JPH0897120A - Charged particle beam exposure method - Google Patents

Abstract

(57) [Abstract] [Purpose] It is possible to perform highly accurate correction of the stage position in and out of the measurement area of the positional displacement amount of the stage coordinates, and to perform exposure with high pattern placement accuracy. [Structure] The position displacement amount of the stage coordinates is measured at an arbitrary measurement point, and the position displacement amount of the stage coordinates is approximated by a single polynomial in the measurement region of the position coordinate amount displacement of the stage. By approximating the approximate residual with a spline function, the positional deviation amount of the coordinates of the stage between the measurement points is interpolated, and an arbitrary position is determined, and outside the measurement area of the positional deviation amount of the coordinates of the stage, An arbitrary position is determined by extrapolating the positional shift amount of the stage coordinates by the solution of the single polynomial.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a charged particle beam exposure method applied to a charged particle beam exposure apparatus such as an electron beam exposure apparatus used in a manufacturing process of integrated circuits.

In recent years, integrated circuits have been increasingly integrated and have improved functions, and are expected to play a role as a core technology for technological progress in a wide range of industries including computers, communications, and machines. .

Of the lithography techniques necessary for manufacturing such integrated circuits, photolithography is said to have a microfabrication limit of about 0.3 μm, but according to electron beam exposure, it is 0.1 μm. The following fine processing is possible.

If an electron beam exposure apparatus having an alignment accuracy of 0.05 μm or less and capable of exposing a 1 cm ² area in 1 second is realized, fineness, alignment accuracy, quick return around and reliability can be achieved. Whichever of the above, 1 to 4 is not allowed to follow other lithography techniques.
Gigabit semiconductor memory and 1 megagate logic LS
I can be manufactured.

[0005]

2. Description of the Related Art FIG. 14 is a diagram schematically showing a main part of an example of an electron beam exposure apparatus used in a manufacturing process of an integrated circuit. In FIG. 14, 1 is an electron gun which is an electron beam generating source. 2 is an electron beam generated from the electron gun 1, 3 is an optical axis, and 4 to 10 are electromagnetic lenses.

Reference numeral 11 denotes a rectangular aperture plate having a rectangular aperture 11A for shaping the electron beam 2 into a rectangular shape, and 12 is used for variable rectangular shaping of the electron beam 2 that has passed through the rectangular aperture 11A. It is a slit deflector.

Further, 13 is a transmission pattern of various shapes,
It is a so-called block mask, which is a mask plate on which a so-called block pattern is formed.

Further, 15 to 18 are mask deflectors,
In particular, the mask deflectors 15 and 16 direct the electron beam 2 onto the optical axis 3
And a deflector for deflecting the block mask 13 to a desired block pattern portion.

Further, the mask deflectors 17 and 18 are deflectors for returning the electron beam 2 deflected outside the optical axis 3 by the mask deflectors 15 and 16 onto the optical axis 3.

Further, 19 is a blanking deflector for controlling passage of the electron beam 2, and 20 is a circular aperture plate formed with a circular aperture 20A.

The circular aperture plate 20 functions as a diaphragm for utilizing only the portions of the electron beam 2 having the same current density, and blocks the passage of the electron beam 2 deflected by the blanking deflector 19. It also functions as a board.

Reference numeral 21 is a main deflector composed of an electromagnetic deflector which deflects the electron beam 2 in a relatively large range, and 22 is a sub-deflector composed of an electrostatic deflector which deflects the electron beam 2 in a relatively small range. is there.

Further, 23 is a wafer which is a sample, 24 is a wafer stage (hereinafter simply referred to as a stage) holding the wafer 23, 25 is a stage controller for controlling the position of the stage 24, and 26 is the position of the stage 24. A laser interferometer that measures coordinates, and 27 is a stage moving mechanism that moves the stage 24.

In this electron beam exposure apparatus, for example, when block exposure is performed, the electron beam 2 generated from the electron gun 1 is shaped into a rectangular cross section by the rectangular aperture 11A, and then the mask deflector. Fifteen, one
It is deflected by 6 to a desired block pattern portion of the block mask 13.

Then, the electron beam 2 whose cross-sectional shape is shaped by a desired block pattern is used as the mask deflector 1.
It is swung back on the optical axis 3 by 7, 18 and reduced / rotated by the electromagnetic lens 8, and the main deflector 21 and the sub deflector 22
Is exposed at a desired position on the wafer 23.

In this case, as a premise, the position coordinate of the stage 24 is measured using the laser interferometer 26, and the stage 24 is positioned by the stage moving mechanism 27.

Here, the laser interferometer 26, as shown in FIG. 15, shows the X laser interferometer 29 for measuring the position coordinate of the stage 24 in the X axis direction and the position coordinate of the stage 24 in the Y axis direction. And a Y laser interferometer 30 for measuring.

Reference numerals 31 and 32 are mirrors fixed to the stage 24, and reference numerals 33 and 34 are mirror surfaces of the mirrors 31 and 32.

Here, as shown in FIG. 16, the mirror 3
If the mirror surfaces 33 and 34 of Nos. 1 and 32 are distorted or the stage 24 itself is twisted, the position coordinates of the stage 24 measured by the laser interferometer 26 and the actual position coordinates of the stage 24 are measured. And the pattern formed on the wafer 23 is displaced from the design position.

Further, in the electron beam exposure apparatus, the exposure pattern may not only have to match the ideal design pattern but also the pattern formed by a stepper (hereinafter referred to as a base pattern). However, in this case, the pattern placement must be performed in consideration of the placement error of the base pattern, the distortion of the wafer 23, and the like.

FIG. 17 is a diagram for explaining the alignment of the design pattern (or the ground pattern) and the exposure pattern, and FIG. 17A is the state before the alignment, FIG.
7B conceptually shows a state after alignment.

In FIG. 17, 36 is a chip arranged on the wafer 23, 37 to 40 are position detection marks of the chip 36, 41 is the shape of the chip 36 in the coordinate system of the electron beam exposure apparatus, which is surrounded by a broken line. The exposed portion 42 is the pattern to be exposed.

That is, when aligning the design pattern with the exposure pattern, first, the positions of the position detection marks 37 to 40 provided at the four corners of the chip 36 are measured by scanning the electron beam 2.

Then, the amount of distortion in the chip 36 is approximated by a correction coefficient representing the displacement of the position of the chip 36, expansion / contraction, rotation, and trapezoidal distortion, and as shown in FIG. 17B, four position detection marks 37-40. The position of the rectangle made from,
The exposure pattern is arranged according to the distortion.

In this way, the design pattern and the exposure pattern are aligned with each other. Equation 1 shows an example of a correction formula used for this alignment, and ai, b
i, ci, di, ei, fi, gi and hi are correction coefficients for the i-th chip.

[0026]

[Equation 1]

Here, as a method for obtaining the correction coefficient for all the chips on the wafer 23, the position detection mark position is detected for all the chips on the wafer 23, and the correction coefficient for each is obtained. , The position of the position detection mark is detected only for a part of the chips on the wafer 23, and the correction coefficient is calculated for each of the measured chips from the result of the position detection,
For non-measurement chips, there is a method of determining a correction coefficient by numerical calculation using an approximate expression.

According to the method of (1), the pattern is arranged at the position measured directly, so that the pattern can be arranged with high accuracy when the coordinate system is distorted relatively uniformly inside the chip. Since it is necessary to detect the position of the position detection mark for all the chips, there is a disadvantage that the processing time per wafer increases.

On the other hand, according to the method (1), the processing time per wafer is shorter than that of the method (2), but in the non-measurement chip, the correction coefficient is estimated by the approximate expression, and therefore the pattern May be misaligned.

The cause of this displacement is the stage 2
Although the positional deviation due to the distortion of the coordinates of No. 4 and the positional deviation of the position detection mark due to the shape of the chip 36 and the deformation of the wafer 23 are conceivable, there is a complicated positional dependence that cannot be measured by the method. Misalignment is due to stage 2
It is considered to be a positional shift due to the distortion of the coordinates of No. 4.

That is, FIG. 18 shows the distortion of the arrangement coordinates of the chip 36, and FIG. 19 shows the distortion of the coordinates of the stage 24. The distortion of the arrangement coordinates of the chip 36 shown in FIG.
The distortion of the coordinates of the stage 24 shown in FIG.
It is considered that this is due to the addition of the shape of the above and the deformation of the wafer 23.

As described above, in the alignment method, the positional deviation amount of the pattern is (a) the distortion component of the coordinates of the stage 24 of the electron beam exposure apparatus, (b) the distortion component of the chip shape, and (c) the wafer itself. Are classified into distortion components due to deformation.

Therefore, in the electron beam exposure apparatus, the positional deviation of the pattern due to the distortion of the coordinates of the stage 24 is removed by highly accurate correction, and then the positional deviation of the pattern due to the distortion of the chip and the pattern deviation due to the deformation of the wafer itself are removed. Regarding the positional deviation, the position detection marks of a part of the chips are detected for each wafer, the approximate function is determined, and the beam deflection position or the beam deflection position for all the stage positions is calculated by the approximate value calculated from the approximate function. The stage movement position is corrected.

In this case, since it is considered that the same chip position detection marks are repeatedly arranged,
Even if it is the same regardless of the position on the wafer or has a dependency on the position on the wafer, it can be approximated by a simple function.

Here, as a method of measuring the error in the position coordinate of the stage 24, (a) between the position of the reference pattern (resist pattern) actually formed on the surface of the wafer 23 by exposure and its design position. The method of directly measuring the amount of deviation of the stage 24 by a length measuring device, and (b) the position detection mark previously arranged on the reference wafer is moved to a certain stage position coordinate to perform position detection, and the amount of position deviation of the stage 24 is determined. There is a method of indirectly measuring the amount of displacement of the reference wafer from the position detection mark.

The method of directly measuring the reference pattern requires a number of steps such as resist coating, exposure of the reference pattern, development, measurement by a length measuring device, and a problem that the time required for measurement increases. have.

On the other hand, the measurement by the position detection mark is easy and the time required for the measurement is short, but since it is an indirect measurement, the position displacement amount due to the distortion of the stage 24 is high only by itself. It has a problem that it is difficult to measure with accuracy.

The correction of the positional deviation of the pattern caused by the distortion of the coordinates of the stage 24 is performed by the stage 24.
The position shift amount of the coordinates can be approximated by a function of the position of the stage 24, and each time the stage 24 is moved, the position shift amount can be moved by an extra amount calculated by the approximation function.

Conventionally, as this approximation function, a single polynomial described by stage position coordinates, for example, the formula shown in Formula 2 has been used. However, x and y are position coordinates of the stage 24.

[0040]

[Equation 2]

[0041]

FIG. 20 shows an example of the positional deviation amount δx of the pattern with respect to the position coordinate x of the stage 24 in the X-axis direction. The black dots are plots of the amount of positional deviation, and are the same in FIGS. 21 to 23.

As is apparent from FIG. 20, the positional deviation amount of the pattern has a complicated stage position dependency and is approximated by a low-order simple single polynomial, for example, FIG.
It is inevitable that the approximation error becomes large in the approximation by the quadratic function as shown in FIG.

On the other hand, as shown in FIG. 22, when a high-order function is used as the approximation function, it is possible to improve the approximation accuracy, but nevertheless, it is about 0.
There is an approximation error of about 05 μm, and the approximation value suddenly increases outside the measurement region. By performing correction using this approximation function, the amount of misalignment is rather increased. .

Further, as shown in FIG. 23, when a spline function which is a piecewise polynomial is used, highly accurate approximation can be performed within the measurement region, but outside the measurement region, the coordinate values referred to are referred to. Since the correction value changes greatly even if is changed a little, it is impossible to extrapolate (extrapolate) an appropriate approximate value.

Here, in the electron beam exposure apparatus,
As shown in FIG. 24, there is a chip 44 having a position detection mark for electron beam adjustment formed immediately outside the wafer 23, and when adjusting the electron beam 2, the absolute amount of positional deviation of the pattern is detected. Since the change amount, not the value, has a large effect, the change in the actual shift amount at the position of the chip 44 cannot be considered to be such a steep value, and thus the change in the correction value is not abrupt. It is not preferable, and it may be considered that the outside of the measurement area is exposed. Note that FIG.
4, the area 46 surrounded by the solid line 45 indicates the measurement area.

In view of the above point, the present invention makes it possible to perform highly accurate correction of the stage position over the inside and outside of the measurement area of the positional displacement amount of the stage coordinates, and to perform exposure with high pattern placement accuracy. An object is to provide a charged particle beam exposure method.

[0047]

According to a charged particle beam exposure method of the present invention, a charged particle beam generated from a charged particle beam generation source is desired for a sample held on a stage which is horizontally movable. It is applied to a charged particle beam exposure apparatus that performs exposure by irradiating a position.

In the present invention, first, as shown in FIG. 1, the positional displacement amount of the coordinates of the stage is measured at an arbitrary measurement point. Note that the black dots in FIG. 1 are plots of the positional deviation amount, and the same applies to FIG.

Then, in the measurement area of the positional deviation amount of the stage coordinates, the positional deviation amount of the stage coordinates is approximated by a single polynomial, and the approximate residual is approximated by a spline function which is a piecewise polynomial. By doing so, the positional deviation amount of the coordinates of the stage between the measurement points is interpolated, and the stage position is corrected.

Outside the measurement area of the positional displacement amount of the stage coordinate, the positional displacement amount of the stage coordinate is extrapolated by the solution of the single polynomial obtained by approximating the positional displacement amount of the stage coordinate described above. By doing so, the stage position is corrected.

[0051]

In the present invention, within the measuring area of the positional displacement amount of the coordinates of the stage, the positional displacement amount of the stage is approximated by a single polynomial, and the approximate residual is approximated by a spline function which is a piecewise polynomial. By this, the positional deviation amount of the stage coordinates between the measurement points is interpolated and the stage position is corrected, so that the stage position can be corrected with a small approximation error in the measurement region.

Further, outside the measurement area of the positional displacement amount of the stage coordinates, the stage position is corrected by extrapolation by the solution of the single polynomial which approximates the positional displacement amount of the stage. Even outside the measurement area
It is possible to perform the approximation that reflects the tendency of the positional deviation amount in the measurement region, and it is possible to correct the stage position with less approximation error.

By the way, when the least squares approximation is performed for all the measurement points in the measurement area to obtain a single polynomial, as shown in FIG. 1, a boundary portion between the measurement area and the outside of the measurement area is used. Then, a relatively steep change may remain in the approximation function.

Therefore, as shown in FIG. 2, the least-squares approximation is performed by using only the outermost measurement data A and B of the measurement area (in plan view, data of at least three outermost measurement points). There is also a method in which a single polynomial is obtained with and the approximate residual to be approximated by the spline function is reduced at the end of the measurement region.

In this case, the plane created by the approximation function according to the present invention is as shown in FIG.
It is possible to set the positional displacement amount of the coordinates of the stage to about 0.07 μm which is an error degree of the pattern arrangement accuracy, not only in the measurement area but also in the boundary portion of the measurement area and the outside thereof.

Incidentally, FIG. 4 shows a plane formed by the conventional method (approximation by a single polynomial equation), and the positional displacement amount of the stage coordinates is 0.2 μm at the maximum (4 mm outside the measurement area). ing.

If the measurement result is flat on the boundary, the approximation function is flat as shown in FIG. 5, but as shown in FIG. If it does not ride over the flat surface (curved surface) created by, it may not be possible to achieve the desired effect.

Therefore, as shown in FIGS. 7 and 8, a pseudo measurement point 4 at which the approximation error in a single polynomial is 0
Least squares approximation is performed on the spline function using the measurement data group to which 7 is added.

In such a case, as shown in FIG. 9 and FIG.
As shown in, it is possible to eliminate a sudden change on the boundary. The pseudo measurement point 47 is preferably separated from the outermost measurement point by at least the distance between the measurement points.

In this way, the error in the position coordinate of the stage can be corrected by changing the length measurement value of the laser interferometer of the stage by the amount of the approximated positional deviation.

[0061]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below by exemplifying a case where the present invention is applied to an electron beam exposure method used in an electron beam exposure apparatus used in an integrated circuit manufacturing process.

In this embodiment, first, (1) a position measuring pattern is formed on the wafer coordinate with respect to the reference wafer.
It is formed by exposing at an arbitrary interval, for example, 1 to several mm.

Then, the position of the position measuring pattern is measured by a length measuring device, and the ideal position of the pattern on the wafer, that is, the design position (Xi ', Y) of the pattern.
The position shift amount (ΔXi, ΔYi) from i ′) is calculated. However, (Xi ', Yi') is the design position of the i-th pattern.

In this case, the stage position (X
i, Yi) (length measurement value by a laser interferometer) is calculated, and the stage position (Xi, Yi) and the position shift amount (ΔXi, ΔY)
i) Correlation with (correction amount).

(2) Next, the outermost measurement points of the stage displacement amount measurement area are extracted, and the X and Y components Xi and Yi of the position displacement amounts of the outermost measurement points are calculated as position coordinates. Linear function fx (X, Y) = AX + BY + E, fy (X,
Y) = CX + DY + F. The coefficient A,
B, C, D, E, and F are approximated by the least square method.

(3) Next, all measurement points (Xi, Yi)
Approximate residuals δXi = ΔXi−fx (X, Y), δYi = Δ for approximate expressions fx (X, Y) and fy (X, Y)
Calculate Yi-fy (X, Y).

Next, a pseudo measurement point with an approximate residual of δXi = δYi = 0 is added to the measurement data at a position outside the measurement area (see FIGS. 7 and 8).

(4) The approximate residuals (δXi, δYi) at all measurement points including the pseudo measurement points are calculated by using the cubic spline function gx (X, Y) = ΣamH (X, Y), gy. (X,
Y) = ΣbnI (X, Y) is approximated.

It should be noted that H (X, Y) and I (X, Y) are b-spline functions serving as basis functions, and the coefficients am and bn.
Is determined by the method of least squares.

(5) The entire area where the stage can be moved is divided at regular intervals, for example, 1 mm, and the grid points (Xs, Y
The positional deviation amounts ΔXst and ΔYst at t) are approximated by fx
(Xs, Yt) + gx (Xs, Yt), fy (Xs, Y
t) + gy (Xs, Yt) is interpolated and calculated.

However, Xs = Px · s and Yt = Py · t, Px and Py are intervals at which interpolation is performed, and s and t are integers.

(6) The position measurement pattern is exposed on a plurality of types of reference wafers at several types of intervals, and the above-mentioned (1) is performed.
The values (ΔXst, ΔYst) of the positional deviation amounts on the plurality of reference wafers are calculated by repeating the steps to (6).

Then, the correction amounts at the same position obtained by a plurality of measurements are arithmetically averaged to eliminate the influence of the measurement error and the difference of the wafer distortion between the wafers, and the position caused by the distortion of the stage coordinate system. Get the amount of deviation.

At the time of exposure, linear interpolation is performed from the correction amounts at the four corners of the area including the current stage position (length measurement value by the laser interferometer) to obtain the correction amount at the current position. To do.

Here, the position interval at which the interpolation calculation is performed is sufficiently smaller than the change in the position shift amount, and the interval (the length measurement value + correction amount) is the interval at which highly accurate interpolation is possible even in the primary interpolation in that section. By treating it as the position coordinate of the stage of
It is possible to correct the amount of positional deviation due to the distortion of the coordinates of the stage.

As for the pattern alignment for each wafer or lot, after the above correction is performed, the mark positions of several to ten or more chips on the wafer are detected, and the positions within the chips shown in Formula 3 are detected. The shift amount (ΔX, ΔY) is approximated to the function of the position coordinate (X, Y) in the chip by least squares.

[0077]

[Equation 3]

Then, the obtained coefficients ai, bi, ci, d
Further, i, ei, fi, gi, hi are further subjected to least-squares approximation by the function of the position coordinates of the chip center position on the wafer shown in Formula 4.

[0079]

[Equation 4]

Note that (Xi, Yi) is the position (X, Y)
Center position of the i-th chip on the wafer, including fxj
(Xi, Yi) is Σα _kl X ^k Y ^l , for example, a quadratic function of X and Y.

Here, at the time of exposure, the correction coefficients ai, bi, ci, di, ei, fi, gi, hi of all the chips are calculated according to the equation 4, and the positional deviation amount is calculated according to the equation 3, and the position thereof is calculated. Correction is performed by changing the electron beam deflection position by the amount of deviation, and alignment is performed.

As described above, according to the present embodiment, in the measuring area of the positional displacement amount of the coordinates of the stage, the positional displacement amount of the stage is approximated by a linear function, and the approximate residual is a cubic spline function. By interpolating the positional deviation amount of the stage coordinates between the measurement points and correcting the stage position, the positional deviation amount of the stage is approximated outside the measurement area of the positional deviation amount of the stage coordinates. Since it is supposed that the stage position is corrected by extrapolation by the solution of the linear function described above, the stage position is corrected with high accuracy over the inside and outside of the measurement area of the position shift amount of the stage coordinates, and the pattern is corrected. It is possible to perform exposure with high placement accuracy.

In this embodiment, the correction amount is recalibrated when the distortion amount of the stage coordinate system changes with time.

FIG. 11 is a diagram for explaining a conventional method for correcting the change with time of the positional deviation amount at a certain point on the stage. In the conventional method, the positional deviation at a certain point on the stage is shown. The amount of positional deviation has been corrected by measuring the amount by a direct measurement method at a predetermined measurement cycle, that is, by directly exposing and forming a reference pattern.

In the present embodiment, the correction method shown in FIG. 12 or FIG. 13 is added during the correction cycle through the direct measurement method to change the position shift amount at a certain point on the stage with time. Is corrected.

Here, the method shown in FIG. 12 uses the indirect measurement method during the correction cycle of the direct measurement method, that is,
The position shift amount of the stage is measured by using the position detection mark formed on the reference wafer in advance, and the above (2) to
The step (6) is performed to correct the stage position.

In this case, by shortening the correction cycle through the indirect measurement method, the temporal change amount of the positional deviation amount per unit time (or one wafer) can be reduced.

The method shown in FIG. 13 predicts the amount of positional deviation during the correction cycle through the indirect measurement method shown in FIG. 12, and changes the amount of positional deviation at a certain point on the stage with time. The change is corrected.

In this method, the first measurement is performed by the direct measurement method. In this case, the correction amount at a certain point on the measured stage is E ₀ .

Further, in this case, in addition to the measurement by the direct measurement method, the measurement by the indirect measurement method is performed on the reference wafer having the position detection mark, and the steps (2) to (6) described above are performed. Then, an approximate value of the error amount is obtained. In this case, the calculated correction amount (error amount) is M ₀ .

If the first measurement is performed, the correction amount E ₀ is calculated from the predetermined stage position each time the stage is moved.
Only, the coordinates are changed and the stage position is corrected.

In the second measurement, only the indirect measurement method is performed with the correction amount E ₀ corrected, and the correction amount M ₁
Ask for.

Then, the time change amount M ₁ from the time when the correction amounts E ₀ and M ₀ are measured and the change amount M ₁ / T per unit time are measured.
₁ (T ₁ is the measurement interval between the first measurement and the second measurement or the number of processed wafers) is obtained.

After that, the correction amount is E ₀ + M ₁ + tM ₁ / T
_The correction is performed as ₁ (M ₁ is the correction amount obtained by the second indirect measurement method and the interpolation, and t is the time since the last indirect measurement method was performed).

Here, the first term (E ₀ ) of the correction amount is a component of the positional deviation amount that does not change with time, and the second term (M ₁ )
Is the amount of positional deviation that has changed with time during the first and second measurements, and the third term (tM ₁ / T ₁ ) is the amount of positional deviation that has changed with time from the second measurement.

In the third and subsequent measurements, only the indirect measurement method was also used, and the correction amount was E ₀ + M _N + tM _N / T
_N (M _N is the correction amount obtained by the N + 1-th indirect measurement method and the interpolation, and T _N is the measurement interval between the N-th measurement and the N + 1-th measurement or the number of processed wafers) Make a correction.

Note that an error may occur only by the indirect measurement method and it is necessary to reconfirm the effectiveness of the correction. Therefore, the measurement by the indirect measurement method and the update of the correction amount are constant. After repeating the number of times, the measurement by the direct measurement method, the approximation and the interpolation are performed again, and the correction amount is updated to set the error amount to zero.

In this way, in the case of correcting the temporal change of the positional deviation amount at a certain point on the stage, FIG.
The stage position can be corrected with higher accuracy than the method shown in FIG.

[0099]

As described above, according to the present invention, in the position displacement amount measuring area of the stage coordinates, the stage position displacement amount is approximated by a single polynomial, and the approximate residual is divided into a piecewise polynomial. By approximating with the spline function, the positional displacement amount of the stage coordinates between the measurement points is interpolated to correct the stage position, and the stage position is outside the measurement region of the positional coordinate amount of the stage. The position of the stage is corrected by extrapolation by the solution of a single polynomial that approximates the amount of displacement. It is possible to perform correction and perform exposure with high pattern placement accuracy.

[Brief description of drawings]

FIG. 1 is a diagram for explaining the present invention (a diagram showing a relationship between a stage position and a positional deviation amount in the present invention).

FIG. 2 is a diagram for explaining the present invention (a diagram showing a relationship between a stage position and a positional shift amount in the present invention).

FIG. 3 is a diagram for explaining an effect in the case of the present invention.

FIG. 4 is a diagram for explaining an effect when a conventional method (approximation by a single polynomial) is used.

FIG. 5 is a diagram showing a curved surface created by an approximation function in the present invention.

FIG. 6 is a diagram showing a curved surface formed by an approximation function in the present invention (when the boundary portion does not ride over the plane formed by the approximation function).

FIG. 7 is a diagram showing pseudo measurement points added in the present invention.

FIG. 8 is a diagram showing pseudo measurement points added in the present invention.

FIG. 9 is a diagram showing an approximation function when a pseudo measurement point is added in the present invention.

FIG. 10 is a diagram showing a curved surface formed by an approximation function when a pseudo measurement point is added in the present invention.

FIG. 11 is a diagram for explaining a conventional method for correcting a temporal change in a positional deviation amount at a certain point on a stage.

FIG. 12 is a diagram for explaining an example of a method of correcting a temporal change of the positional deviation amount at a certain point on the stage added in the embodiment of the present invention.

FIG. 13 is a diagram for explaining another example of the method of correcting the temporal change of the positional deviation amount at a certain point on the stage added in the embodiment of the present invention.

FIG. 14 is a diagram schematically showing a main part of an example of an electron beam exposure apparatus used in a manufacturing process of an integrated circuit.

15 is a diagram showing a part of the electron beam exposure apparatus shown in FIG.

FIG. 16 is a diagram showing distortion on a mirror surface.

FIG. 17 is a diagram for explaining alignment of a design pattern (or a base pattern) and an exposure pattern.

FIG. 18 is a diagram showing distortion of arrangement coordinates of chips.

FIG. 19 is a diagram showing distortion of stage coordinates.

FIG. 20 is a diagram showing the stage position dependency of the positional displacement amount of the stage coordinates.

FIG. 21 is a diagram showing a case where a positional deviation amount of stage coordinates is approximated by a quadratic function of position.

FIG. 22 is a diagram showing a case where a positional deviation amount of stage coordinates is approximated by a higher-order function of position.

FIG. 23 is a diagram showing a case where a positional displacement amount of stage coordinates is approximated by a position spline function.

FIG. 24 is a diagram for explaining a problem in the case of approximating a positional displacement amount of stage coordinates by a position spline function.

[Explanation of symbols]

Continuation of front page (51) Int.Cl. ⁶ Identification number Office reference number FI Technical display location G03F 7/20 521 G21K 5/04 MH01J 37/305 B 9508-2G (72) Inventor Yoshihisa Ohba Kawasaki, Kanagawa 1015 Kamiodanaka, Nakahara-ku, Yokohama, Fujitsu Limited (72) Inventor Junichi Kai, 1015, Kamikodanaka, Nakahara-ku, Kawasaki, Kanagawa Prefecture, Fujitsu Limited

Claims (9)

[Claims] 1. A charged particle beam exposure apparatus for performing exposure by irradiating a charged particle beam generated from a charged particle beam generation source to a desired position of a sample held on a stage movable in the horizontal direction. , Measuring the position displacement amount of the stage coordinates at any measurement point, in the measurement region of the position displacement amount of the stage coordinates,
The displacement amount of the coordinate of the stage is approximated by a single polynomial, and the approximate residual is approximated by a spline function to interpolate the displacement amount of the coordinate of the stage between the measurement points, and the stage Correct the position,
Outside the measurement area of the positional deviation amount of the stage coordinates, the positional deviation amount of the stage coordinate is extrapolated by the solution of the single polynomial to correct the stage position. Particle beam exposure method. 2. The charged particle beam exposure according to claim 1, wherein the single polynomial takes a least-squares approximation value using at least three outermost measurement points in the measurement region. Method. 3. A pseudo measurement point which is a solution of the single polynomial is provided outside the outermost measurement point in the measurement area,
The charged particle beam exposure method according to claim 1 or 2, wherein the spline function is obtained by including the pseudo measurement points. 4. The charged particle beam exposure according to claim 3, wherein the pseudo measurement point is separated from the outermost measurement point in the measurement area by a distance equal to or more than the distance between the measurement points. Method. 5. The measurement of the positional displacement amount of the coordinates of the stage at an arbitrary measurement point is performed by measuring the displacement amount between the position of a reference pattern actually formed on a sample surface by exposure and its design position. 5. The charged particle beam exposure method according to claim 1, wherein the measurement is performed by directly measuring with a long device. 6. The measurement of the positional displacement amount of the stage coordinates at an arbitrary measurement point is performed at a desired time interval or a desired number of processed samples.
The charged particle beam exposure method according to 2, 3 or 4. 7. The measurement of the position displacement amount of the coordinate of the stage at an arbitrary measurement point is performed by detecting the position of a position detection mark arranged on a reference sample. 3. The charged particle beam exposure method according to 3 or 4. 8. A length measuring device for measuring the amount of displacement between the design pattern and the position of a reference pattern actually formed on the sample surface by exposure, for the measurement of the amount of displacement of the coordinate of the stage at an arbitrary measurement point. And the measurement by the length measuring device, and the measurement by the position detection of the position detection mark arranged on the reference sample, which is performed after a desired time interval or a desired number of processed samples. Claims 1, 2, 3
Or the charged particle beam exposure method described in 4 above. 9. The detection time is calculated from the difference between the detection position at the N-th position detection time point of the position detection mark (where N is a positive integer) and the detection position at the (N + 1) th position detection time point of the position detection mark. The amount of positional deviation per interval or the number of processed samples is calculated, and the positional deviation amount between the Nth position detection time point of the position detection mark and the (N + 1) th position detection time point of the position detection mark is calculated as the detection time interval. Alternatively, the position shift amount per unit time or the number of unit samples is predicted by linearly interpolating the position shift amount per number of processed samples, and the position detection mark N + 2nd time from the N + 1th position detection time of the position detection mark. During the position detection time of, based on the amount of positional deviation added the predicted amount of positional deviation per unit time or unit number of samples,
9. The charged particle beam exposure method according to claim 8, wherein the stage position is corrected.