JP2001057169A - Optical element, astigmatism corrector, charged particle beam exposure apparatus, and method for manufacturing semiconductor device - Google Patents

Abstract

(57) [Summary] [Problem] Even when a large astigmatism correction output is output,
Provided is an electrostatic astigmatism corrector which does not generate aberration due to higher order components such as 6θ component and 10θ component. SOLUTION: Electrodes 1A, 1B, 1C, 1D, 1E, 1
F, 1G, 1H, 1I, 1J, 1K, 1L, 1M, 1
N, 10 and 1P are formed in a shape that is 16 times symmetric with respect to the optical axis. If the voltages applied to the electrodes 1A, 1B, ..., 1P are VA, VB, ..., VP, VA: VB: V
C: VD: VE: VF: VG: VH: VI: VJ: V
K: VL: VM: VN: VO: VP = + Vx: + (Vx
^{+ Vy) / 2 1/2} : + Vy :-( Vx-Vy) / 2 1/2:
-Vx:-(Vx + Vy) / ^21/2 : -Vy: + (Vx
-Vy) / ^21/2 : + Vx: + (Vx + Vy) / ^21/2 :
+ Vy:-(Vx-Vy) / 2 ^1/2 : -Vx:-(Vx
+ Vy) / 2 ^1/2 : −Vy: Voltages Vx and Vy are applied from two power sources so as to be + (Vx−Vy) / 2 ^1/2 .

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an electron beam exposure apparatus,
An optical system element used for a charged particle beam optical system of a charged particle beam apparatus such as an ion beam exposure apparatus, an electrostatic astigmatism corrector, and a charged particle beam exposure apparatus using the optical system element, and furthermore, The present invention relates to a method for manufacturing a semiconductor device using a charged particle beam exposure apparatus.

[0002]

2. Description of the Related Art In recent years, as the degree of integration of semiconductor integrated circuits has increased, finer circuit patterns have been required. With the required minimum circuit pattern width of 0.1 μm or less, the use of conventional optical exposure and transfer equipment has reached its limit. To overcome this, both high resolution and high throughput during exposure have been achieved. Investigation of a charged particle beam exposure system having both of the above has been advanced.

The method first developed for this purpose is the batch transfer method. In this method, one chip or a plurality of chips are exposed at a time. However, this method has recently been used recently because it is difficult to manufacture a mask serving as a master for transfer, and it is difficult to reduce aberration to a necessary degree within a large optical field such as one chip or more. , Its development is declining.

[0004] Instead of the batch transfer method, a method which has been studied recently is a method in which one cell or a plurality of cells are not exposed at a time, but are divided into small areas (sub-fields) of about several hundred μm square and exposed and transferred. This is generally called a division transfer method. At the time of exposure transfer, exposure is performed for each subfield while correcting aberrations such as the focus of an image formed on the surface to be exposed and the distortion of the deflection field. As a result, exposure with high resolution and accuracy can be performed over an optically wider area as compared with batch transfer.

FIG. 5 is a schematic view showing a transfer optical system of such an electron beam exposure apparatus. The mask 3 is irradiated by an illumination optical system (not shown), and the electron beam passing through the pattern thereon is imaged on the wafer by the two lenses 1 and 2, and the pattern on the mask 3 is reduced and transferred onto the wafer 4. . An aperture 5 for cutting scattered rays is provided between the lens 1 and the lens 2. The deflector 7
Two pieces of C1 and C2 are provided on the mask side of the aperture 5, and two pieces of P1 and P2 are provided on the wafer side of the aperture 5.

In these deflectors, an electron beam starting from a predetermined position on the mask 3 rides on an orbit 8 of the predetermined electron beam, passes through the aperture 5, and forms an image at a predetermined position on the wafer 4. In addition to deflecting the electron beam as described above, it has the function of removing image distortion and aberration. In the astigmatism corrector 9, one each of S1 and S2 is arranged on both sides of the aperture 5,
Eliminates deflection astigmatism and deflection hybrid astigmatism generated by deflection.

FIG. 6 shows an example of an electrostatic astigmatism corrector of a conventional charged particle beam apparatus. The figure shows a sectional view of the electrostatic astigmatism corrector cut along a section perpendicular to the optical axis.
Electrodes 1a, 1b, 1c, 1d of the electrostatic astigmatism corrector
Reference numerals 1e, 1f, 1g, and 1h denote eight-fold symmetric electrodes centered on the optical axis. Usually, the inner surface is obtained by dividing a cylindrical surface. In this specification, n-fold symmetry refers to a state in which a figure becomes the same as the original figure in any case where the figure is rotated n times 1 / n times around the rotation axis.

[0008] The material of each electrode is a metal or an insulator provided with a conductive film, and is given a voltage to create an electric field near the optical axis. Assuming that voltages applied to the electrodes 1a to 1h are Va to Vh, the voltages Va to Vh are, for example, Va: Vb: Vc for correcting a beam expanded and contracted in the directions of the electrodes 1a, 1c, 1e, and 1g due to astigmatism. : Vd: Ve: Vf: Vg: Vh = +
1: 0: -1: 0: +1: 0: -1: 0, and the electrodes 1b, 1b
Va: Vb: Vc: Vd: Ve: Vf: Vg: Vh =
It is applied so as to keep the ratio of 0: +1: 0: -1: 0: +1: 0: -1.

In general, the applied voltages Vx and Vy are changed in accordance with the direction and magnitude of astigmatism. Va: Vb: Vc: Vd: Ve: Vf: Vg: Vh = V
x: Vy: -Vx: -Vy: Vx: Vy: -Vx: -V
y. With this setting, the potential near the optical axis of the astigmatism corrector can be calculated using φ [r, θ, z] ≒ g [z] r ² cos [2 ( θ-χ)].

Here, g [z] is an on-axis astigmatism field generated by the astigmatism corrector, where a is the inner diameter of the astigmatism corrector, and θo is the potential angle of the electrode of the electrostatic astigmatism corrector. z] = 4 sin (θ ₀ ) (Vx ² + Vy ² ) ^1/2 / (πa ² ), and χ is calculated as χ = tan ^-1 (Vy / Vx) / 2 Angle.

[0011]

In a recent charged particle beam apparatus, for example, a split transfer type electron beam exposure apparatus, a reticle portion (subfield) of about 1 mm square is reduced to 1/4 and transferred onto a wafer. I do. Further, the electron beam for illuminating the reticle is deflected about 20 mm on the reticle surface, and the electron beam for exposing the wafer is correspondingly deflected about 5 mm on the wafer surface. In such a charged particle beam optical system in which the size and deflection width of an image transferred at a time are large, astigmatism increases. Therefore, it is necessary to increase the output of the astigmatism corrector accordingly to cancel large astigmatism.

However, in the conventional electrostatic astigmatism corrector as shown in FIG. 6, when the deflection width is increased, a higher-order field, such as the 6θ component, which is not a problem in the range where the deflection width is small, is not required. Becomes large, and there is a problem that an aberration is generated due to this. The 6θ component is a potential near the optical axis represented by φ [r, θ, z] ≒ k [z] r ⁶ cos [6 (θ−χ)]. Here, k [z] is the 6θ field on the axis generated by the astigmatism corrector, where a is the inner diameter of the astigmatism corrector, and θo is the expected angle of the electrode of the electrostatic astigmatism corrector. 4sin [3θ ₀ ] (Vz ² + Vy ² ) ^1/2 / (3πa ⁶ ), χ is the angle calculated by χ = tan ^-1 (Vy / Vx) / 6 .

The present invention has been made in view of such circumstances. Even when a large output is output, the 6θ component, 10
The main object is to provide an optical system element which does not generate aberration due to a higher order component such as a θ component, and furthermore, an electrostatic astigmatism corrector using the same, a charged particle beam exposure apparatus having a small aberration, An object of the present invention is to provide a method for manufacturing a semiconductor device using a particle beam exposure apparatus.

[0014]

A first means for solving the above problems is an optical system element used in a charged particle beam optical system, which is 16 times symmetric with respect to the optical axis of the charged particle beam optical system. An optical system element having a simple electrode (claim 1).

In this means, since 16 electrodes are provided which are 16 times symmetrical with respect to the optical axis of the charged particle beam optical system, the voltage applied to each electrode is different from the conventional case of eight electrodes. By adjusting, even when a large output (for example, an astigmatism correction output) is output, the higher-order component can be reduced, and thus the aberration caused by the higher-order component can be reduced.

A second means for solving the above-mentioned problem is as follows.
In the first means, the electrodes are arranged in the order of one round with respect to the optical axis with respect to an arbitrary one electrode, and approximately: +1:
+2 ^-1/2 : 0: -2 ^-1/2 : -1: -2 ^-1/2 : 0: +2
^-1/2: +1 +2 ^{-1 /} 2: 0: -2 ^-1/2: -1: -
A first voltage having a voltage ratio of 2 ^-1/2 : 0: +2 ^-1/2 is applied and superimposed on the first voltage, and an electrode at a position rotated 45 degrees with respect to the previous reference electrode is used as a reference. In the order of one round with respect to the optical axis, approximately +1: +2 ^−1/2 : 0: −
^{2-1 / 2} : -1: -2 ^-1/2 : 0: +2 ^-1/2 : +1: +2 ^{-1
/ 2} : 0: -2 ^-1/2 : -1: -2 ^-1/2 : 0: +2 ^-1/2. 2).

In a cylindrical coordinate system with the optical axis as the z axis, the Laplace equation of the electric field φ [r, θ, z] near the optical axis is

[0018]

(Equation 1)

When φ is Fourier-expanded to θ, θ is m
The cos [mθ] component of the next term is φc [m, r, z], and the sin [mθ] component is φ
Assuming that s [m, r, z], φ [r, θ, z] is

[0020]

(Equation 2)

Φc [m, r, z] and φs [m, r, z] are Taylor-expanded in the r direction.

[0022]

(Equation 3)

Then, the potential φ becomes

[0024]

(Equation 4)

## EQU2 ## Here, by substituting the equation (5) into the equation (1), the coefficient of r becomes equal to 0.
s ^{terms, m 2 c [m, 0} , z] = 0 (1-m 2) c [m, 1, z] = 0 (4-m 4) c [m, 2, z] + c "[ m, 0, z] = 0 A similar relationship is obtained for the term of sin.

It is known that m = 2, 6, 10,... Occur for a symmetric device such as an astigmatism corrector. That is, ideally, only the term of m = 2 is generated, but the term of m = 6, 10,...
It occurs parasitically due to the angle configuration. The ideal astigmatism field is φ [r, θ, z] = (r ² c [2,2, z] -r ⁴ c "[2,2, z] / 12 +…) cos2θ +
(r ² s [2,2, z] -r ⁴ s "[2,2, z] / 12 + ...) cos2θ. Similarly, the term corresponding to 6θ is φ [r, θ, z] = (r ⁶ c [6,6, z] -r ⁸ c "[6,6, z] / 28 +…) cos6θ +
(r ⁶ s [6,6, z] -r ⁸ s "[6,6, z] / 28 + ...) cos6θ.

Here, c [m, n, z] and s [m, n, z] are obtained. First, when cosm ₀ θ is multiplied by equation (5) and integrated by θ,

[0028]

(Equation 5)

Is obtained. Since the lowest order term of the field in the astigmatism corrector is 2, from equation (6),

[0030]

(Equation 6)

From this, assuming that the inner diameter of the astigmatism corrector is a ₀ ,

[0032]

(Equation 7)

Is obtained as follows.

Similarly, for higher order astigmatism, 6θ
The ingredients are

[0035]

(Equation 8)

The 10θ component is

[0037]

(Equation 9)

Is calculated. For s [m, n, z],
If cos is changed to sin, it can be obtained in the same way.

Here, consider the conventional astigmatism corrector shown in FIG. That is, the electrode is symmetrical eight times,
There are two electrodes on the x-axis, and the angle at which one electrode is viewed from the optical axis is θ ₀ . Then, φ [a ₀ , θ, z] = + Vx (θ = -θo / 2 ~ + θo / 2) = + Vy (θ = π / 4 -θo / 2 ~ π / 4 + θo / 2) = -Vx (θ = 2π / 4 -θo / 2 ~ 2π / 4 + θo / 2) =-Vy (θ = 3π / 4 -θo / 2 ~ 3π / 4 + θo / 2) = + Vx (θ = 4π / 4 -θo / 2 to 4π / 4 + θo / 2) = + Vy (θ = 5π / 4 -θo / 2 to 5π / 4 + θo / 2) =-Vx (θ = 6π / 4 -θo / 2 ~ 6π / 6 + θo / 2) =-Vy (θ = 7π / 4-θo / 2 ~ 7π / 4 + θo / 2) ... (11), and the electric field φ is 0 in other θ ranges .

Of these, the non-dotted field c [m, m, z] due to Vx is c
Assuming x [m, m, z], from equation (6), cx [m, m, z] = (Vx / m) {+ sin [m (π 0/4 + θo / 2)] − sin [ m (π 0/4 -θo / 2)]-s in [m (π 2/4 + θo / 2)] + sin [m (π 2/4 -θo / 2)] + sin [m (π 4 / 4 + θo / 2)]-sin [m (π 4/4 -θo / 2)]-sin [m (π 6/4 + θo / 2)] + sin [m (π 6/4 -θo / 2)]} / (πa ₀ ^m ) (12)

Therefore, cx [2,2, z] = 4 Vx sin [θo] / (πa ₀ ² ) (13) cx [6,6, z] = (4/3) Vx sin [θo] / (πa ₀ ⁶ )… (14) cx [10,10, z] = (4/5) Vx sin [θo] / (πa ₀ ¹⁰ )… (15) Non-point field cy [m, m, z by Vy ] In the same way, cy [2,
Since 2, z] = cy [6,6, z] = cy [10,10, z] = 0, c [2,2, z] = 4 Vx sin [θo] / (πa ₀ ² )… (16) c [6,6, z] = (4/3) Vx sin [3θo] / (πa ₀ ⁶ )… (17) c [10,10, z] = (4/5) Vx sin [5θo] / (πa ₀ ¹⁰ ) (18) is obtained.

Similarly, for s [m, m, z], s [2,2, z] = 4 Vy sin [θo] / (πa ₀ ² ) (19) s [6,6, z] = ( 4/3) Vy sin [3θo] / (πa ₀ ⁶ )… (20) s [10,10, z] = (4/5) Vy sin [5θo] / (πa ₀ ¹⁰ )… (21) The 6θ component and the 10θ component of the non-point field remain.

On the other hand, in the present means, 16 electrodes are provided 16 times symmetrically, and the center of two electrodes is on the X-axis, and the angle at which one electrode is seen from the axis is set. θ _{0 is} assumed. When the first voltage Vx and the second voltage are Vy, the electric field due to the first voltage Vx is φ [a ₀ , θ, z] = + Vx (θ = −θo / 2 to + θo / 2) ) = + Vx / √2 (θ = π / 8 -θo / 2 ~ π / 8 + θo / 2) =-Vx / √2 (θ = 3π / 8 -θo / 2 ~ 3π / 8 + θo / 2 ) =-Vx (θ = 4π / 8 -θo / 2 to 4π / 8 + θo / 2) =-Vx / √2 (θ = 5π / 8 -θo / 2 to 5π / 8 + θo / 2) = + Vx / √2 (θ = 7π / 8 -θo / 2 to 7π / 8 + θo / 2) = + Vx (θ = 8π / 8 -θo / 2 to 8π / 8 + θo / 2) = + Vx / √ 2 (θ = 9π / 8 -θo / 2 to 9π / 8 + θo / 2) =-Vx / √2 (θ = 11π / 8 -θo / 2 to 11π / 8 + θo / 2) =-Vx (θ = 12π / 8 -θo / 2 to 12π / 8 + θo / 2) =-Vx / √2 (θ = 13π / 8 -θo / 2 to 13π / 8 + θo / 2) = + Vx / √2 (θ = 15π / 8-θo / 2 to 15π / 8 + θo / 2) (22), and the electric field φ is 0 when θ is in the other range.

Therefore, the astigmatism field cx [m, m, z] by Vx is
From equation (6), cx [m, m, z] = (Vx / m) {(sin [mθo / 2]-sin [m-θo / 2]) + (sin [m (π / 8 + θo / 2) )]-sin [m (π / 8 -θo / 2)] / √2-(sin [m (3π / 8 + θo / 2)]-sin [m (3π / 8 -θo / 2)]) / √2-(sin [m (4π / 8 + θo / 2)]-sin [m (4π / 8 -θo / 2)])-(sin [m (5π / 8 + θo / 2)]-sin [m (5π / 8 -θo / 2)]) / √2 + (sin [m (7π / 8 + θo / 2)]-sin [m (7π / 8 -θo / 2)]) / √2 + (sin [m (8π / 8 + θo / 2)]-sin [m (8π / 8 -θo / 2)]) + (sin [m (9π / 8 + θo / 2)]-sin [m (9π / 8 -θo / 2)]) / √2-(sin [m (11π / 8 + θo / 2)]-sin [m (11π / 8 -θo / 2)]) / √2-(sin [m (12π / 8 + θo / 2)]-sin [m (12π / 8 -θo / 2)])-(sin [m (13π / 8 + θo / 2)]-sin [m (13π / 8 -θo / 2)]) / √2 + (sin [m (15π / 8 + θo / 2)]-sin [m (15π / 8 -θo / 2)]) / √2)} / (πa ₀ ^m )… (23)

From this, cx [2,2, z] = 8 Vx sin [θo] / (πa ₀ ² ) (24) cx [6,6, z] = 0 (25) cx [10,10 , z] = 0 (26) is obtained. When the applied voltage Vy is similarly calculated, cy [2,2, z] = cy [6,6, z] = cy [10,10, z] = 0,
After all, c [2,2, z] = 8 Vx sin [θo] / (πa ₀ ² )… (27) c [6,6, z] = 0… (28) c [10,10, z] = 0… (29) is obtained.

Similarly, for s [m, m, z], s [2,2, z] = 8 Vy sin [θ ₀ ] / (πa ₀ ² ) (30) s [6,6, z] = 0 ... (31) s [10,10, z] = 0 (33), and 6θ, 10 even if the applied voltages Vx and Vy are increased.
The θ component remains at 0. Therefore, no higher-order component of the on-axis astigmatism field is generated.

Further, comparing the expressions (16) and (27), since θ ₀ = π / 4 in the expression (16), the on-axis astigmatic field is 2.2 · ^1/2 Vx / ( πa ₀ ² ), and θ ₀ = π / 8 in equation (27).
2 ^1/2 ) ^1/2 Vx / (πa ₀ ² ). Therefore, in this means, when the same voltage is used, 4 ・ (2-2 ^1/2 ) ^1/2 / (2 ・ 2 ^1/2 )
Since it is 81.08, it is about 8 times as compared with the conventional astigmatism corrector.
% Large astigmatism can be generated. The same result can be obtained by comparing the expressions (19) and (30).

In the present means, the fact that the ratio of the voltage applied to each electrode is set to "approximately" a predetermined value means that it is not always necessary to set the ratio exactly to the predetermined ratio. A person skilled in the art can easily calculate the degree of deviation from the predetermined value based on the magnitude of the higher-order component of the on-axis astigmatism that is allowed.

A third means for solving the above-mentioned problem is as follows:
In the first means, the electrodes are arranged in the order of one round with respect to the optical axis with respect to an arbitrary one electrode, and approximately: +1:
+2 ^1/2 -1: -2 ^1/2 +1: -1: -1: -2 ^1/2 +
1: +2 ^1/2 -1: +1: +1: +2 ^1/2 -1: -2 ^1/2
A first voltage having a voltage ratio of +1: -1: -1: -2 ^1/2 +1: +2 ^1/2 -1: +1 is applied, superimposed thereon, and applied to the above-mentioned reference electrode. With reference to the electrode at the position rotated 45 degrees,
Almost +1: +2 ^1/2 -1: -2 ^1/2 +1: -1: -1:
−2 ^1/2 +1: +2 ^1/2 −1: +1: +1: +2 ^1/2 −
1: -2 ^1/2 +1: -1: -1: -2 ^1/2 +1: +2 ^1/2
The present invention is characterized in that a second voltage having a voltage ratio of -1: +1 is applied (claim 3).

In this case, there is no electrode in the x-axis direction,
The calculation is performed assuming that the electrode center is placed at a position forming an angle of π / 16 with the x axis. The signs and the like used for the calculation are the same as those used in the description of the second means. Then, the electric field due to the first voltage Vx is: φ [ao, θ, z] = + Vx (θ = π / 16−θo / 2 to π / 16 + θo / 2) = + Vx (√2-1) (θ = 3π / 16 -θo / 2 to 3π / 16 + θo / 2) =-Vx (√2-1) (θ = 5π / 16 -θo / 2 to 5π / 16 + θo / 2) =-Vx (θ = 7π / 16 -θo / 2 ~ 7π / 16 + θo / 2) =-Vx (θ = 9π / 16 -θo / 2 ~ 9π / 16 + θo / 2) =-Vx (√2-1) (θ = 11π / 16 -θo / 2 to 11π / 16 + θo / 2) = + Vx (√2-1) (θ = 13π / 16 -θo / 2 to 13π / 16 + θo / 2) = + Vx (θ = 15π / 16 -θo / 2 ~ 15π / 16 + θo / 2) = + Vx (θ = 17π / 16 -θo / 2 ~ 17π / 16 + θo / 2) = + Vx (√2-1) (θ = 19π / 16 -θo / 2 to 19π / 16 + θo / 2) =-Vx (√2-1) (θ = 21π / 16 -θo / 2 to 21π / 16 + θo / 2) =-Vx (θ = 23π / 16 -θo / 2 ~ 23π / 16 + θo / 2) =-Vx (θ = 25π / 16 -θo / 2 ~ 25π / 16 + θo / 2) =-Vx (√2-1) (θ = 27π / 16 -θo / 2 to 27π / 16 + θo / 2) = + Vx (√2-1) (θ = 29π / 16 -θo / 2 to 29π / 16 + θo / 2) = + Vx (θ = 31π / 16−θo / 2 to 31π / 16 + θo / 2) (34), and θ is 0 in other ranges.

Therefore, c [m, m, z] = (Vx / m) {(Sin [m (π / 16 + θo / 2)] − Sin [m (π / 16−θo / 2)]) + (Sin [m (3π / 16 + θo / 2)]-Sin [m (3π / 16 -θo / 2)]) (√2-1)-(Sin [m (5π / 16 + θo / 2)] -Sin [m (5π / 16 -θo / 2)]) (√2-1)-(Sin [m (7π / 16 + θo / 2)]-Sin [m (7π / 16 -θo / 2)] )-(Sin [m (9π / 16 + θo / 2)]-Sin [m (9π / 16-θo / 2)])-(Sin [m (11π / 16 + θo / 2)]-Sin [m (11π / 16-θo / 2)]) (√2-1) + (Sin [m (13π / 16 + θo / 2)]-Sin [m (13π / 16 -θo / 2)]) (√2 -1) + (Sin [m (15π / 16 + θo / 2)]-Sin [m (15π / 16 -θo / 2)]) + (Sin [m (17π / 16 + θo / 2)]-Sin [m (17π / 16 -θo / 2)]) + (Sin [m (19π / 16 + θo / 2)]-Sin [m (19π / 16 -θo / 2)]) (√2-1)- (Sin [m (21π / 16 + θo / 2)]-Sin [m (21π / 16 -θo / 2)]) (√2-1)-(Sin [m (23π / 16 + θo / 2)] -Sin [m (23π / 16 -θo / 2)])-(Sin [m (25π / 16 + θo / 2)]-Sin [m (25π / 16 -θo / 2)])-(Sin [m (27π / 16 + θo / 2)]-Sin [m (27π / 16 -θo / 2)]) (√2-1) + (Sin [m (29π / 16 + θo / 2)]-Sin [m (29π / 16 -θo / 2)]) (√2-1) + (Sin [m (31π / 16 + θo / 2)]-Sin [m (31π / 16 -θo / 2)])} / ( πa ₀ ^m ) (35)

From this, c [2,2, z] = 8 Vx (4-2 · 2 ^1/2 ) ^1/2 sin [θo] / (πa ₀ ² ) (36) c [6,6, z] = 0… (37) c [10,10, z] = 0… (38) s [2,2, z] = 8 Vy (4-2 · 2 ^1/2 ) ^1/2 sin [θo] / (πa ₀ ² )… (39) s [6,6, z] = 0… (40) s [10,10, z] = 0… (41) Even if the applied voltages Vx and Vy are increased 6θ, 10
The θ component remains at 0. Therefore, no higher-order component of the on-axis astigmatism field is generated.

When Equations (16) and (36) are compared, since θ ₀ = π / 8 in Equation (36), the on-axis astigmatic field is 8 (2 ^1/2 -1) Vx / (πa ₀ ² ).
When the same voltage is used, since 8 (2 ^1/2 -1) / (2 · 2 ^1/2 ) ≒ 1.17, the astigmatism field which is about 17% larger than that of the conventional astigmatism corrector is obtained. Can be generated. Equation (19) and (3
The same applies to the case where the expressions 9) are compared. "Almost"
Has the same meaning as described in the second means.

A fourth means for solving the above-mentioned problem is:
An electrostatic astigmatism corrector (claim 4) comprising an optical system element of any one of the first to third means.

By using each optical system element as an electrostatic astigmatism corrector, even if a large astigmatism is generated to correct a large astigmatism, no 6θ component and 10θ component are generated. Large astigmatism can be corrected without generating aberration.

A fifth means for solving the above problems is as follows.
A charged particle beam exposure apparatus (Claim 5), comprising an optical system element of any one of the first to third means.

Since this means has an optical system element in which a high-order component of the astigmatism field hardly occurs even when a large voltage is applied, it is possible to correct the astigmatism even when a large deflection amount is applied. it can. Therefore, the aberration can be reduced, and the size of the main field can be increased, so that the throughput of the exposure transfer can be increased.

A sixth means for solving the above-mentioned problem is:
Using the charged particle beam exposure apparatus as the fifth means,
A method for manufacturing a semiconductor device, comprising a step of transferring a pattern formed on a mask or a reticle onto a wafer (claim 6).

According to this means, since a charged particle beam exposure apparatus having a small aberration and a high throughput is used, a semiconductor device having a fine line pattern can be efficiently manufactured.

[0060]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a schematic view when an optical system element according to a first embodiment of the present invention is used as an electrostatic astigmatism corrector, and is a cross-sectional view cut along a cross section perpendicular to the optical axis. . In FIG. 1, electrodes 1A, 1B, 1C,
1D, 1E, 1F, 1G, 1H, 1I, 1J, 1K, 1
L, 1M, 1N, 10 and 1P are formed in shapes that are 16 times symmetric with respect to the optical axis. That is, the shape of each electrode is the same, and is arranged every π / 8. Further, the inner surfaces of the electrodes are arranged on the same circumference. That is, the inner surface of each electrode is obtained by dividing the cylindrical surface. For convenience of explanation, the two electrodes are arranged so as to be on the x-axis. However, it is obvious from the nature of the astigmatism corrector that this is not always necessary.

If the voltages applied to the electrodes 1A, 1B, ..., 1P are VA, VB, ..., VP, VA: VB: VC: VD: VE: VF: VG: VH: V
I: VJ: VK: VL: VM: VN: VO: VP = + V
x: + (Vx + Vy) / ^21/2 : + Vy:-(Vx-V
y) / ^21/2 : -Vx:-(Vx + Vy) / ^21/2 : -V
y: + (Vx-Vy) / 2 ^1/2 : + Vx: + (Vx + V
y) / ^21/2 : + Vy:-(Vx-Vy) / ^21/2 : -V
x:-(Vx + Vy) / ^21/2 : -Vy: + (Vx-V
y) / 2 so that ^half, giving two of the power supply voltage Vx, the Vy. In such a case, the on-axis astigmatism field is as shown in the equations (27) to (33), and the 6θ component and the 10θ component, which are higher-order components of the astigmatism field, are not generated. Further, as described above, an astigmatism field that is about 8% larger than that of the conventional electrostatic astigmatism corrector can be generated.

FIG. 2 is a schematic diagram of an optical system element according to a second embodiment of the present invention, which is also an example in which the optical system element is used as an electrostatic astigmatism corrector. It is sectional drawing cut | disconnected by the cross section perpendicular | vertical to an optical axis. In FIG. 2, electrodes 1A, 1B, 1
C, 1D, 1E, 1F, 1G, 1H, 1I, 1J, 1
K, 1L, 1M, 1N, 10 and 1P are 16
It is created in a symmetrical shape. That is, the shape of each electrode is the same, and is arranged every π / 8. Further, the inner surfaces of the electrodes are arranged on the same circumference. That is, the inner surface of each electrode is obtained by dividing the cylindrical surface. In this case, the arrangement is such that the x-axis is at the center of the electrodes 1A and 1P, and 1H and 1I.

When the voltages applied to the electrodes 1A, 1B,..., 1P are VA, VB,.
C: VD: VE: VF: VG: VH: VI: VJ: V
K: VL: VM: VN: VO: VP = + Vx + (2 ^1/2
-1) Vy: (2 ^1/2 -1) Vx + Vy: (1-2 ^1/2 )
Vx + Vy: -Vx + (2 ^1/2 -1) Vy: -Vx-
(2 ^1/2 -1) Vy: (1-2 ^1/2 ) Vx-Vy: (2
^1/2 -1) Vx-Vy: + Vx + (1-21 ^{/ 2} ) Vy: +
Vx + (2 ^1/2 -1) Vy: (2 ^1/2 -1) Vx + Vy:
^{(1-2 1/2) Vx + Vy:} -Vx + (2 1/2 -1) V
y: -Vx- (21 ^/ 2-1) Vy: (1-21 ^{/ 2} ) Vx-
Vy: (2 ^1/2 -1) Vx-Vy: + Vx + (1-
2 ^1/2 ) Vy is applied from the power supply.

In such a case, the on-axis astigmatism field is expressed by (3)
From equation 6) to equation (41), the 6θ component and the 10θ component, which are higher-order components of the astigmatism field, are not generated. Further, as described above, an astigmatism field that is about 17% larger than that of the conventional electrostatic astigmatism corrector can be generated.

The present optical system member can be used, for example, as the astigmatism corrector 9 for the charged particle beam exposure apparatus shown in FIG. Not only used but also other charged particle beam optical members having a similar shape and function, for example, a Q lens (4
(Polarizer lens).

Hereinafter, a method of manufacturing a semiconductor device using a charged particle beam exposure apparatus using such an optical member will be described. FIG. 3 is a flowchart showing an example of the semiconductor device manufacturing method of the present invention. The manufacturing process of this example includes the following main processes. Wafer manufacturing process for manufacturing a wafer (or wafer preparing process for preparing a wafer) Mask manufacturing process for manufacturing a mask to be used for exposure (or mask preparing process for preparing a mask) Wafer processing for performing necessary processing on a wafer Steps Chips formed on a wafer are cut out one by one to make them operable Chip assembling step Chip inspection step to inspect the chips that have been made Each of these steps further comprises several sub-steps.

Among these main steps, the main step that has a decisive effect on the performance of the semiconductor device is the wafer processing step. In this step, designed circuit patterns are sequentially laminated on a wafer, and a number of chips that operate as a memory or an MPU are formed. This wafer processing step includes the following steps. A thin film forming step (using CVD, sputtering, etc.) for forming a dielectric thin film, a wiring portion, or a metal thin film for forming an electrode portion, which serves as an insulating layer. A lithography process of forming a resist pattern using a mask (reticle) in order to selectively process etc. An etching process of processing a thin film layer or a substrate according to a resist pattern (for example, using a dry etching technique) An ion / impurity implantation diffusion process Resist stripping step Inspection step of inspecting the processed wafer Further, the wafer processing step is repeated by a necessary number of layers to manufacture a semiconductor device that operates as designed.

FIG. 4 is a flowchart showing a lithography step which is the core of the wafer processing step shown in FIG. This lithography step includes the following steps. A resist coating step of coating a resist on a wafer on which a circuit pattern has been formed in the preceding step An exposing step of exposing the resist A developing step of developing the exposed resist to obtain a resist pattern Stabilizing the developed resist pattern The above-described semiconductor device manufacturing process, wafer processing process, and lithography process are well known, and will not require further explanation.

By using a charged particle beam exposure apparatus using the optical system element according to the present invention in the above exposure step, the pattern on the mask and reticle can be accurately transferred onto the wafer, and the fine pattern can be transferred. Can be manufactured with good yield.

[0070]

As described above, according to the first aspect of the present invention, since the electrodes having 16 times symmetry with respect to the optical axis of the charged particle beam optical system are provided, the number of electrodes is eight as in the prior art. By adjusting the voltage applied to each electrode, unlike in the case where the electrodes are formed, even when a large astigmatism correction output is output, the high-order component can be reduced, and therefore, the aberration caused by the high-order component is reduced. can do.

According to the second aspect of the present invention, since a voltage having an appropriate ratio is applied to each of the 16-fold symmetrical electrodes, 6θ and 10θ components, which are higher-order components of the astigmatism field, do not occur. Therefore, it is possible to reduce aberrations caused by higher order components. Further, an astigmatism correction output that is about 8% larger than that of the conventional astigmatism corrector can be output.

According to the third aspect of the present invention, since a voltage having an appropriate ratio is applied to each of the 16-fold symmetrical electrodes, 6θ and 10θ components, which are higher-order components of the astigmatism field, do not occur. Therefore, it is possible to reduce aberrations caused by higher order components. Further, an astigmatism correction output that is about 17% larger than that of the conventional astigmatism corrector can be output.

In the invention according to claim 4, even if a large astigmatism field is generated to correct a large astigmatism,
Since the 6θ component and the 10θ component are not generated, large astigmatism can be corrected without generating aberration.

In the invention according to claim 5, even if a large amount of deflection is given, the aberration can be reduced. Also,
Along with this, the size of the main field can be increased, so that the throughput of exposure transfer can be increased.

In the sixth aspect of the present invention, since a charged particle beam exposure apparatus having small aberration and high throughput is used, a semiconductor device having a fine line pattern can be manufactured efficiently.

[Brief description of the drawings]

FIG. 1 is a schematic diagram of an optical system element according to a first embodiment of the present invention.

FIG. 2 is a schematic diagram of an optical system element according to a second embodiment of the present invention.

FIG. 3 is a flowchart illustrating an example of a semiconductor device manufacturing method according to the present invention.

FIG. 4 is a flowchart illustrating a lithography process.

FIG. 5 is a diagram showing an outline of an example of a charged particle beam exposure apparatus.

FIG. 6 is a schematic diagram of a conventional electrostatic astigmatism corrector.

[Explanation of symbols]

1A, 1B, 1C, 1D, 1E, 1F, 1G, 1H, 1
I, 1J, 1K, 1L, 1M, 1N, 1O, 1P ... 16
Electrodes of polar electrostatic astigmatism corrector 1a, 1b, 1c, 1d, 1e, 1f, 1g, 1h ... 8
Electrodes of polar electrostatic astigmatism corrector

Claims (6)

[Claims] 1. An optical system element used in a charged particle beam optical system, comprising an electrode which is 16 times symmetrical with respect to the optical axis of the charged particle beam optical system. 2. The optical system element according to claim 1, wherein
With reference to an arbitrary electrode, the electrodes are arranged in the order of one round with respect to the optical axis, and approximately +1: +2 ^−1/2 : 0: −2 ^−1/2 : −
1: -2 ^-1/2: 0: +2 ^-1/2: +1 +2 ^{-1 /} 2: 0: -
A first voltage having a voltage ratio of 2 ^-1/2 : -1: -2 ^-1/2 : 0: +2 ^-1/2 is applied, and further superimposed on the first voltage, with respect to the above-mentioned reference electrode. With reference to the electrode at a position rotated by 45 degrees, the electrodes are substantially +1:
+2 ^-1/2 : 0: -2 ^-1/2 : -1: -2 ^-1/2 : 0: +2
^-1/2: +1 +2 ^{-1 /} 2: 0: -2 ^-1/2: -1: -
An optical system element characterized in that a second voltage having a voltage ratio of 2 ^-1/2 : 0: +2 ^-1/2 is applied. 3. The optical element according to claim 1, wherein:
With reference to an arbitrary electrode, the electrodes are arranged in the order of making a round with respect to the optical axis, and approximately +1: +2 ^1/2 -1: -2 ^1/2 +1:
-1: -1: -2 ^1/2 +1: +2 ^1/2 -1: +1: +1:
+2 ^1/2 -1: -2 ^1/2 +1: -1: -1: -2 ^1/2 +
A first voltage having a voltage ratio of 1: +2 ^1/2 -1: +1 is applied and superimposed thereon, and light is applied with reference to the electrode at a position rotated 45 degrees with respect to the previous reference electrode. In order of making one round with respect to the axis, approximately +1: +2 ^1/2 -1:
-1 ^/ 2 + 1: -1: -1: -21 ^/ 2 + 1: + 21 ^/ 2-
1: +1: +1: +2 ^1/2 -1: -2 ^1/2 +1: -1:-
An optical system element wherein a second voltage having a voltage ratio of 1: -2 ^1/2 +1: +2 ^1/2 -1: +1 is applied. 4. One of claims 1 to 3
13. An electrostatic astigmatism corrector comprising the optical system element described in the item [6]. 5. One of claims 1 to 3
A charged particle beam exposure apparatus, comprising the optical system element described in the above section. 6. A method for manufacturing a semiconductor device, comprising a step of transferring a pattern formed on a mask or a reticle onto a wafer using the charged particle beam exposure apparatus according to claim 5.