JPS6262046B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an improvement in an electron beam exposure method used to form fine patterns in semiconductor devices.

Generally, when performing electron beam exposure, electrons irradiated onto a resist film are scattered. This scattering is
This includes forward scattering that occurs in the resist film and back scattering that occurs depending on the substrate. Due to such a scattering phenomenon, exposure is performed over a wider area than the area irradiated with electrons. Therefore, if a plurality of drawing figures exist closer than the scattering distance of electrons, the patterns are mutually affected by scattering, and a pattern with the desired size cannot be obtained. This is conventionally known as the proximity effect.

Therefore, in order to form fine patterns using electron beam exposure and obtain accurate pattern dimensions, an exposure method that corrects the proximity effect is required.

Generally, in order to perform proximity effect correction, it is necessary to consider the electron scattering intensity distribution, pattern shape, and proximity, and it is only when these factors are satisfied that it is possible to give the optimal dose to each figure. Become.

By the way, in order to correct the proximity effect as described above, it is necessary to obtain an electron scattering intensity distribution equation. This scattering intensity distribution is based on the resist material used,
It depends on resist development conditions, substrate material, electron beam acceleration voltage, etc. Therefore, if even one of these conditions changes, the scattering intensity distribution must be recalculated each time.

The conventional method for determining the scattering intensity distribution equation is as follows.

In other words, as shown in Figure 1A, multiple line patterns are separated to the extent that the proximity effect does not affect them.
a ₁ , a ₂ , a ₃ ...a _i with different irradiation doses Q ₁ ,
After exposing the _sample to light with Q ₂ , Q ₃ . Obtain the side cross section as seen in the pattern a ₁ ,
Measure the bottom widths w ₁ , w ₂ , w ₃ ... w _i of a ₂ .... still,
In FIG. 1B, 1 is a substrate, and 2 is a photoresist film.

Next, the radiation dose Q versus pattern width w/2 is plotted to obtain the scattering intensity distribution curve shown in FIG. 1C. Then, the scattered intensity distribution equation is constructed by finding the coefficients of the assumed scattered intensity distribution equation using the least squares method from these measurement points.

However, in this method, the required dimension measurement point is the boundary between the resist film 2 and the substrate 1, and as seen in FIG. 1B, in the case of a positive resist, the dimensions of the resist pattern The bottom surface is larger than the top surface, and the side cross-sectional shape is trapezoidal, so it cannot be accurately determined by measuring only from the top surface. Therefore, as mentioned above, the bottom widths w ₁ and w ₂
In order to obtain..., it is necessary to at least cut the resist film 2. In addition, the dimensions must be measured using an electron microscope because they are in the submicron range.

As described above, determining the scattering intensity distribution of an electron beam is a task that requires a great deal of effort and time.

The present invention provides a pattern in which the pattern dimensions at the interface between the resist film and the substrate can be easily determined by observation from above, even if the cross-sectional shape of the resist film is as shown in FIG. The purpose of this invention is to make it possible to easily obtain the scattering intensity distribution using the pattern formation conditions, thereby shortening the measurement time of resist patterns and simplifying the work.This will be described in detail below.

In the present invention, the pattern shown in FIG. 2A is used in principle as the pattern used to derive the scattering intensity distribution.

In other words, a donut-shaped pattern with an inner diameter of 2w ₁
p ₁₁ , p ₁₂ ... p _1n in m horizontal directions, and n stages of patterns with different inner diameters in the vertical direction, for example, for the second stage, patterns p ₂₁ , p with inner diameters of 2w _{2 twenty two}
...Expose so that p _2n is arranged. At this time, the electron beam irradiation amount Q is set to be the same for objects in the same row. For example, for the leftmost example, Q ₁ ,
The second row from the left is exposed as Q ₂ , etc. Now, assuming that Q ₁ < Q ₂ ... < Q _n ,
If you develop each pattern and look at the cross section, you will see the second pattern.
As shown in Figure B.

As is clear _from FIG. 2B, for example, in the uppermost patterns p ₁₁ , p ₁₂ . _. .
If Q ₃ is set, the resist film remains in the center, but if the irradiation dose is set to Q ₄ , there will be overexposure and the resist in the center will completely disappear. And the irradiation dose Q ₃ is between Q ₂ and Q ₄ ,
The center of the resist film appears to be supported by one point at the center of the pattern (actually, the center falls and is removed). Irradiation amount at this time Q ₃
Let be the critical dose for this pattern. Furthermore, it can be seen that in the second-stage patterns p ₂₁ , p ₂₂ . . . , the irradiation amount Q ₄ for pattern p ₂₄ has become the critical irradiation amount.

Now, when the pattern shown in FIG. 2B is observed from above, it can be easily determined whether there is a center part in the resist film. That is, specifically, the resist pattern is viewed from above, and the presence or absence of the center of the resist film is determined in patterns with different irradiation doses at each stage, starting with the pattern with the lowest irradiation dose. , the irradiation amount in the pattern where the central part disappears for the first time is taken as the critical irradiation amount at that stage.

Next, a case will be described in which a scattering intensity distribution formula is derived from the results obtained in the observation.

Normally, the critical dose Q _c at the center of the pattern
When resist melting energy level E
is expressed by the following formula.

E=Q _c・∫ _s f(r)ds...(1) Here, f(r) is an expression expressing the scattering intensity distribution of the minute beam at a point r away from the center, and s is the expression for the scattering intensity distribution of the minute beam is the irradiated area.

Generally, f(r) is often expressed as the sum of forward scattering and backward scattering of electrons, f(r)=e−(r/C) ² +ηe−(r/C _B
)(2).

In this equation (2), if C _f , η, G _B , and E are determined, the scattering intensity distribution can be derived.
That is, since there are four unknowns, at least four types of patterns are required. If approximation is to be performed using another function, different types of patterns will be used as many as the required number of unknowns.

Assuming a scattering intensity distribution equation as shown in equation (2) above, and assuming that the critical dose Q _ci for each pattern is obtained by observing four types of patterns, Q _c1 ∫ _s1 f(r ) ds=E 〓 Q _c4 ∫ _s4 f(r) ds=E Since each equation holds true, the unknowns C _f , η, C _B , and E can be found from this equation. in this way,
If the critical dose Q _c is known, the electron scattering intensity distribution can be easily determined.

By the way, the above explanation relates to the principle of the present invention, and specifically, it is preferable to use a rectangular pattern as shown in FIG. That is, in electron beam exposure, the basic pattern is a rectangle, so exposure of a circular pattern as shown in Figure 2 must be covered with a series of small rectangular patterns, and microscopically is not the circumference.

However, if you use the pattern in Figure 3,
The exposure of the pattern is performed accurately and, even more preferably, the calculations are simplified. That is, in this case, the conditional expression is as follows.

4・Q _i ∫ ^yi2 _yi1 ∫ ^xi2 _xi1 f(r)d
xdy=E Here, r=√ ² + ^2However , xi ₁ ≦x≦xi ₂ and yi ₁ ≦y≦yi _2In addition, in order to expose the pattern in Fig. 3, the same rectangular pattern represented by the solid line is exposed four times, so the hatched area is exposed twice as much, but this does not cause any practical problems. Furthermore, although the above embodiments have been described with respect to positive resists, they can be similarly applied to negative resists. In this case, the resist contacts at the center of a pattern such as a circle, square, polygon, etc., and the substrate surface The irradiation amount at the limit of coverage may be taken as the critical irradiation amount.

As can be seen from the above explanation, according to the present invention, the various coefficients, that is, the unknowns, of the equation for determining the electron scattering intensity distribution necessary for performing proximity effect correction in electron beam exposure can be calculated from the surface of an extremely simple pattern. Calculations can be made based on the critical irradiance obtained through observation, and the forward and backward scattering distances of the electron beam can be determined, allowing exposure to be performed with the optimum irradiance for each figure. be able to.

[Brief explanation of the drawing]

Figure 1A is an explanatory diagram of a conventional exposure pattern.
Figure B is a cross-sectional view taken along line A-A' in Figure 1A, Figure 1C is a line diagram showing the scattering intensity, and Figure 2A is a plan explanatory diagram showing an example of the exposure pattern in the present invention. , FIG. 2B is a sectional view of FIG. 2A, and FIG. 3 is a plan view showing another example of the exposure pattern in the present invention. In the figure, p ₁₁ , p ₁₂ ... are patterns, Q ₁ , Q ₂ ...
is the irradiation amount, w ₁ , w _{2 .} . . are the widths, 1 is the substrate, and 2 is the resist film.

Claims (1)

[Claims] 1. Coefficients necessary to solve the electron scattering intensity distribution equation for the process of drawing a predetermined number of circular, rectangular, polygonal, etc. annular patterns on a positive resist film with different irradiation doses. The size of the annular pattern is changed and repeated a number of times corresponding to the number of unknown numbers, and then development is performed and the central part of the annular pattern drawn by changing the irradiation amount is removed during irradiation in the first pattern. An electron beam exposure method characterized in that the unknown quantity is determined for each step, the determined irradiation amount is set as a critical irradiation amount, the unknown quantity is determined, and an electron beam scattering intensity distribution is obtained based on the unknown quantity, and appropriate exposure is performed. 2. The process of drawing a predetermined number of annular patterns in the form of circles, squares, polygons, etc. on a negative resist film with different doses corresponds to the number of unknowns, which are the coefficients necessary to solve the electron scattering intensity distribution equation. Repeat the process by changing the size of the annular pattern a number of times, then perform development and change the irradiation amount to determine the irradiation amount for the first pattern in which the base no longer appears in the center of the annular pattern drawn. 1. An electron beam exposure method characterized in that the unknown quantity is obtained by determining the irradiation amount for each time, and the obtained irradiation amount is set as a critical irradiation amount, and based on the unknown quantity, an electron beam scattering intensity distribution is obtained and proper exposure is performed.