JPS5816613B2 - Electron beam exposure method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an electron beam exposure method capable of drawing figures while maintaining a constant resolution.

In recent years, the development of electron beam exposure methods as a means of manufacturing LSI and VLSI devices has been remarkable.

One of the representative methods is the variable electron beam cross section type electron beam exposure method.

This method is based on high-speed and high-precision exposure, and the cross-sectional shape and size of the electron beam emitted from the electron beam generating means are controlled by two mask plates having polygonal holes; The cross-sectional shape and size of the electron beam is varied by means for changing the cross-sectional shape and size of the electron beam, which is composed of an electron lens and a deflector placed between the two, and is projected onto the sample and at the same time deflected by the sample 2, so that the electron beam is placed at a desired position. It is a method of drawing shapes.

However, in this electron beam exposure method, when the cross-sectional area of the electron beam is changed by the electron beam cross-sectional shape and size variable means, a hole is generated in the electron beam image on the sample, and a constant resolution cannot be maintained. Therefore, there is a problem in that it becomes impossible to draw a figure, resulting in a decrease in exposure accuracy.

The present invention has been made in view of these points, and it is an object of the present invention to provide a novel electron beam exposure method in which almost no blurring occurs in an electron beam image on a sample.

First, the principle of the present invention will be described below.

Assuming that the electron density from the electron beam generating means is constant, the electron beam cross-sectional area and the electron beam current are approximately in a proportional relationship.

Furthermore, according to the inventor's experiments, the electron beam current/IO (μA
] and blur F [μm] as shown in FIG.

That is, as shown in FIG. 2, the electron beam current is I.

At this time, the projection lens L focuses the smear beam onto the sample S, and the electron beam current is set to I, (<Io),
This is a study of the degree of blur when setting I3 (>Io), etc. Characteristic line A is the one focused when the electron beam current was 1 μA, and characteristic line B is the one focused when the electron beam current was 6 pA. It is.

The curve C in Figure 1 indicates the optimal focus points FA, FB, ・
It is a combination of...

In this experiment, two mask plates with 100 μm square holes were used, and two mask plates with vertical sides of 5 μm and 12.5 μm were used.
Electron beams with various rectangular cross-sectional shapes were formed, and the blurring was measured when the horizontal sides of these electron beams were changed to vary the electron beam current, and similar characteristics were obtained in all cases.

Moreover, it is clear that this is also the case theoretically.

In other words, blur related to the space charge effect in electron beams can be divided into Wsc, which cannot be controlled by adjusting the focus of the electron lens, and blur △Wf, which can be controlled by adjusting the focus of the electron lens (this blur is called out-of-focus). .

For the former, approximately every △Wsc is ■/(V3/2
・α)...Represented by (1).

(2) is the current value of the electron beam, and the coefficient is the physical constant J2m/2e.

However, m and e represent the rest mass and charge of the electron.

α indicates the aperture angle.

This equation (1) is an equation that indicates the blur in the case where there is no out-of-focus, that is, in the optimum focus state, and corresponds to the curve C in FIG.

actually appear.

The blur further includes the blur caused by the out-of-focus described above.

The effective focal length of an electron lens changes depending on the electron beam current due to the space charge effect, and the deviation from the focal length f when the current value is 0, fs, is approximately proportional to the current and is approximately expressed by the following equation. .

△f s = K f・■ ・・・・・・・・・・・・
(2) Now, a certain current value ■.

To achieve the optimum focus state, if the working distance is Lw, then Lw=
fo+Kf-I.

The focal length f at the current value O of the electron lens is set so as to satisfy the following.

It is to adjust. For magnetic lenses, this is done by adjusting the excitation current.

Now, if the electron beam current changes to 11 in this state, the focus position becomes Lf=fo+Kf・■1, and Lw
The image will be out of focus.

The amount of out-of-focus ΔWf is expressed by the following equation in a range where the deviation of the focus point from Iw is not too large.

△Wf 1 △r1α = lLf - Lwlα = IKf - I - Kf - Iolα = KflI - IIα ...... (3) O In other words, the amount of blur due to out of focus is the current value ■ at the optimal focus point.

The absolute value of the deviation from 111 (ol is proportional to □.

Therefore, as shown in Figure 1, when the focus is optimized at 6 μA, the blur will be as shown by curve B in the figure.
When the adjustment is made using lμA, the curve becomes like curve A, and the blurring becomes larger as the deviation of the current value increases.

Therefore, in order to minimize the blurring of the electron beam due to out-of-focus, it is important to minimize the deviation of the electron beam current from the current value at the optimum focus point.

Incidentally, as shown in FIG. 3, the blur F is determined from the size of the non-uniform portion of the electron dose distribution of the rectangular beam projected onto the sample.

Now, from the characteristic line between the nuclei, it can be seen that if the electron beam current is changed by changing the electron beam cross section, the blur gradually increases after the optimal focus point.

Therefore, in order to reduce blur, it is desirable to prevent the multipleton beam current from changing from the value at the optimal focus point to any other value.

That is, it is desirable to prevent the cross-sectional area of the electron beam projected onto the sample from changing from the current value to any other value.

In reality, as shown by the broken line, if we take into account that blurring up to about 0.1 μm does not significantly affect exposure accuracy, for example, when focusing when the electron beam current is 6 μA, It is considered that there is no problem in exposure accuracy even if the voltage is changed from approximately 2 μA to 8 μA.

Now, this electron beam current is the product of the electron beam cross section and the electron beam density.

Therefore, if the electron beam density is kept constant, when the electron beam cross-sectional area is increased, the electron beam current also increases accordingly, and the blur also increases accordingly.

Therefore, once the projection lens is set to focus the electron beam with the cross-sectional area on the sample, it is desirable to prevent the cross-sectional area from changing to other values.

Therefore, in the present invention, the electron beam density is kept constant, the cross-sectional area of the electron beam is kept approximately constant (however, some variation is allowed within the permissible blurring range), and the cross-sectional shape of the electron beam is varied. However, a figure was drawn on the sample.

According to such a method, graphics can be drawn using an electron beam with almost no blur, resulting in highly accurate drawings.

FIG. 4 shows an example of an exposure apparatus for carrying out such an electron beam exposure method, in which the electron beam emitted from the electron gun 1 is transmitted through the lens 2.
is projected onto a first mask M1 having a rectangular hole H1.

The electron beam conducted through the hole H1 is passed through the rectangular hole H by the lens 3.
At the same time, the electron beam is projected onto a second mask M2 having a mask M2 having a shape and size, and is simultaneously deflected by a deflector 5 which deflects the electron beam in an appropriate direction and size based on a shape and size command signal from an electronic computer 4.

The electron beam passing through the rectangular hole H2 is reduced and projected onto the sample γ by the lens 6, and at the same time is deflected to a desired position on the sample by the deflector 8 which receives a position designation signal from the computer 4. Ru.

When drawing a figure as shown in FIG.
, a lens 3 and a deflector 5 arranged between the two masks.
While keeping the cross-sectional area of the electron beam approximately constant, the shape is sequentially changed to q11q31, keeping in mind the size that can be formed using
A command is sent from the electronic computer 4 to the electron beam cross section variable means to change the cross section to q4, and the figure is drawn.

Assuming that the current value of the electron beam with the cross-sectional area (SA) at this time is, for example, 6 μA, the excitation of the projection lens 6 is controlled by the electronic computer 4 so that the electron beam is focused on the sample at this current value. be done.

When drawing a small figure q whose area is slightly smaller than sA, or a small figure QIO which is slightly larger than sA, the current of the electron beam corresponding to either area is 6 μA (within the allowable blur range), so the projection lens 6 The excitation is constant, and only the cross-sectional shape of the electron beam is changed to a shape corresponding to the small figures q91 qIO for drawing.

Incidentally, as shown in drawing Q' of FIG. 6, there are cases where it is not possible to draw the figure Q by dividing it into an area SA or an area similar to SA.

At this time, for example, the computer 4 uses a fixed projection projection so that the electron beam is focused at an electron beam current value whose cross-sectional size is sB or close to sB (for example, this value is 1 μA). Control the excitation of lens 6 again.

As in the case of FIG.
'.

q2'l (13'l q4'l..., q9'
The figure Q' is drawn by varying this.

Next, experimental examples are shown in Table 1.

In this experiment, mask plates having square holes of 100 μm square were used as the two masks of the electron beam cross-section variable means.

Using an electron beam cross-section variable means, electron beams were formed with cross-sectional shapes of (support) a square, (to) a rectangle with a length-to-width length ratio of 4:1, and (0) a rectangle with a length-to-width length ratio of 16:1. .

In addition, in the shapes θj, [Yes]) and (C), the cross-sectional area was changed as shown in ■ to ■, respectively, and the current value and blur of the electron beam having each cross-sectional area were measured.

In addition, the excitation current of the electron lens is adjusted so that the electron beam is just focused when the cross-sectional area is (■).

When the cross-sectional area was 2, 2, or 2, the excitation current was fixed at this value.

In addition, to measure the current value and blur, a knife installed to replace the sample is placed at the sample position, a Faraday tube is placed under the knife, and the knife is passed through the hole in the second mask plate. The electron beam focused by the electron lens is scanned by a deflector on the horizontal plane where the knife is placed.
The measurement was carried out by measuring the electron beam that entered the two Faraday tubes with an ammeter.

In other words, the current value is the detected value when the electron beam is not shielded by the knife, and the blur measurement is performed by firstly differentiating the electron beam current detected by the Faraday tube.
The measurement was carried out by measuring the waveform width corresponding to 10% to 90th amplitude of the second differential waveform.

According to the present invention, since an electron beam with almost no blur is projected onto the sample, figures can be drawn while maintaining a constant resolution.
As a result, highly accurate exposure becomes possible.

[Brief explanation of drawings]

1, 2, and 3 are diagrams used for detailed explanation of the present invention, FIG. 4 is a schematic diagram of an electron beam exposure apparatus for carrying out the electron beam exposure method of the present invention, and FIGS. FIG. 6 is a diagram for explaining the present invention in detail. 1: Electron gun, Ml: First mask, M2: Second mask, 3
:Lens, 4:Electronic computer, 52 deflector, 6:Lens,
7: Sample, 8: Deflector.

Claims (1)

[Claims] 1. Projecting the electron beam from the electron beam generating means onto the sample while changing the cross section by an electron beam cross section variable means consisting of two masks having polygonal holes, a lens and a deflector arranged between the masks. In a method of drawing a desired figure by projecting an electron beam with a lens and simultaneously deflecting the electron beam to a desired position with a deflector, the density of the electron beam from the electron beam generating means is kept constant, and the cross section of the electron beam is An electron beam exposure method characterized in that a figure is drawn while varying the cross-sectional shape of the electron beam while keeping the cross-sectional area of the electron beam approximately constant using a variable means.