JPS63248129A - Charged particle beam lithography - Google Patents

Abstract

PURPOSE:To improve the accuracy of directly written patterns extending over in X and Y directions while correcting any shifting errors of stage and slip of beams etc., by means of providing fields having respectively overlapping framed regions before writing patterns. CONSTITUTION:Before writing rectangular patterns, when the data inside the pattern from a magnetic disc 13 extend over multiple fields, the patterns are divided into two regions, i.e. the framed regions and the residual regions excluding the framed regions by a controller 7. The position, type of fields containing respectively divided patterns, initial positions and respective dimensions in X and Y directions are stored either in built-in memory or in an outer memory. When the divided and stored patterns are read out and fed to deflectors 5 and blanking deflectors 8 to be written, the electron beams from an electron gun 1 are focussed by focussing lenses 2, 3 to be directly written by the deflectors 5 on the specified position so that the accuracy of directly written patterns extending over in X and Y directions may be improved.

Description

[Detailed description of the invention] [Industrial application field] The present invention relates to a drawing method with improved drawing accuracy.

[Prior art 1] When a pattern is drawn on a material by irradiation with a charged particle beam, the pattern is usually divided into areas (fields) in which the pattern can be drawn while keeping the deflection aberration within an allowable range according to the deflection length of the beam. Each time a pattern in each field is drawn, the stage on which the material is placed is moved so that the beam comes to the center of each field, and the beam is placed at a predetermined position on the material to draw the pattern.

By the way, as shown in FIG. 6, the pattern P is at least
, Y. A plurality of (two in this case) fields Fa, Fb1 . : If the data of the pattern P is divided into the fields Fa and Fb, for example, first move the stage on which the material is placed so that the axis of the beam is at the center of the field Fa, and divide the beam. Pattern Pa
Then, the stage is moved so that the axis of the beam is at the center of the field Fb, and the beam is shot to the position where the divided pattern Pb is to be drawn. The pattern P is drawn on the material by drawing the divided pattern.

[Problems to be solved by the invention] However, when drawing such a pattern, it is difficult to move the stage.
Due to the J error, beam shift, etc., as shown in FIG.
b occurred, deteriorating the drawing accuracy.

The present invention is aimed at solving such problems.

[Means for Solving Problems 5] Therefore, in the present invention, before pattern drawing, fields F+ and Fz each having four overlapping border areas are provided, and the data of the pattern extending over each of the fields F+ and F2 is , the field position where each divided pattern exists, the field position where each divided pattern exists, and the field position where each divided pattern exists, Data regarding whether or not the pattern is drawn in the 7-edge area and the shot position of the pattern are stored in a memory, and each data is read from the memory when the pattern is drawn, and the data is written in an area obtained by subtracting the 7-edge area from the field. When drawing a divided pattern, a beam of prescribed m is irradiated, and when drawing a pattern divided into bordered areas, a beam of half the prescribed ω is irradiated to write each pattern.

[Example] FIG. 1 is a schematic diagram of an electron beam lithography apparatus for carrying out the lithography method of the present invention, in which an electron beam from an electron gun 1 is focused onto a material 4 by a focusing lens 2.3, and at the same time The material is appropriately scanned with a beam using a deflector 5, and a pattern is directly drawn at a predetermined position on the material. In the figure, 6 is a deflection signal generating circuit which supplies a beam shot position signal to the deflector 5 based on a command from the control device 7. 8 is a deflector for blanking, 9 is a slit for blanking, 1
0 is a blanking signal generating circuit, which supplies a blanking signal to the blanking deflector 8 based on a command from the control device 7. Reference numeral 11 denotes a stage drive mechanism that moves the stage 12 based on commands from the control III device 7. Reference numeral 13 denotes a magnetic disk that stores data such as the position and size of a pattern created by, for example, a CAD (computer aided design) device based on a blueprint.

In the following description, for convenience of explanation, a case will be described in which a rectangular pattern is drawn.

First, before actually drawing a pattern, in the control device 7, among the pattern data from the magnetic disk 13, if the data spans multiple fields, it is divided into a border area and an area obtained by subtracting the border area from the field. The field position where each pattern exists, the type of pattern, the initial position of the pattern, the dimension of the pattern in the X direction and the dimension in the Y direction, etc. are stored in a built-in memory or an external memory.

For example, patterns P+, F2 . F3. F4, Ps, F6,...
The data of . The field position, the type of pattern, the initial position of the pattern, the dimension of the pattern in the X direction and the dimension in the Y direction, etc. are stored in the memory.

First, regarding the pattern P1, the X coordinate X1 of the initial position of the pattern P1 and the rightmost position fff of the field F1.
By comparison with By comparing the fixed position and the right end position of field F1 x 1, the pattern P1 is
It can be seen that there is a part that straddles to the right side of the rightmost position M X I of field F. In this way, for a pattern that has a part that straddles to the right side of the right edge position x 1 of field F+, the data of the part in field F2 has a border f that has a size d that is considerably smaller than the dimension f of one side of the field.
[Divided by E and the area (F+E) obtained by subtracting the bordered area E from the field F1, the data of the portion within field F2 is divided by the bordered area E and the bordered area E from the field F2. The territory bib drawn (Fz
Divide by E). As a result of this division, pattern P1 is divided into (F+E) areas Plo, F12 which is divided into bordered areas E in field F2,
(F2 E) F13 divided into areas and F12 divided into bordered area E in field F2
becomes.

Next, regarding pattern P2, by comparing the X coordinate eX X z of the initial position of pattern P2 with the right end position x1 of field F1, the Further, by comparing the sum of the X coordinate of the initial position of the pattern P2 x 2 and the X direction dimension H2 with the right end HXt of the field F1, it is found that the pattern P2 is located at the right end of the field F1. It is found that there is no straddling part to the right of position x1, and this binding data of the pattern P1 becomes data drawn by field F+length.

The position of the right end of pattern P3 is the right end position X of field F1! However, since it is not to the right of the right end position x 1, the data of the pattern P3 is data drawn by field F+length in the same way as pattern P2.

Next, regarding pattern P4, a comparison similar to that described above is performed, but in this comparison, a comparison is made between the right end portion 11×2 of field F2.

As a result of this comparison, it is found that the X coordinate x 4 of the initial position of pattern P4 is on the left side of the right end position x 2 of field F2, and furthermore, the X coordinate x 4 of the initial position of pattern P4 and the X direction dimension H4 are added. Field F2
By comparing the right end position x 2, it is found that the pattern P4 has no part that straddles to the right side of the right end M becomes.

Next, a similar comparison is made for the pattern Ps, but in that case, the pattern P5 is as shown in (A) in the figure.
If there is no part extending to the right side of the right end I x 2 of field F2, the data of pattern P5
The data is drawn in 2 lengths. However, as shown in (b), if there is a part that straddles to the right side of the right end position x 2 of field F2, the data of the part in field F2 is divided into the border area E' and the border area from field F2. 6 minus area E'! ? (F2-E-), and the data of the portion within field F2 is divided into the bordered area E' and the area (F3--E-) obtained by subtracting the bordered area E' from the field F3'. Divide by. As a result of this division, the pattern P5 is divided into (F2-E-) regions P s t , Ps z divided into 7 plover regions E- in field F2, (F3--E
-) divided into regions F53 and field F3
This is F52, which is divided into bordered areas E' within -.

Then, regarding the next pattern P6, if there is no pattern in field F2 that straddles the field to the right as in the former case (a), the same comparison as above is performed and the field "drawn in 3 lengths" is However, if there is a pattern in field F2 that spans the field to the right as in the latter case, the data will be drawn in field F3-.

In this way, among all the pattern data from the magnetic disk 13, data that spans multiple fields is divided into a border area and an area obtained by subtracting the border area from the field, and the field position where each divided pattern exists, The type of pattern, the initial position of the pattern, the dimensions of the pattern in the X direction and the Y direction, etc. are stored in a built-in memory or an external memory. FIG. 3 shows the data of each pattern and divided pattern stored at each address in the memory. Six addresses are used for data for each pattern and one divided pattern, and the first The field position where the pattern exists, starting from the top,
Pattern type, X coordinate of the initial position of the pattern,
The Y coordinate, the dimension in the X direction, and the dimension in the Y direction of the pattern are shown. Figures 3(b) and 3(c) show the continuation of (a) showing memory addresses 1 to 42, respectively, and (b) shows the above (
(C) shows the continuation of case (B) above, and (C) shows the continuation of case (B) above. In each figure, as described on the right side of each address, addresses 1 to 6 are data of division pattern P11, and addresses 7 to 12 are data of division pattern P.
12 data, addresses 13 to 18 show data of pattern P2, . . . , addresses 67 to 72 show data of divided pattern PS3, and so on.

Note that the lowest data "1" of the data stored at each second address indicates that the pattern is rectangular, and when the second data from the highest is 111 II, it is data within the border area. When it is II OII, it indicates that the data is within the field or within the area obtained by subtracting the 7 Chidori area from the field.

Furthermore, when drawing a pattern in a field and/or when drawing a pattern divided into areas obtained by subtracting the border area from the field, a beam of a prescribed irradiance is used, that is, a beam of sufficient amount to expose the resist coated on the monthly rate. Shot time correction data is stored in the storage device of the control device so that when drawing a pattern divided into bordered areas, half of the prescribed amount of positive beam is irradiated.

Thus, each pattern P+, P2. P3, P4゜Ps
, Pg, . . . directly on the monthly interest rate, the control device 7 sequentially reads data from address 1 of the memory and issues each command. First, the control device 7 issues a command based on the field position data at address 1 to the stage drive mechanism 1.
Send to 1. Then, the stage drive mechanism 11 moves to the field F! The stage 12 is moved so that the axis of the beam is centered at the center of the stage 12. Further, the control device sends a command based on the data at address 2 to the blanking signal generation circuit 10. Then,
The blanking signal generation circuit 10 uses this pattern (P+
Since the pattern 1) is not a border IA 14 pattern, a blanking signal having a shot time signal such that a prescribed amount of beam is irradiated when drawing the pattern is supplied to the blanking deflector 8. Further, the X and Y coordinate data of the initial positions of the patterns at addresses 3 and 4 and the dimension data of the X and Y directions of the patterns at addresses 5 and 6 are sent to the deflection signal generation circuit 6. Then, the deflection signal generation circuit sends to the deflector 5 a position signal that causes the beam to be shot at the drawing position of the divided pattern P1□. As a result, the divided pattern Pa+ is drawn in the field F1 with a prescribed amount of beam.

Similarly, based on the command based on the field position data at address 7, the stage drive mechanism 11 moves the stage 12 so that the beam axis is located at the center of the field F1. In this case, since the previous process and command are the same, the stage does not actually move from its previous position.

Also, according to the command at address 8, the blanking signal generation circuit 1
Since this pattern (P+z) is a pattern in a border area, a blanking signal having a shot time signal such that a beam of half the prescribed amount is irradiated when drawing the pattern is supplied to the blanking deflector 8. In addition, the X and Y coordinate data of the initial position of the pattern at addresses 9 and 10, and the X direction and Y coordinate data of the pattern at addresses 11 and 12.
The directional dimension data is sent to the deflection signal generation circuit 6. Then,
The deflection signal generation circuit sends a position signal to the deflector 5, which causes the beam to be shot at the drawing position of the divided pattern P12. As a result, the divided pattern P12 is drawn in the bordered area E of the field F1 with a pattern that is half of the specified ω.

Similarly, the pattern P is determined by the data at addresses 13 to 18.
2. A pattern P3 is drawn within each field using the data of addresses 19 to 24 using a prescribed amount of beam.

Next, in response to a command based on the field position data at address 25, the stage drive mechanism 11 moves the stage 12 in the X direction (f-
d) move the distance. Also, according to the command at address 26, the blanking signal generation circuit 10 generates this pattern (P+2
) is a pattern in a fudge area, so a blanking signal having a shot time signal is supplied to the blanking deflector 8 so that half of the prescribed amount of beam is irradiated when drawing the pattern. Further, the X and Y coordinate data of the initial positions of the patterns at addresses 27 and 28 and the dimension data of the patterns in the X and Y directions at addresses 29 and 30 are sent to the deflection signal generation circuit 6. Then, the deflection signal generating circuit sends a position signal to the deflector 5, which causes the beam to be shot at the drawing position of the divided pattern P12. As a result, division 1<turn P12 is drawn in the bordered area E of the field F2 using a beam that is half the prescribed amount.

Similarly, the pattern P is determined by the data at addresses 31 to 36.
+ 3. Pattern P based on data from address 37 to address 42
4 are each written with a prescribed amount of beam in the field F2.

Next, when the pattern Ps is in the state shown in FIG. 2(a), it is drawn according to the address data shown in FIG. 3(b). That is, the pattern P5 is written in the field F2 using the data of addresses 43 to 48 using a prescribed amount of beam. Then, based on the command based on the field position data at address 49, the stage 12 is moved a distance f in the X direction so that the beam axis is at the center of field F3, and based on the command based on the data at addresses 50 to 54, In field F3, pattern P6 is written with a prescribed amount of beam.

However, when in the state (b), drawing is performed according to the data at the address shown in FIG. 3(c). That is, addresses 43 to 48
The pattern Ps1 is written in the field F2 with the specified amount of beam according to the address data, and the pattern PS2 is written in the field F2 with the beam of half the specified amount according to the data of addresses 49 to 54. Then, with a command based on the field position data at address 55, field F3
The stage 12 is moved a distance (f-d) in the X direction so that the beam axis is centered at the center of the field F3', and the pattern Pg is placed in the field F3' by a specified amount according to the command based on the data at addresses 56 to 60. It is drawn in the throat. Furthermore, from 61st
Pattern P52 is field F due to the data at address 66.
Pattern P53 is written in field F3- with a beam of half the specified amount in field F3', and based on the data at addresses 67 to 72, pattern P53 is written in field F3- with a beam of a specified amount.

Now, among the drawn patterns, the pattern divided into the border area and the area obtained by subtracting the border area from the field (Ps in the case of P+ or (0)) is the difference between fields due to stage movement error or beam shift. If there is no deviation at all, as shown in Figure 4(a), there will be no step in the drawn pattern at all, and since the pattern in the border area has been double exposed, there will be no difference in the inner gold of the pattern. A specified hanging beam is illuminated. In addition, when a shift between fields occurs due to a stage movement error, a beam shift, etc., for example, using pattern P1 as a representative, as shown in FIG. 4(b), the (Fl-E) area Steps Go and Go − occur at the boundary between pattern P++ and pattern P12′ of bordered area E in field F2, and between pattern PI3 in the (F2-E) area and pattern P+z of bordered area E in field +. However, such a difference in level does not occur, and the connection between these boundaries becomes extremely smooth as shown in FIG. 4(C).

The reason for this is that the portions where the patterns P+ + and P+ 3 and the patterns P12 and P12' in the bordered area overlap are irradiated with a specified amount of beam, and the portions where the patterns P12 and F12- in the bordered area do not overlap are irradiated with a specified amount of beam. Since half of the amount of beam is irradiated, the area near the upper boundary culm of pattern P++, which is irradiated with the specified amount, Q1. Near the upper boundary of the part where patterns PL2 and P12' of Q2 and 7 Chidori areas overlap Q3.04 The charges of each Q3.04 are scattered between each other, so that the beam of half of the specified amount is irradiated on the Chidori area. This is because, among the parts where the patterns P12 and P12' do not overlap, only the parts between Q1 and 03 and between Q2 and 04 are exposed.

That is, this is because a proximity effect occurs between them.

Note that the present invention is not limited to the above embodiments. For example, in the above embodiment, if there is no pattern that extends over an adjacent field within a field, no border area common to that field and the adjacent field is provided, but as shown in FIG. As shown, whether there is one or more patterns that span multiple fields or not,
All pattern data is stored in each field F1. F2. F3. FA, etc., and as in the above embodiment, the field position where there is a pattern that does not span multiple fields, and the bordered area because it spans multiple fields. Create data such as the field position where the divided pattern exists, the type of pattern, the initial position of the pattern, the dimension in the X direction and the dimension in the Y direction, etc. If it is stored in the memory, when drawing a pattern in each field, the stages Fl, F2, . F3. F4. It suffices to move the beam a certain distance so that the axis of the beam is centered on the center of the beam, which simplifies control.

In the above embodiment, a pattern that is long in the X direction has been described, but a pattern that is long in the Y direction or a pattern that is long in both the X and Y directions can be similarly formed.

Further, in the above embodiments, writing using an electron beam was taken as an example, but writing using other charged particle beams is also applicable.

[Effects of the Invention] According to the present invention, when drawing a pattern spanning a plurality of fields on a material at least in either the X or Y direction,
For example, even if there is a stage movement error or beam shift, there will be no step-like misalignment at the joints between patterns as in the conventional method, and there will be extremely smooth connections near the pattern joints. The accuracy of direct drawing can be improved by only generating .

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram of an electron beam lithography apparatus for carrying out the lithography method of the present invention, FIGS. 2 to 4 are diagrams used to explain one embodiment of the present invention, and FIG. 6 and 7 are diagrams used to explain the conventional drawing method. 1: M gun 2, 3: Focusing lens 4: Material 5
: Deflector 6: Deflection signal generation circuit 7: Control device 8
Deflector for Ni Blanking 9 Slit for Ni Blanking
10 Ni blanking signal generation circuit 11: Control device stage drive mechanism 12: Stage 13: Magnetic disk Patent applicant JEOL Ltd. Figure 1 12 l/ (3) (b) (C) Figure 7

Claims (1)

[Claims] Before pattern writing, fields F_1 and F_2 each having an overlapping border area are provided, and each field F_1 is
, F_2 is divided into a bordered area and an area obtained by subtracting the bordered area from the field, so that the parts divided by each bordered area are overlapped, and at least each of them is divided. Data regarding the field position where the pattern exists, whether or not the pattern is drawn in the border area, the shot position of the pattern, etc. are stored in a memory, and when the pattern is drawn, each data is read from the memory and the field When drawing a pattern divided into areas obtained by subtracting the border area from the border area, a prescribed amount of beam is irradiated, and when drawing a pattern divided into border areas, a beam of half the specified amount is irradiated to write each pattern. A charged particle beam drawing method that resembles drawing.