JP2506122B2 - Electronic beam exposure method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Outline] Regarding correction of electron beam deflection when a device or reticle is created by directly drawing a pattern on a sample with an electron beam exposure apparatus, a plurality of regions (fields) are combined. The aim is to improve the splicing accuracy of each field when drawing one area so that even a large-scale chip can be drawn with high accuracy. An electron beam exposure method for a sample substrate having a large area, in which the same gain and rotation correction coefficient are applied to the entire surface of the sample substrate with respect to electron beam positioning deflection signal data (M _X , M _Y ) from the field center. (G _X , G _Y , R _X , R _Y ) is defined, and local gains and rotation correction coefficients (g _X , g _Y , r _X , r different for each area subdivided into the areas are defined. define the _Y) Correction operation formula using the distance from the area center to the field center (M _OX, M _OY), Result of correction calculation by Is configured to output.

[Industrial applications]

The present invention relates to correction of electron beam deflection when a device or reticle is created by directly drawing a pattern on a sample with an electron beam exposure apparatus.

The electron beam exposure apparatus forms a pattern using an electron beam. Pattern formation using an electron beam is characterized by being able to draw a fine pattern of a micron or less. Electron beam exposure apparatuses are roughly classified into a Gaussian beam system and a shaped beam system. These are further classified into a stage continuous movement method in which the sample is drawn while continuously moving it, and a step-and-repeat method in which the sample is moved by a predetermined amount for each drawing area.

[Conventional technology]

FIG. 3 is a block diagram of a general electron beam exposure apparatus. The central processing unit (hereinafter referred to as CPU) 1
Controls the entire electron beam exposure apparatus. The disk 2 stores various data including pattern data which is graphic information to be drawn and control information. The interface 3 is an interface between the CPU 1 and each unit described later. Data memory 4
Is a large capacity memory for temporarily storing the pattern data read from the disk 2. The pattern generation unit 5 reads pattern data for one area (a group of a plurality of fields surrounded by alignment marks at four corners) from the data memory 4, and moves the pattern data to a plurality of fields (step and repeat method for stage movement). Without this, disassemble into a range that can be drawn only with the positioning deflector: usually 1-10 mm ^□ ) and create the coordinate data required for exposure. The pattern correction unit 6 corrects the beam deflection amount according to the state (expansion / contraction, rotation, parallel displacement, and height variation) of the drawing area of the sample. The amplifier (AMP) 7a amplifies the blanking signal. Reference numerals 7b, 7c and 7d are circuits each including a D / A converter (DAC) that converts the digital signal from the pattern correction unit 6 into an analog signal and an amplifier (AMP) that amplifies the analog signal. The exposure unit 8 irradiates the sample 9 with an electron beam emitted from the electron gun 8a to draw an image.

As mentioned above, in the step-and-repeat method,
The inside of the area is divided into multiple fields and connected to draw.

FIG. 4 (A) is a diagram showing one area. area
10 is composed of a plurality of fields 11 ₁ , 11 ₂ , ... (In the example shown, one area 10 is composed of 9 fields). Positioning marks 12 ₁ , 12 ₂ , 12 ₃ and 12 ₄ are provided at the four corners of the area 10, respectively. These alignment marks 12 ₁ to 12 _4, the exposure time, while these alignment marks 12 ₁ to 12 ₄ to move the stage is scanned by an electron beam to detect the position, pattern correction based on the detection result The unit 6 corrects the beam deflection amount related to the area 10 and is used for correcting the drawing position.

Here, in general, the following equation (1) is established between the input position data input to the pattern correction unit 6 and the output position data output by the pattern correction unit 6.

Here, M _X , M _Y : Input position data ▲ ▼, ▲ ▼: Output position data G _X , G _Y : Field gain (expansion / contraction) correction coefficient R _X , R _Y : Field rotation correction coefficient H _X , H _Y : Field Trapezoid (height fluctuation) correction coefficient O _X , O _Y : Field offset (parallel displacement)
Correction Coefficients D _X , D _Y : Field Deflection Distortion Correction Coefficients FIG. 5 shows a circuit diagram of the pattern correction unit 6 based on the equation (1). Signal in the figure Indicates a digital multiplier, and the symbol "+" indicates a digital adder.

The correction performed by the above general formula is only for area 10
It is applied relatively accurately when it is considered as one drawing area (because the field correction coefficients G _X , G determined by using the alignment marks 12 _{1 to} 12 ₄ at the four corners of the area 10).
_Y ; R _X , R _Y ; H _X , H _Y ; O _X , O _Y ), and does not apply to each field 11 ₁ , 11 ₂ ,. This is because it is not possible to place alignment marks at all four corners of each field. However, conventionally, even when a plurality of fields are joined to draw one area, a correction coefficient is calculated based on the alignment marks 12 _{1 to} 12 ₄ placed at the four corners of the area 10, and each field is calculated. The method of drawing is commonly used by using these field correction coefficients.

Another conventional method is as follows. This method, as shown in FIG.
The lower left corner of the field 11 ₁ of the alignment mark 13 ₁ placed in three points, 13 reads the ₂ and 13 ₃ by the electron beam, the correction coefficient G _X of Thereby three fields, G _Y; R _X, R _Y; O _X, seek O _Y, by fixing these values in area 10, the other fields 11 _2, to draw a ... in stage movement.

[Problems to be solved by the invention]

However, the above conventional electron beam exposure method has the following problems.

4 alignment marks placed in the 4 corners of area 10 ₁
In the method of obtaining the correction factor with 12 _4, as described above, each field 11 _1, 11 _2, so to draw ... by the same correction factor, the correction value of the area 10 at each position is not reflected. Accordingly, as shown in FIG. 4 (C), which may each field 14 _{1-14 4} does not lead well.

Further, in the method of obtaining the correction coefficient using the _three alignment marks 13 _{1 to} 133, the trapezoidal (height fluctuation) correction coefficient H _X , H _Y
Cannot be calculated, so if area 10 is trapezoidally distorted, it cannot be corrected. Therefore, conventionally, the trapezoidal correction coefficients H _X and H _Y are ignored, or the trapezoidal correction coefficients of the reference chip (1 field) or the adjustment chip are calculated and approximated. Not good.

Therefore, an object of the present invention is to improve the splicing accuracy of each field when drawing a single area by splicing a plurality of fields so that even a large-scale chip can be drawn with high accuracy. And

[Means for solving problems]

1 (A) and 1 (B) are explanatory views of the principle of the present invention.
The present invention is configured by providing an area correction calculation unit 15 in addition to the conventional pattern correction unit 6. The area correction calculator 15
With respect to electron beam positioning deflection signal data, the same gain and rotation correction coefficient are defined over the entire surface of a sample substrate such as a wafer or a glass substrate, and the sample substrate is subdivided (for example, FIG. 1). A local (corresponding to a plurality of fields constituting the area 16 or beam deflection range) gain which is different for each area 16 centered on P shown in (A),
A rotation correction coefficient is defined, the following calculation is performed using this constant, and the result is added to the output value of the pattern correction unit 6.

Here, M _OX, M _OY: distance M _X from the area center P to the field center Q, M _Y: field center (positioning deflection signal data of the electron beam) position coordinates from Q G _X, G _Y: Field gain correction Coefficients R _X , R _Y : Field rotation correction coefficient g _X , g _Y : Area gain correction coefficient r _X , r _Y : Area rotation correction coefficient ▲ ▼, ▲ ▼: Field correction output ΔX, ΔY: Area correction output [Operation] According to the equation (2), a field correction output in which the same gain and rotation correction coefficient (G _X , G _Y , R _X , R _Y ) are defined over the entire surface of the sample substrate such as the wear or the glass substrate is obtained. To be In addition, the area correction output defined by the local gain (for example, for each field) and the rotation correction coefficient (g _X , g _Y , r _X , r _Y ) different for each area can be obtained by the above equation (3). . Then, according to equation (4), the final correction value output Is obtained.

In this way, by adding the area correction calculation unit 15, the correction value at the position in the area 16 (regardless of the field position and the deflector position) can be obtained. Therefore, the joining accuracy of the fields 17 ₁ , 17 ₂ ... In the area is improved, and a large chip can be drawn with high accuracy.

〔Example〕

 FIG. 2 is a circuit diagram of an embodiment of the present invention.

In the figure, the pattern correction unit 6 is one that has been conventionally used, and includes a multiplier 18 ₁ , 18 ₂ , 18 ₃ and 18 ₄ and an adder.
It is configured by including 19 ₁ and 19 ₂ . Each adder and each multiplier add and multiply the illustrated inputs, respectively, and obtain the field correction outputs ▲ ▼, ▲
▼ is output to the adders 20 ₁ and 20 ₂ , respectively.

On the other hand, the area correction calculation unit 15 includes adders 21 ₁ , 21 ₂ , 21 ₃ and 2
1 and _{4 and} multipliers 22 ₁ , 22 ₂ , 22 ₃ and 22 ₄ . The adders 21 ₁ and 21 ₂ execute a part of the operations in the equation (3). The adders 21 ₃ and 21 ₄ and the multipliers 22 _{1 to} 22 ₄ are
The remaining operations in the equation (3) are executed. The adders 20 ₁ and 20 ₂ execute the remaining arithmetic operations of the equation (3). As a result, the final correction value output from each of the adders 20 ₁ and 20 _2. Is output.

Final correction value output obtained in this way Is sent to the D / A converter in the next stage.

As described above, since the correction value at an arbitrary position within the area is obtained, the joint accuracy of each field within the area is improved, and a large-scale chip can be drawn with high accuracy.

Next, another embodiment of the present invention will be described with reference to FIG.

The embodiment shown in FIG. 6 has a configuration in which trapezoidal distortion, offset correction, and deflection distortion amount are corrected in addition to the above embodiment described with reference to FIG. This can be expressed in the following formula.

_{ΔX = g X + r X +} h X + o X (6) ΔY = g Y + r Y + h Y + o Y Here, M _OX , M _OY : Distance from area center P to field center Q M _X , _MY : Position coordinates from field center Q G _X , G _Y : Field gain correction coefficient R _X , R _Y : Field rotation Correction coefficient H _X , H _Y : Field trapezoidal correction coefficient O _X , O _Y : Field offset correction coefficient D _X , D _Y : Field deflection distortion correction coefficient g _X , g _Y : Area gain correction coefficient r _X , r _Y : Area Rotation correction coefficient h _X , h _Y : Area trapezoidal correction coefficient o _X , o _Y : Area offset correction coefficient ▲ ▼, ▲ ▼: Field correction output ΔX, ΔY: Area correction output : Final correction value output The distance from the area center P to each position is calculated by the above equation (5). By substituting this into the formula (6), the area correction values g _X , g for matching the shape of an arbitrary area to be exposed
_Y , r _X , r _Y , h _X , h _Y , o _X , o _Y are obtained. This result and the field correction values G _X , G _Y , R _X , R _Y , H _X , H _Y , O _X , for each conventional field related to the tilt of the stage or the sample obtained by the equation (7)
O _Y , D _X , and D _Y are added as shown in equation (8). Then, the added value is the final correction output, that is, the deflection signal of the electron beam. Is output to the next stage D / A converter.

In this way, by adding the area correction calculation unit 15, the correction value at the position in the area 16 (regardless of the field position and the deflector position) can be obtained. Therefore, the joining accuracy of the fields 17 ₁ , 17 ₂ ... In the area is improved, and a large chip can be drawn with high accuracy.

In FIG. 6, the pattern correction unit 6 is conventionally used, and the multipliers 18 ₁ , 18 ₂ , 18 ₃ , 18 ₄ , 18 ₅ , 18, ₆ and 1 are used.
8 _{7 and} adders 19 ₁ and 19 ₂ are provided.
Each adder and each multiplier adds and multiplies the input shown, the (7) field compensation output ▲ ▼ obtained by the equation, and outputs ▲ ▼ to adders 20 ₁ and 20 _2.

On the other hand, the area correction calculation unit 15 includes adders 21 ₁ , 21 ₂ , 21 ₃ and 2
1 and _{4 and} multipliers 22 ₁ , 22 ₂ , 22 ₃ , 22 ₄ , 22 ₅ , 22 ₆ and 22 ₇ . The adders 21 ₁ and 21 ₂ execute the operation of the equation (5). Adders 21 ₃ and 21 ₄ and multipliers 22 _{1 to} 22 ₇
Executes the operation of the equation (6). Adders 20 ₁ and 20 ₂
Executes the operation of the equation (8). As a result, the adder 20
Each ₁ and from 20 ₂ final correction value output Is output.

Final correction value output obtained in this way Is sent to the D / A converter in the next stage.

As described above, also in the embodiment shown in FIG. 6, since the correction value at any position in the area can be obtained, the splicing accuracy of each field in the area is improved, and the large-scale chip has high accuracy. Can be drawn on.

The present invention is not limited to a method of directly drawing a pattern on a wafer, and can be applied to mask dry plate drawing of a reticle, a mask, etc., and an object having an arbitrary area shape can be created.

Further, the present invention is not limited to the one-deflector system (spot beam or fixed rectangular beam system) having only the positioning deflector, and a variable rectangular electron beam exposure apparatus having two variable rectangular beam forming and positioning deflectors. Applicable.

〔The invention's effect〕

As described above, according to the present invention, the correction output for adjusting the deflection signal of the electron beam to the inclination of the stage or the sample and the correction output for adjusting the shape of the arbitrary region to be exposed are added. Since it is decided to output the data, it is possible to improve the joint accuracy of each field and to draw a large-scale chip with high accuracy.

[Brief description of drawings]

FIG. 1 is a diagram for explaining the principle of the present invention, FIG. 2 is a circuit diagram of an embodiment of the present invention, FIG. 3 is a block configuration diagram of an electron beam exposure apparatus, and FIG. 4 is for explaining a field joining method. FIG. 5, FIG. 5 is a circuit diagram of a conventional pattern correction unit, and FIG. 6 is a circuit diagram of another embodiment of the present invention. In the figure, 6 is a pattern correction unit, 12 _{1 to} 12 ₄ are alignment marks, 15 is an area correction calculation unit, 16 is an area, and 17 ₁ , 17 ₂ , ... Are fields.

Claims (1)

(57) [Claims] 1. An electron beam exposure method for a sample substrate having an area composed of a plurality of fields and provided with alignment marks at the four corners, wherein electron beam positioning deflection signal data (M _X , M) from the center of the field. _Y ), the same gain and rotation correction coefficient (G _X , G _Y , R _X , R _Y ) are defined over the entire surface of the sample substrate, and different localities are obtained for each area subdivided into each area. Gain and rotation correction coefficient (g _X , g _Y , r _X , r _Y ) are defined, and the correction calculation formula is calculated using the distance (M _OX , M _OY ) from the area center to the field center. Result of correction calculation by The electron beam exposure method is characterized in that: