JPH06188180A - Charged particle beam exposure - Google Patents

Abstract

PURPOSE:To calculate a true correction coefficient not depending on the scanning time of every mark by removing thermal influences of an apparatus. CONSTITUTION:N pieces out of a plurality of marks formed on an exposure plane are scanned with a charged particle beam one after another to obtain a discrepancy di (1<=i<=n) between a beam position and the center position of every mark, and the firstly scanned mark is finally scanned again to obtain a discrepancy d'1; when the (d'1-d1) of the measurement discrepancies in twice scannings of the first mark is DELTA, and the total scanning time T, the drift DELTAi at the scanning time of every mark is determined by an expression DELTAi=DELTAX(i-1)/n, and the true discrepancy Di of every mark is calculated by an expression Di=di-DELTAi to correct a deflection position of a beam by means of D,. Instead of the expression DELTAi, an expression DELTAi=DELTAX(ti-t1)/T (t is time) can be used.

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 (electron beam etc.) exposure method used in a manufacturing process of semiconductor devices and the like, and in particular, it corrects a beam deflection position drift caused by a thermal change of an exposure apparatus. Regarding the method.

In recent years, charged particle beam exposure apparatuses have been used for pattern formation of LSIs, etc., but with higher density and finer LSIs, higher precision and higher resolution are required.
According to the present invention, in order to correct the beam deflection position of a field (a practical deflection range of a beam, a drawing range) (so-called field correction), when a beam is scanned over a mark to measure a positional deviation amount (offset). In addition, it can be used to correct the drift of the positional deviation amount caused by the thermal influence of the device due to the heat generation of the deflection coil.

[0003]

2. Description of the Related Art The deflection coil of a main deflector 20 of a charged particle beam exposure apparatus (see FIG. 3) is arranged in a narrow space below the column (in the final lens). Generates heat. Due to the thermal expansion of the coil itself due to heat generation, the coil shape is deformed and the position is displaced. In addition, the Joule heat generated in the coil causes deformation and displacement of the pole piece that forms the magnetic field of the lens due to heat conduction and heat radiation. As a result, a deviation in deflection or a deviation in the origin occurs.

In the case of actual exposure, an operation of obtaining and updating a beam correction coefficient, for example, a field correction coefficient at a certain fixed time and the exposure are repeated.
The conventional process of calculating the correction coefficient is performed as follows. By moving the stage, the marks placed at specified positions in the drawing area on the wafer are sequentially moved to the beam position, the beam is deflected and the marks are scanned, and the amount of deviation between the beam position and each mark position is acquired. Then, the correction coefficient was calculated from the obtained shift amount.

However, the deflection position of the beam is deviated due to the thermal deformation of the coil itself due to the heat generation of the deflection coil. Therefore, the amount of deviation between the beam and each mark acquired by long-time beam scanning still includes the amount of drift due to the thermal effect of the device in the true amount of deviation, and the correction coefficient was calculated based on this. .

[0006]

In the conventional example, as the position measurement of each mark progresses, the amount of displacement of each mark is subject to the thermal change of the apparatus. The amount depends on the scanning time of the area, which causes a problem that an accurate correction coefficient cannot be calculated.

The present invention removes the thermal influence of the exposure apparatus to calculate a true correction coefficient that does not depend on the scanning time of each mark,
The purpose is to achieve high precision and high resolution.

[0008]

[Means for Solving the Problems] 1)
The n marks of the plurality of marks formed on the exposed surface are sequentially scanned by the charged particle beam to obtain the deviation amount d _i (1 ≦ i ≦ n) between the beam position and the center position of each mark. , The first scanned mark is scanned again and the last one is scanned to obtain the displacement amount d ₁ ′, and the difference (d ₁ ′ −d ₁ ) between the measured displacement amounts of the first mark in two scans is set as Δ. If the total scanning time is T, the drift amount Δ _i at each mark scanning time is
Equation (1), Δ _i = Δ × (i-1) / n (1) The true deviation amount D _i of each mark is calculated by equation (2), and D _i = d _i −Δ _i (2 ) A charged particle beam exposure method for correcting the deflection position of a beam using this D _i value, or 2) the method according to claim 1.
Instead of equation (1), the following equation (3) Δ _i = Δ × (t _i −t ₁ ) / T (3) where t ₁ is the scanning time of the first mark and t _i is the i-th Mark scan time. Is achieved by the charged particle beam exposure method.

[0009]

In the present invention, after scanning a plurality of marks with a beam, the time-dependent change of the offset drift amount (positional shift amount (offset) drift) caused by the thermal influence of the apparatus with the passage of the scanning time is acquired, Based on this, the drift amount of the thermal effect is subtracted from the measured amount of misalignment of each mark to correct it, so that the scanning time dependence of the measured amount is eliminated. As a result, the positional deviation caused by the thermal effect of the device is calculated from the correction coefficient that includes the conventional thermal effect, which is simply calculated from the amount of deviation between the beam position and each mark position by deflecting the beam and scanning the mark. It is possible to calculate an accurate correction coefficient from which the drift amount has been subtracted.

[0010]

Embodiments FIGS. 1A to 1C are explanatory views of an embodiment of the present invention. FIG. 1A shows a layout of n marks formed on the wafer. The stage on which the wafer is placed is moved in the direction of the arrow, the beam is scanned when each mark comes directly under the beam, and the amount of deviation between the beam and the mark center is analyzed by analyzing the waveform of the reflected electrons. d _i (1 ≦
i <n) is acquired. This is performed for all the marks on the wafer, in which case the first scanned mark is again scanned last.

FIG. 1B shows the relationship between the amount of measurement deviation and the elapsed time. The difference between the measured deviation amounts (d ₁ '-d ₁ ) in the two scans indicates the offset drift amount Δ caused by the passage of the time T during which all marks are scanned. The offset drift amount Δ _i at each mark scanning time is obtained by linear interpolation. At this time, the interval between the scanning times of the respective marks is approximated by a time obtained by equally dividing the scanning time T of all the marks by the number n of the marks to obtain the following equation.

Δ _i = (d ₁ ′ −d ₁ ) × (i−1) / n (1) Using this equation, the true deviation amount D considering the offset drift amount Δ _i due to the thermal change of the device. _i is calculated by the following equation.

D _i = d _i −Δ _i (2) Using these values, gain G, rotation R, trapezoid H,
A correction coefficient for the offset O is calculated.

FIG. 1C is an explanatory diagram of linear interpolation. Figure shows the time course of the offset drift amount, the first scanning time t ₁ of a mark, when the last scan time of the first mark again scanned in the t ₁ ^', the scanning time t _i of the i-th mark The offset drift amount Δ _i is obtained by linear interpolation as shown in the figure.

FIG. 2 is an explanatory diagram of the correction for the measurement deviation amount. The value obtained by subtracting the drift amount Δ _i due to the thermal change of the apparatus from the deviation amount d _{i of} each mark at the scanning time is the true deviation amount D _i .

In the embodiment, the offset drift amount Δ is obtained by scanning all the marks on the wafer. However, the marks are divided into a certain number of units, and all the marks in the unit are scanned to obtain the offset drift amount for each unit. May be asked.

Further, in the embodiment, the interval of the scanning time of each mark is approximated by a time obtained by equally dividing the total scanning time T by the number n of marks, but in the actual exposure, the scanning time of each mark is sequentially stored in the memory. It is more strict to store the offset drift amount Δ _i at the scanning time t _i of the i-th mark by the following equation.

Δ _i = (d ₁ ′ −d ₁ ) × (t _i −t ₁ ) / T (3) where t ₁ is the scanning time of the first mark. Next, the process flow of the embodiment will be described.

(1) Sequential irradiation of marks: movement by stage (2) Acquisition of mark position information: beam deflection for mark scanning (3) Comparison of mark position information and deflection control information: mark center and deflection center Deviation (4) Completion of irradiation of all marks (5) Re-irradiation of first mark (6) Acquisition of position information of first mark: d ₁ , d ₁ '(7) Comparison of position information of first mark twice : Extraction of displacement due to drift Δ = d ₁ '-d ₁ (8) Interpolation of displacement due to drift of each mark: Δ _i (9) Removal of displacement due to drift from displacement of each mark: D _i = d _i −Δ _i (10) Correction for the pattern information exposed in the area corresponding to each mark: (A) Field correction The value corrected here is the deflection position x, y of the beam by Gain G to correct to ', y', rotation R, trapezoid H, displacement amount ( Offset) is a correction coefficient of the O.

X ′ = Gx * x + Rx * y + Hx * x * y + Ox y ′ = Gy * y + Ry * x + Hy * x * y + Oy Here, the gain G is a correction coefficient for the size of the pattern, and the rotation R is a correction coefficient for the direction of the pattern. The trapezoid H is a correction coefficient for the shape distortion of the rectangular pattern, and the positional deviation amount (offset) O is a correction coefficient for the positional deviation of the pattern.

In order to perform field correction, the deflector is provided with a field correction register containing a correction coefficient in the pattern generator of the exposure apparatus. This corrects the deflection positions x and y of the beam with the above-mentioned correction factors G, R, H, and O.

In the field correction, the true positional deviation amount D _i of the present invention when one mark is moved to four corner points of the field and the beam is made to follow the mark is moved by moving the stage.
The correction items G, R, H, and O described above are calculated and stored in the field correction register.

(B) Distortion correction In actual exposure, after the field correction is completed, the positional deviation (distortion) of each point in the field is further corrected with respect to the beam deflection position, and the beam position after the field correction is obtained. The position of the deflector is corrected by adding them.

Next, a configuration example of the exposure apparatus used in the embodiment is shown. FIG. 3 is a diagram showing an example of the configuration of an electron beam exposure apparatus. In the figure, 3 is a sample, 9 is an electron beam exposure apparatus, 10 is a cathode, 11 is a grid, 12 is an anode, 13
a is a rectangular hole, 14 is a focusing lens, 15 is a variable deflector, 16a is a rectangular hole, 17 is a focusing lens, 18a is an aperture,
19a is a refocus lens, 20 is a main deflector, 22 is a central processing unit, 24a, 24b, 24c, 24d are memories, 25 is a correction arithmetic circuit, 26, 27, 33, 34 are DACs (DA converters), 28, 29, 3
5, 36 is an AMP (amplifier), 30 is a register, 31 is an integrator, 32 is an addition and correction arithmetic circuit (related to the correction of the present invention).

Next, position deviation amount D _i than the correction coefficient G,
A specific example of calculating R, H, and O will be briefly described. Letting each component of the amount of deviation from the mark position when the mark is detected by scanning the electron beam on the wafer is (ΔX, ΔY),
Equation (1) is obtained.

[0026]

[Equation 1] Here, X and Y are the deflection positions of the electron beam, A, B, ...
.., O and P are correction factors. Now, assuming that the higher-order terms of the equation (1) are εx (X, Y) and εy (X, Y), the equation (2) is obtained.

[0027]

[Equation 2] In direct writing, generally, marks are provided at four corners of a chip, mark detection is performed to determine correction coefficients A, B, ..., G, and H, and the deflection positions of electron beams are aligned.

Further, the higher-order terms εx (X, Y), εy
Correcting (X, Y) enables logically correct electron beam deflection. Now, with the marks at the four corners (ΔX, ΔY)
(ΔX ₁ , ΔY ₁ ) _, (ΔX ₂ , ΔY ₂ ) _, (Δ
X _3, [Delta] Y _3), when the ([Delta] X _4, [Delta] Y _4), lower order term in equation (2), Equation (3) to obtain (4).

[0029]

[Equation 3] From equations (3) and (4), the following equations (5) and (6) are obtained.

[0030]

[Equation 4] The correction factors A, B, ..., G, H are calculated from the equations (5) and (6).
To obtain the electron beam at the correct position,
Irradiate the next position indicated by (7).

[0031]

[Equation 5] Equation (7) is generally written as the following equation (8).

[0032]

[Equation 6] _{Here, G x = B, G y} = G: Gain _{R x = C, R y =} F: Roteishon _{H x = D, H y =} H: trapezoidal _{O x = A, O y =} E: Offset D _x = εx (X, Y), D _y = εy (X, Y): distortion.

[0033]

According to the present invention, it is possible to eliminate the thermal influence of the exposure apparatus, calculate a true correction coefficient that does not depend on the scanning time of each mark, and realize high accuracy and high resolution. did it.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of an embodiment of the present invention.

FIG. 2 is an explanatory diagram of correction for a measurement deviation amount.

FIG. 3 is a diagram showing an example of a configuration of an electron beam exposure apparatus.

Claims (2)

[Claims] 1. A deviation amount d _i (1 ≦ i ≦) between a beam position and a center position of each mark is obtained by sequentially scanning n marks of a plurality of marks formed on an exposed surface with a charged particle beam.
n) is obtained, the mark scanned first is scanned again, and the displacement amount d ₁ ′ is acquired by scanning the mark lastly, and the difference (d ₁ ′ −d ₁₎ between the measured displacement amounts in the two scans of the first mark is acquired. ) As Δ,
When the total scan time is T, the drift amount delta _i in each mark scanning time obtained in (1), the _{Δ i = Δ × (i-} 1) / n (1) true shift amount D _i of the marks D _i = d _i −Δ _i calculated by equation (2) (2) A charged particle beam exposure method characterized in that the beam deflection position is corrected using this D _i value. 2. In place of the equation (1) according to claim 1, the following equation (3) Δ _i = Δ × (t _i −t ₁ ) / T (3) where t ₁ is the first The scanning time of the mark, t _i is the scanning time of the i-th mark. A charged particle beam exposure method comprising: