JPS596530A - Correction of deflection strain at charged beam exposure device - Google Patents

Abstract

PURPOSE:To enhance the precision of positioning of a charged beam onto the surface of a sample by a method wherein deflection strain at the level surface of the arbitrary height is measured precisely to obtain the correction formulas of deflection strain. CONSTITUTION:The charged beam is focused in height (z1) of the level surface to measure deflection strain. The correction equations of deflection strain at the level surfaces of the same heights with a low standard mark and a high standard mark are decided using the standard marks of the two faces of high and low holding the focus in this condition as it is. Then, the values of Ai(z1), Bi(z1), (i=0-9) are calculated performing substitution as z=z1 at the formula I . As a result, the equations II can be obtained as the correction equations of deflection strain at height (z1). Because the position of the focus of the beam is not altered when the formulas II are obtained, the beam injected on the surface of the sample comes moving on a straight line. Accordingly, the quantity of deflection strain measured according to the standard marks of the two faces of high and low can be expressed by the linear function of height (z), and the formula II have the coefficients of the more precise values as the correction formulas of the deflection strain at height z=z1 than the usual method.

Description

Detailed Description of the Invention [Technical Field of the Invention] The present invention provides a charged beam exposure apparatus that deflects a charged beam to expose a /4' tan, and focuses a charged beam optical system on a level plane at an arbitrary height. When combined, the present invention relates to a method for accurately measuring deflection distortion on the level plane, and a method for accurately positioning a charged beam on a sample surface based on the measurement results of the deflection distortion.

(Technical background of the invention and its problems) In order to accurately position the charged beam on the surface of a sample such as a silicon wafer using a charged beam exposure device and expose the holes, the optical system of the charged beam exposure device is required. It is necessary to measure the eigenplane deflection distortion and correct the deflection distortion according to the height of the sample surface.

A conventional method of this type will be explained.

A standard mark consisting of two surfaces with heights different by 100 to 200 Itrn is used as a standard mark fixed to a sample moving stage for measuring deflection strain. First, the charged beam is focused on a level plane at the same height as the lower standard mark, and the standard mark is placed within the deflection field by nXn (n=
Move the stage to the lattice points of 3.5, 7°9...),
Deflection distortion is determined from the position of the stage measured by the Rade length measuring device at each grid point and the mark detection position of the standard mark by the beam deflection system. From this deflection distortion measurement result,
Determine the deflection distortion correction formula for a level plane at the same height as the low standard mark as shown below.

Here, (x, y) are designed values of the charged beam irradiation position within the deflection field. (XtD t YtD)' is the position coordinate after deflection distortion correction. When exposing the armpit on the sample surface at the same height as the low standard i-
The beam is deflected based on the value of Yzn ).

From the deflection distortion measurement using the higher standard mark in the same manner as above, the following deflection distortion correction coefficients (x z y ) and (xhn - Yho ) on the level plane at the same height as the higher standard mark are calculated. has the same meaning as equation (1).

Next, when exposing a pattern on a sample surface such as a silicon wafer, first move the center of the deflection field to be drawn directly below the beam optical axis, and use a height measuring device to measure the height of the sample surface corresponding to the deflection field. Measure carefully. Here, the height of the lower standard mark is measured using the height measuring device, and the result is t.
Similarly, let h be the height measurement result of the higher standard mark, and 2 be the height measurement result of the sample surface. Using equations (1) and (2), the deflection distortion correction equation at height 2 is determined as follows.

Equation (3) is a deflection distortion correction equation at a height of 2, and the equation (
Each coefficient on the right side of 3) is obtained by linearly interpolating the corresponding coefficients of equations (1) and (2). When exposing a pattern to a deflection field with a height of 2, first focus the charged beam at a height of 2, and then calculate the position coordinates (x t y ) in the design data of each pattern as ( X, Y) and convert it to (x, , yD),
The charged beam is deflected based on (xD, yD) and a /4 tan is exposed at a predetermined position on the sample surface.

In the conventional method, the deflection distortion was corrected based on the deflection distortion measurement results at the height of the standard marks on the two sides.
When a pattern is drawn on a sample surface at a height apart from each height of the standard mark on the surface, the positioning accuracy of the /4 button decreases. Furthermore, when attempting to use a technique such as dynamic focus correction, linear interpolation of coefficients does not hold, so there is a drawback that the positioning accuracy of the tactile lens further deteriorates.

[Purpose of the invention]

In order to eliminate these drawbacks, the present invention makes it possible to measure the deflection distortion that occurs when deflecting a charged beam when focused at an arbitrary height 2, and also to measure the deflection distortion at an arbitrary height 2. This method corrects deflection distortion when drawing a pattern on a sample surface based on measurement results, and will be described in detail below.

[Embodiment of the Invention] First, in the deflection distortion correction method explained as the conventional technique, each coefficient was linearly interpolated to determine the deflection distortion correction formula at a height of 2. Let's explain the case of standing up.

Here, for the sake of simplicity, only the correction of the vowed coordinates will be explained.

Equations (1) and (2) are transformed to obtain the following equation.

XtD-) (-4to+At, X+At2Y+-+ A
L, Yj (4)XhD-X=Aho+Ah, X+Ah2
Y+-+Ah9Y5 (5) The constant result of deflection distortion using the standard mark is input to the beam deflector and fc
It is calculated based on the data (Xl p Ys ) and the value measured by the Radhe length measuring device (Xl.

The moving distance of the stage corresponding to Y) (XLi, YL
It is given as a set of j). If deflection distortion i = does not exist, (
, Xl, Yj) = (XLi, YLj). Each coefficient of equation (1) and equation (2) is calculated by (X,
(XLl, yLj) for Y), and (Xt
D, Yzn) or (Xhn-YhD) with (xl
, Yj).

In equations (4) and (5), (X, y)=(Xt
, ty YLJ). (XLl,
Yt J) is the reference coordinate value obtained by the laser performance 1j length machine l), giving the amount of deflection strain in the X direction at the reference coordinate value (Xt4 p ”LJ) of the formula When the amount of deflection distortion changes in the height direction, that is, in two directions, as a linear function of 2, the reference coordinate value (Xt, 1 - Yt, j
) is given by (6).

Equation (6) matches the equation obtained by linearly interpolating the coefficients corresponding to each of Equation (4) and Equation (5). That is, the deflection distortion of the present invention utilizes such a principle. First, a charged beam is focused on the height zl of the level plane whose deflection distortion is to be measured. Next, keep the focus as it is and change the height to 2.
Formula (1) is calculated in the same way as before using the standard mark on the surface.
, determine equation (2). Next, in equation (35), K
=* Putting 1, Ai(zl)y Bl(tl)(1=
0 to 9). As a result, the following equation is obtained as a deflection distortion correction equation at the height zl.

When formula (7) is obtained, the focal position of the beam is not changed, so the beam incident on the sample surface moves on a straight line. Therefore, the amount of deflection distortion measured by the standard marks on the two high and low surfaces is completely a linear function of the height 2. That is, equation (7) is a deflection distortion correction equation at the height of Z:Z1, and the conventional method also has coefficients with exact values. Note that, since the depth of focus of the charged beam is deep, standard marks having a height difference of 100 to 200 μm can be detected without any problem.

Next, an application example of the deflection distortion correction method according to the present invention will be explained.

The deformation in the height direction of the silicon wafer used to draw the surface and the surface of the surface is measured in advance. Note that this technique is a well-known technique for electron beam exposure apparatuses. As a result, it is assumed that the sea urchin that draws the glossy tongue is between the heights 22 and zh. By the method according to the invention! ``Determine the deflection distortion correction formula for 1t and t, =Zh, and when drawing each field, calculate the
==22 and each correction coefficient of z = z > is linearly interpolated. Furthermore, if the difference between 5 = zl and w = z hO is large, or if more precise deflection distortion correction is required, z
=xl and z=z1 between z=zh. zl, z5・
..., etc., and select two measured level planes for linear interpolation according to the height of the sample surface in the actual drawing. It is possible to reduce errors when setting correction coefficients.

〔Effect of the invention〕

As explained above, we have made it possible to strictly measure the deflection distortion on a level plane of any height and obtain the deflection distortion correction formula. positioning accuracy on the sample surface can be improved. Another advantage is that deflection distortion can be corrected accurately even when focusing the charged beam according to the height of the sample or performing dynamic focus correction within the π plane according to the height of the sample. There is. The present invention is particularly effective in cases where the demand for pattern positioning accuracy becomes stricter as the mother tongue becomes finer, such as in LSI manufacturing using submicron A tongues.

Claims (1)

[Claims] A method for correcting deflection distortion that occurs when deflecting a charged beam and positioning the charged beam on a sample surface in a charged beam exposure device that uses an electron beam, ion beam, etc., and used as a standard mark for measuring deflection distortion. Using a standard mark with marks set on two level planes with different heights, with the focus of the charged beam set at an arbitrary height, deflection distortion is measured with the standard marks on two planes of high and low heights by the charged beam, A method for correcting deflection distortion in a charged beam exposure apparatus, characterized in that a deflection distortion correction formula for a focused level plane is determined from measurement results for the two high and low planes.