 28b has blankers 41 corresponding to the openings 31 of the array 28a,
Each blanker 41 is provided with a blanking electrode 43,44 supported by an insulating substrate 42. And opening 3
Each beam passing through one blanker 41 is made incident on one blanker 41. Blanking electrode 43.44
The interval between the blanking electrodes 43 . This is to avoid hitting '44. If voltage is applied to the blanking electrodes 43 and 44,
The beam passing through the corresponding aperture 31 is selectively
N.

Can be turned off. That is, it becomes possible to switch NxN beams independently.

An example of the configuration of the blanker device 28b is shown in FIG. A square hole 58 is made in the insulating substrate (for example, a 5i02 substrate) 42, and metal is deposited on the side surface of the hole to form the blanking electrode 43.
44, and wirings 53 and 54 are formed on the surface of the insulating substrate 42 by photolithography or the like to form blanking electrodes 4.
3.44 and wiring 53.54 are connected. The blanking electrode 43 is grounded, and the flanking electrode 44 is connected to a power source 56 via a switching circuit 55. Further, the insulating substrate 42 constituting the blanker device 28b is processed, as necessary, to attach a shield electrode or remove exposed portions of the insulator in order to prevent charge-up. For example, an electrode may be attached to the non-chamfered insulator surface of the flanking electrodes 43 and 44 in the hole 58 and connected to the ground, or between the blankers 41 arranged in a direction perpendicular to the direction in which the blanking electrodes 143 and 44 are connected. The insulator may be removed by etching, beam processing, or the like.

A method of generating a drawing pattern using the multi-beam generating mechanism consisting of the array slit 28a and the blanker device 28b constructed as described above will be explained. FIG. 6 ta+ shows the relationship between the pattern to be drawn and the pixels.
This is further divided into 62 pixels, which are the minimum unit of the pattern.
Divide it into mesh-like pieces. Note that the hatched portion represents the drawing pattern 63. Drawing is performed in units of partial areas 61. The drawing pattern 63 is limited to a pattern whose width is an integral multiple of the pixel 62. The partial area 61 is formed by N'xN openings 31 of the array slit 28a.
(Note that FIG. 6 shows the case where N=6). Furthermore, one pixel 62 is provided in one aperture 31.
Corresponds to 1. Now, to draw pattern 63,
Each blanker 41 corresponding to all the pixels 62 included in the pattern 63 is turned off, and each blanker 41 corresponding to the pixel 62 not included in the pattern 63 is turned on.

Then, as shown in FIG. 6(b), the blanker device 28
From b, the beam 64 corresponding to the pixel 62 included in the drawing pattern 63 is output without being deflected. Here, since the array slit 28a has a grating portion 32 as shown in FIG. 3, the beams are blocked by this, so that each beam 64 is separated by an interval corresponding to the width of the grating portion 32. Turn on each blanker 41 at the same time.

If an OFF signal is given, the drawing pattern 63 will be drawn at once.
.

A corresponding beam 64 is generated. The beam that has passed through the blanker device 28b is focused by a reduction lens 29 and focused at the position of a blanking aperture 210, and the beam deflected by the blanker device 28b is cut by the blanking aperture 210. Furthermore, the beam that is not deflected by the blanker device 28b is reduced by a reduction lens 211, and an image is formed on the surface of a sample 214 by an objective lens 212. At this time, the intensity of the output beam of the blanker device 28b has a distribution as shown in FIG. 7, for example. That is, a
The part corresponds to the beam passing through the aperture 31, and the b part corresponds to the part blocked by the grid part 32. When reducing this beam, assuming that there is no aberration in the electron optical system, Fig. 7 t
The beam having the intensity distribution of al is directly reduced and irradiated onto the surface of the sample 214. Therefore, since the portion corresponding to portion b in FIG. 7fal is not exposed, the pattern is cut off at this portion. Therefore, the projected image of the lattice part 32'
It is necessary to prevent this from appearing on the surface of the sample 214.

In other words, the image of the array slit (28a) on the surface of the sample 214 is blurred due to the aberration of the electron optical system, so that the grating portion 32 is not visible due to the blur (FIG. 7(1)).
), the beam corresponding to each interval 1'' 31 indicated by the dotted line has a beam distribution due to the brushing in the part corresponding to the grating part 32, and when the respective beam intensities are added, The region 61 is made to have a uniform intensity distribution except for the end portions. Then, by turning the beams ON and OFF, a plurality of rectangular beams having beam sag corresponding to the width of the grating section 32 are obtained, as shown in FIG. 7tc+. Further, even when these beams are adjacent to each other, the beam intensity at the boundary is uniform, so that no breaks occur in the drawn pattern. Here, specific numerical values will be given and explained. First, arrange the play slits 28a at 20X.
20 (N=20), and the length of one side of each opening 31 is 8.
0 μm, and the width of the grating portion 32 is 20 μm. Therefore, the array slit) 28aJ second partial area 61 -:
The length of the side is 2 mm. Condenser lens 25°2
7. If the total magnification by the objective lens 212 is IAoo, then the length of one side of the partial area 61 on the surface of the sample 214 is 5μ
m. In this case, the length of one side of the pixel 62 is 025μ
It is m. The length of one side of the image of the aperture 31 on the surface of the sample 214 is 02 μm, and the width of the image of the grating portion 32 is 0.05 μm. Here, since the electron optical system normally has an aberration of about 0.05 μm, some aberration occurs in each beam forming an image of the aperture 31 on the surface of the sample 214. If this beam deflection is about 0.05 μm, the influence of the grating portion 32 can be ignored, and the beam intensity of the grating portion 32 can be made comparable to the beam intensity of the aperture 31. Finally, let's compare it with the variable shaping beam method, which is a conventional drawing method. The maximum beam size for variable shaping is 5μm x 5μm, and the length is 5μm.
2 Considering the case of exposing lines and spaces with a width of 0.5 μm, the variable shaped beam method has a width of 0.5 μm x 5.
It is necessary to shoot the μm beam five times. Also, 0
, 5μm x 05μm through holes at 05μm intervals, 05μm x O, 5μm beam is 2
You need to take 5 shots. On the other hand, in the method of the present invention, as described above, the array slit I.
If the array is 20 x 20, the lines and spaces and through holes can be exposed in one shot. Therefore, in the variable shaped beam method, the actual drawing time is 5 times longer and 2 times longer, respectively, compared to the method of the present invention.
It will be 5 times longer. Furthermore, considering that dead time such as data transfer increases with each shot, the method of the present invention greatly reduces the number of shots, which has a great effect on productivity. Further, assuming that the length of one side of the opening 31 of the array slit 28a is 80 μm and the pitch is 100 μm, the dimensions of the blanker device 28b are set as follows, for example. That is, the length of one side of the hole 58 is 90 μm.
, the depth is 1 mm. Then, set the acceleration voltage to 30kV
, set the blanking voltage to 5V, and blanking! Aperture 2
If the position of the blanker device 28b is 200 mm below the blanker device 28b, the beam deflected by the blanker device 2 can be moved by about 200 μm, and the beam can be easily cut.

As explained above, the pattern of the partial area 61 is drawn, so in order to write one chip, the deflector 213 and sample stage 216 shown in FIG.
Repeat drawing until it covers the entire area of one chip. The pattern data is input from the computer 223 to the pattern data generation circuit 218 via the interface 222, and a blanking signal is sent to the blanker device 28b. Mark detection is performed by extracting a mark signal by a mark detector 215 and processing the signal by a mark detection circuit 220. When writing a pattern that exceeds the deflection area of the deflector 213, the stage may be moved by the stage control circuit 221. The blanker 24 is turned on and off by a blanking control circuit 217, but this is provided to reduce the time during which the array lift 28a is irradiated with the beam and to prevent sea contamination. This blanker 24 does not pose any problem in terms of functionality, and distortion in the optical system can be corrected with high precision by correcting the deflection voltage generated by the deflection control circuit 219.

Note that in the above embodiment, the blanking electrode 43.
Although the beam that is not deflected by the blanking electrodes 43 and 44 is reduced and imaged onto the sample 214 surface, the beam that is deflected by the blanking electrodes 43 and 44 may be reduced and imaged onto the sample 214 surface.

〔Effect of the invention〕

As explained above, in the charged beam exposure apparatus according to the present invention, complex patterns can be exposed using a large number of blanking electrodes, so that productivity can be improved. Furthermore, overlapping drawing and distortion correction can be easily performed in the same manner as in the conventional variable transformation method. Furthermore, the image of the lattice portion of the array slit does not appear on the sample surface, and a uniform beam can be obtained. Therefore, exposure 1 for pattern drawing of an LSI having a submicron pattern
It has advantageous features as a device and is suitable for use in the LSI manufacturing process.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a conventional multi-beam type charged beam exposure device, FIG. 2 is a diagram showing the overall configuration of a charged beam exposure device according to the present invention, and FIG. 3 is a diagram showing the specific shape of the array slit. 4 shows the relationship between the array slit and the blanker device, FIG. 5 shows the configuration of the blanker device, FIG. 6 [a] shows the relationship between the drawing pattern and the pixels, FIG. (1)・ is a diagram showing the beam output without being deflected from the Fran force device, and Figure 7 is a diagram showing the correspondence between the output beam intensity of the blanker device and the beam intensity imaged on the sample surface. be. 27...Condenser lens °28a...Array slit 28b...Blanker device 29...Reducing lens 31
・Aperture 32...Grid portion 41...Blankers 43, 44...Blanking electrode 2''0...Frankink diaphragm 2]1...Reducing lens 212...Objective lens 214...Sample 218...Pattern Data generation circuit patent applicant Junnosuke Nakamura, patent attorney representing Nippon Telegraph and Telephone Public Corporation Figure 3 Figure 5 Figure 6 Figure 3 Off diagram (C) -βVIA-

Claims (1)

[Claims] an array slit having apertures arranged two-dimensionally, blanking electrodes corresponding to the apertures, a pattern data generation circuit applying a voltage to each of the blanking electrodes, and a beam not deflected by the blanking electrodes; and an electron optical system that forms a reduced image of either one of the deflected beams on the sample surface, so that the width of the reduced image of the lattice portion between the apertures and the aberration of the electron optical system are approximately the same. A charged beam exposure apparatus characterized in that the width of the grating portion and the aberration of the electron optical system are set.