JPS58130522A - Electron beam exposure device - Google Patents

Abstract

PURPOSE:To contrive a large improvement of drawn image pattern dimensions, by setting and keeping the ratio of the exposure amount due to the primary electron to that due to the re-reflected electron at a specific value or less. CONSTITUTION:The primary incident electron 4 is reflected, based on the cosine law, from a substrate 1 to a direction making an angle theta with the beam axis. The reflected electron 5 thereof is next reflected on the rear surface of the upper structural body 2, then advances to a direction making an angle theta with the normal of the structural body 2, and approximates when incidence is into the point 7 on the substrate 1. The rear surface of the structural body 2 is kept apart from the substrate 1, or the rear surface thereof is processed, or a substance satisfying the Formula described hereunder is deposited thereon, in order that the ratio of the exposed amount Qo being received by the substrate 1 due to the incidence of the electron beam 4 to the exposed amount QR being received by the substrate 1 due to the electron beam which is re-reflected on the rear surface of the structural body 2 becomes QR/Qo<=0.1. Thus, the pattern dimensional error due to the influence by the re-reflected electron can be reduced to an almost negligible value and drawn image patterns of good accuracy can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to an electron beam exposure apparatus for improving the precision of drawn patterns.

[Technical background of the invention and its problems]

Recently, electron beam rattan optical equipment has been used to form fine patterns on sample substrates such as semiconductor wafers and masks. , Proximity Effect Correction [0 (1) to (5) and the like are conventionally known as methods for correcting the proximity effect that need to be performed.

+11 Design dimensions [change 0 (
2) Change the exposure amount t-0(3) according to the drawing pattern.
) Film thickness of exposed sample [to be as thin as possible 0 (4)
Accelerating voltage of electron beam [up.

(5) Resist of exposed sample [Multi-layered resist]
Proximity effect research [While proceeding, the present inventors discovered that the reflected electrons from the sample substrate to be exposed are transmitted to the objective lens pole piece facing the scissors substrate, the aperture, the heat sink↓, the reflected electron detector, etc. (hereinafter these t After being reflected twice by the lower surface of the upper structure (generally referred to as the upper structure), the resist on the sample substrate to be exposed is exposed once again, and the effect is a non-negligible shadow*1 on the desired pattern dimensions to be microfabricated. Ta. The first factor is a characteristic diagram showing the dependence of the exposure pattern size (punched pattern size) on the exposure area, and the second factor is a characteristic diagram showing the dependence of the exposed pattern size (punched pattern size) on the exposure area ratio.

If the exposure area d is larger than 3[-angle], the pattern size at the center of the exposure area is constant, but if the exposure area d is larger than 3[-angle]
Lcvs angle] or less, the pattern size becomes smaller as the exposed area decreases. Also, the exposed area ratio is large!
The pattern size at the center of the exposure area becomes larger. In this way, the influence on twice-reflected electrons, that is, re-reflected electrons, is the same as the influence on the conventionally known proximity effect, and moreover, the pattern size area where the proximity effect is significant is larger than the number 10. [βm] A similar effect appears for the above patterns. In other words, it has been found that the effect of re-reflected electrons is as important a problem as the proximity effect.0 [Object of the Invention] The object of the present invention is to An object of the present inventors is to provide an electron beam exposure apparatus that can reduce the effects of re-reflected electrons and significantly improve drawing pattern accuracy. [Summary of the Invention] Analysis and experiments were conducted to confirm that the nine-pattern dimensional difference shown in FIG. 2 is an effect of re-reflected electrons. Figure 3 shows the model used in the analysis. In the figure, 1 is the sample substrate to be exposed, 2
J indicates the objective lens aperture, and the incident primary electron 4 is
Based on the cosine law, the beam axis and angle 01 from the substrate 1! It is reflected in the direction of t. The backscattered electrons 5 are then reflected by the lower surface of the upper structure 2, forming an angle mar with the normal to the upper structure 20.
The approximation is obtained by proceeding in the direction of 0 and re-entering point 7 on the substrate 1. Substrate 10 reflected electron coefficient tη1. Letting the reflected electron coefficient tη2 of the l partial structure 2, the ratio of the re-reflected electrons 6 re-entering the point 2 to the primary electrons is nl/n. Here, hI is the distance between the substrate 1 and the upper structure 2, nl is the number of primary electrons 4, and nl is the number of 60 re-reflected electrons. #! The effect of the re-reflected electrons 6 in the central part 9 of the exposure area 8, which is indicated by diagonal lines in FIG.
It is obtained by integrating the equation, and becomes as follows.

Here, η is the reflected electron coefficient of the objective lens aperture 2, do is the diameter of the beam axis aperture 1 of the upper structure 2, and h
! is the distance between the objective lens aperture 3 and the substrate 1. Further, N is the number of re-reflected electrons per unit area that re-enters the central portion 9 from the exposure region I, and No is the number of primary electrons per unit area of the exposure region 10. Above JI! Equation 2 shows the sum of the effect due to the objective lens aberdeer 3 and the effect due to the upper structure 2. As a result of numerical calculations, it was found that this equation can be approximated by the following equation.

Based on the fourth equation above, the substrate 1 [silicon (da + =
0.15), upper structure 2t-iron (η!=0.2
9), distance between upper structure 2 and substrate 1, txz (m)
and L1 time t), NI/No 10 exposure area dependence [shown in Figure 1ILb. In addition, the exposure dependence of pattern dimensions is as shown in JII6WJlc.
As you can see from the figure, the exposure amount.

When changes by ±10 (1G), the change in pattern dimensions is ±
It is about 0.05 (voice m). However, N, /'N
, is a large number at most, as shown in Figure 5, and cannot fully explain the pattern size difference shown in Figures 1 and 2.3 Therefore, the effect of re-reflected electrons is not expected. However, if reinterpreted based on the recent theory of the proximity effect, the exposure ability of the backscattered electrons is due to the fact that the energy of the backscattered electrons is of the first order. Since reflected electrons are smaller than electrons and incident obliquely on the resist, they are considered to be larger than the exposure ability of primary electrons. Therefore, the ratio of the exposure ability Qk due to re-reflected electrons to the exposure ability Qa due to primary electrons must be expressed by the following equation. .

However, η1' and η8' are 1 of the exposure ability due to backscattered electrons.
It is the ratio to the exposure ability by secondary electrons (hereinafter referred to as the effective backscattered electron coefficient), and in the case of silicon, η,' =
1, which is said to be 0.5 to 1.0, and in the case of iron, according to experiments by the present inventors, η/=05 to 1°0.

FIG. 7 shows the values obtained from Equation 45 and the values obtained from the experiments of the 11 present invention. In the figure, the solid line P is the conventional device, and the solid line Q is the theoretical O value when the upper structure is coated with beryllium [
It shows. Further, in the figure, marks O and Δ indicate experimental data obtained under the respective conditions. From this figure, by depositing beryllium on the upper structure, Q position/Q can be obtained.

It can be seen that becomes smaller. Beryllium backscattered electron coefficient η
le is 0.05@degree, but the effective backscattered electron coefficient ηn
, / is 0.25 to 0.51! It was degree.

ηBe'/ηle is η,'/η1. The reason why η,′/η is larger is because the reflected electrons from light elements have more energy than the primary electrons. Then, the theoretical value Q obtained from the K11k5 formula based on the effective reflected electron coefficient of beryllium dl・' = 0.25 to 0.5 is in good agreement with the two experimental data (△ mark). However, according to Section 7, the above dimensional difference cannot be achieved even if beryllium is coated on the lower surface of the upper structure. To achieve this, the distance between the substrate 1 and the upper structure is increased.

[By increasing the size or by processing the lower surface of the upper structure, the exposure amount Q due to re-reflected electrons becomes less than 1/'1o of the exposure amount Q due to primary electrons, that is, Q Q0≦O,l ・・・・・・・・・(6) must be maintained 0 The present invention focuses on this point and irradiates an electron beam onto a sample substrate to which atomized light is applied to form a desired surface on the substrate. In an electron beam exposure apparatus that exposes a pattern, the amount of exposure Q0 that the substrate receives due to the incidence of primary electrons, the pole piece of the objective lens that is reflected from the substrate and faces the skeleton substrate, the aperture, the backscattered electron detector, and The lower surface of the upper structure [covered] is adjusted so that the ratio of the exposure amount q to the substrate due to the re-reflected electrons reflected again on the lower surface of the 1-ha structure consisting of these surroundings becomes Q''/Qe≦01. Either separate it from the exposed sample substrate, process the lower surface of the upper structure, or deposit a material that satisfies the above formula on the lower surface of the upper structure.

〔Effect of the invention〕

According to the present invention, the exposure amount Q by primary electrons.

The exposure amount Ql due to re-reflected electrons and the O ratio are Q''/
By setting and maintaining Qo≦0.1, pattern dimension difference t±0.05Cμm due to re-reflected electrons QIK
], and the drawing pattern precision can be significantly improved. For this reason, even when performing microfabrication with a pattern size of several μm or less, exposure can be performed with sufficient accuracy, which is extremely effective in microfabrication.

[Embodiments of the invention]

FIG. 8 is a diagram showing the main part of the first embodiment of the present invention. In this embodiment, the lower surface 12 of the objective lens facing the sample substrate 11 to be exposed is coated with a light element film 13'l such as beryllium or aluminum. At the same time, the distance 1111 between the substrate 11 and the lower surface 12 of the objective lens is set such that the above-mentioned equation jI6 holds true.

If the dimension of the sample substrate 11 to be exposed is k100 (asφ), then ks t-satisfying the formulas 5 and 6 above should be selected, '=0.
．． 5-1.0. If η,'=0.25=0.5, ≧! According to experiments conducted by the present inventors, the light element 1 [JJ] was deposited on the lower surface 12 of the objective lens, and the distance was set to = sO (sa). As a result, it was confirmed that the above-mentioned formula 6 holds true, and in this case, almost no difference in pattern dimensions due to the influence of re-reflected electrons was observed, and a highly accurate drawing pattern could be obtained. and Fig. 1θ are main part configuration diagrams showing the embodiment of @2, respectively.
The surface roughness of η increases by t, so that the sixth equation holds true.
O that is small and flat, that is, the lower surface 12 of the objective lens
In addition, the distance 1*h+ between the sample substrate 11 to be exposed and the lower surface 12 of the objective lens is 12 (ax), as shown in FIG. is coated with beryllium (not shown).

According to the above-mentioned FIG. 7, simply depositing beryllium on the lower surface 12 of the objective lens cannot obtain Q"/Qoζ of 0.25L. However, according to the experiments of the present inventors, the above-mentioned Keyaki of $14 Under the condition that the ratio of the width to the width is larger than 2, it was found that the ratio α of the area of the convexity of the groove 14 @ 14 m to the total area of the lower surface is approximately proportional to η, ○ Therefore, if -≦04 Said [6
formula? You can make friends. The solid line R and the mark ``l'' for the seventh factor indicate the theoretical value and experimental value t of jl2 according to the ninth factor together with beryllium evaporation. In the case of the groove 14, the same effect was obtained even in the case of the groove 14, which has a sawtooth shape as in JllO.In addition, the processing of the groove 14. This could be done by machining.Also, instead of providing the grooves 14 and 155f, if carbon, beryllium, or aluminum colloid particles were deposited on the lower surface 12 of the objective lens, the grooves 14 and l1
The same effect as when lk is provided can be obtained. Furthermore, it is also effective to form the objective lens lower surface 12f with a thin tube (bundled channel, plate-like shape, or nicum structure).
t.

FIG. 1111 is a diagram showing the main part of the third embodiment. In this embodiment, the lower surface of the objective lens 12 is processed into a convex lid. If the dimension of the portion 16 of the objective lens lower surface 12 parallel to the atomized light sample substrate 11 is [dL], then the effect corresponding to the fifth equation due to the re-reflected electrons is as follows. According to experiments conducted by the present inventors, the lower surface of the objective lens 1
2. The above-mentioned formula 6 can be established by setting dL≦10 (sue) under the conditions of beryllium coating, h and z12["s].

It should be noted that the present invention is not limited to the above-mentioned embodiments. All of the embodiments are applied to the lower surface of the objective lens, but in an actual apparatus, the lower surface of the objective lens and the exposed sample substrate A backscattered electron detector and a heat sink for controlling the temperature of the sample substrate to be exposed are placed on the surface of the substrate. Just do it. That is,
Is the lower surface of the upper structure facing the sample substrate to be exposed treated as shown in the example? Give and make sure that the 6th equation holds true.
It's okay.

In addition, although a silicon substrate is used as the sample substrate to be exposed in each of the examples, it may be applied to a compound semiconductor substrate such as GIAs (H), which has been attracting attention recently. In the case of (

[Brief explanation of drawings]

Figures 1 and 2 show the operation of the conventional device kii1, respectively.
Figure 1 is a characteristic diagram showing the dependence of the exposure pattern size on the exposure value range, and Figure 2 is a characteristic diagram showing the single light area dependence of the exposure pattern size. - Which of the following is a detailed explanation of the present invention? Figure 3 is a seven-chill O-type diagram used for the analysis of re-reflected electrons, and Figure 4 is a diagram of Tomonamitsu Yoshijo and Kao used in the model TT. Fig. 5 shows the exposure value range dependence of the re-reflection force ratio N, /No by the conventional 4-dimensional method, and Fig. 6 shows the dependence of the exposed pattern size on the exposure amount. FIG. 7 is a characteristic diagram showing the theoretical value and experimental value of the exposure ability ratio Q''/Qo of re-reflected electrons, and FIG. 8 is a characteristic diagram showing the main part configuration of the first embodiment of the present invention. 9 and 10 are main part configuration diagrams showing the second embodiment, respectively, and Figure 11 is a main part configuration diagram showing the third embodiment. 1.11... Fog light Sample substrate, 2...upper structure,
3...Objective lens aperture, 4...Primary electron,・
5... Reflected electron, 6... Re-reflected electron, 8... Exposure area, 9... Central part, 12... Lower surface of objective lens, 1
3... Light element film, 14.15... Groove. Tachito agent ff Landlord Suzue Takehiko Figure 1 Figure 2 Figure 4 Figure 5 Figure 6 1tlQO, +c/crn2) - Figure 7 Sentence Figure 8

Claims (1)

[Claims] (11) By irradiating the exposed sample substrate with an electron beam and
In an electron beam exposure device that exposes a desired pattern,
The exposure amount Q0 of the substrate due to the incidence of primary electrons, the exposure amount Q0 of the substrate due to the re-reflected electrons that are reflected from the substrate at this time and reflected again by the upper structure of the exposure device facing the wrinkled substrate. Electron beam exposure apparatus 0 in which the above exposure amount Qlk is made small so that the ratio of the above exposure amounts Qo=Q'' is 01 (Ql/Q)≦01 (2) The ratio of each of the above exposure amounts Qo=Q" The electron beam exposure apparatus according to claim 1 is characterized in that 'It' is the effective reflected electron coefficient of the sample substrate to be exposed, and 'Is' is the effective reflected electron coefficient of the upper structure. electronic coefficient, d
is the diameter of the exposure area on the sample substrate to be exposed, and h is the distance between the sample substrate to be exposed and the upper structure. (3) As a means for reducing the exposure amount Ql t due to the re-reflected electrons, the distance between the sample substrate to be exposed and the upper structure described in claim 1 or Jlli2 is characterized in that: Section 0 Electron beam exposure device. (4) To reduce the exposure amount Ql due to the re-reflected electrons, the upper structure may be formed of a material with a small effective backscattered electron coefficient, or the upper structure may be made of a material with a small
2. The electron beam exposure apparatus according to claim 2, wherein ii is coated with a substance having a small effective backscattered electron coefficient h'O. (5) As a means of reducing the fog light Q" caused by the re-reflected electrons, the surface of the first part structure is processed into an uneven groove shape or sawtooth shape, or a large number of holes are densely formed on the surface of the upper structure. The electron beam processing device according to claim 1 or 2, characterized in that the device is formed by integrating the A patent claim characterized in that the upper structure is formed into a convex structure, and where the diameter tdL of the planar portion at the tip of the convex portion is approximated by the ratio t of the fog light covers Qo and Q''. Range W!JI
The electron beam exposure apparatus according to section K2. (7) colloidal particles of carbon, beryllium, or aluminum are used as a substance with a small effective backscattered electron coefficient η,′ that adheres to the surface of the upper structure;
An electron beam exposure apparatus according to claim 4.