JPS614228A - Charged beam lithography - Google Patents

Abstract

PURPOSE:To facilitate data processing, increase the data processing speed and reduce the amount of required data, by making the best use of the repetition of data which has already been clearly recognized at the designing stage in such a manner that cell data is referred to from a subfield and utilized repeatedly. CONSTITUTION:A decoder 52 interprets lithography data B to obtain subfield specifying data C, offset repeat data D and cell specifying data E. A cell data buffer 51 reads out pattern data F specified by the cell specifying dara E, and delivers the data F to a pattern data decoder 55. In addition, the buffer 51 delivers cell size data G representing the size of the cell concerned to a controller 54. The pattern data decoder 55 interprets the data F and delivers signals for generating a rectangle, a triangle and a segment of a line to a driver 56 for a variable beam shaping device 57. In response to signals from the driver 56, the variable beam shaping device 57 shapes an electron beam 69 from an electron gun 68 so as to form specified pattern data.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention relates to a charged method for drawing patterns of semiconductor integrated circuits such as LSIs on samples such as masks and wafers at high speed and with high precision using electron beams, ion beams, etc. It relates to a beam writing method.

[Technical background of the invention and its problems] Among the drawing methods in an electron beam exposure apparatus, there is a so-called double deflection method that uses two deflectors: a large deflection deflector and a small deflection deflector. This is a DAC (digital-to-analog converter) with a large number of bits and fast response characteristics necessary to draw a wide deflection scanning area at high speed and with high precision.
Since it is difficult to obtain DACs with a large number of bits, positioning of the large deflection area is performed using a DAC with a large number of bits, and a high-speed DA with a small number of bits.
The small deflection area is positioned in step C, the positioning speed of the entire deflection area is increased to the same side, and the drawing throughput is performed in step 9. The aim is to improve performance. An example of a conventional drawing method is shown in FIG.

Here, drawing is performed by sequentially scanning the subfields from the bottom to the top in a survein-like pattern2.Here, the area determined by the deflection width of the electron optical system is - 5 Re・207・−″
'1505'' *KO!!1 is the subfield.Furthermore, when drawing another field, the stage on which the sample is placed, etc. is moved in a step-like manner.

In such conventional drawing methods, the setting of the area and size of the subfield is determined by the conditions of the deflector and the electron optical system, and is determined by taking into account the characteristics and amount of data used for drawing. For this reason, a figure (or a group of figures) of the same shape appears repeatedly within a pattern like a memory element, and CA
D (Even if the Tsukuturn data output from a design system such as Computer Eighth Design 9 is greatly compressed with repetitive information, the repetitive information is not utilized and the repetitive data is divided at subfield boundaries and the data is There was a problem that the processing was complicated, and that the output of data expanded into subfield data was huge. This situation is shown in Figure 2. In this case, the pattern group of (,) Since the repetition pitch and subfield size are different, changing the subfield horizontally (b
) (c), and when the subfields are changed in the vertical direction, the arrangement of the pattern groups shifts as shown in (d) and (e), impairing repeatability. Furthermore, with the rapid advancement of LSI, the number of turn patterns is increasing, and the data processing time is also rapidly increasing, which poses a problem in increasing the drawing throughput of electron beam exposure equipment.

In addition, the positioning of the Zabu field can be performed at the same time as the main deflection scan.
Since this is done by repeating %' turns, if you want to change the positioning of the subfield depending on the conditions of the sample or drawing pattern, the synchronization control of reading and transferring the drawing data from the storage device and the electron beam deflection becomes complicated. There was also a problem that the drawing throughput decreased.

Incidentally, the above-mentioned problem also applies to an ion beam drawing method that uses an ion beam instead of an electron beam.

[Purpose of the invention]

SUMMARY OF THE INVENTION An object of the present invention is to provide a charged beam lithography method that solves the above problems, facilitates and speeds up data processing, reduces data size, and contributes to improved throughput.

[Summary of the invention]

The gist of the present invention is to take advantage of the repetition of tater data, which is clearly considered at the design stage, by referring to cell data from subfields and using it repeatedly.

For the above reasons, in the present invention, each turn data group used for subfield drawing is stored in a storage device as cell data, and when the cell data is read from this storage device and subfield drawing is performed, the drawing data is (1) Subfield position specification data that specifies the position of the subfield to be drawn (2) Cell data name (
(or cell data storage address); (3) offset data indicating the drawing start position of cell data in the relevant subfield; and (4) repeat data representing repetition information of cell data in the relevant subfield. It is possible to read cell data specified by the sack yield from the storage device and repeatedly draw the same setan group within a subfield or between subfields while changing the drawing start position of the cell data.

That is, the present invention divides LSI chip data into field regions determined by the lateral width of the charged beam of the charged beam optical system, and divides this 74-field into a plurality of minute subfield regions. In this electron beam lithography method, in which the field is drawn sequentially, and the pattern of the entire sample is drawn by drawing a plurality of 1-fields, the pattern data group used for subfield lithography is Subfield position specification data that is stored in the storage device in units of cell data defined in the coordinate system of the field, and specifies the position of the subfield where the drawing data should be drawn, and the cell that specifies the cell data that should be drawn with its sub 2 yield. 6 in the specified data, the subfield, and
゛''-7-p tv '4 n!iii M # --
Consists of b ~ part data, repeat data representing repeat information in the X direction and Y direction, and when drawing, the above sack (
- Positioning is performed in the subfield specified by the field position specification data, and the cell data specified by the cell specification data is transferred to the cell data.

In this method, cell data is read out from a subfield storage device, cell data is drawn within a subfield according to the offset data and repeat data, and cell data is drawn overlapping at subfield boundaries as necessary.

〔Effect of the invention〕

According to the present invention, it is possible to refer to common cell data for drawing in multiple subfields, and the drawing start position and number of repetitions can be arbitrarily set for each subfield.
Moreover, since they can be determined based on the periodicity of the /setan without being constrained by subfield boundaries inherent in the device, there are the following advantages. That is, the repetition that was clearly considered at the design stage can be utilized as it is by referring to the cell data from the subfield and repeatedly using it, and it is impossible to keep up with the rapid increase in drawing data as in the past. Data processing becomes easier and faster because graphic divisions at sack boundaries are reduced. Since fewer data numbers are required, there is no need to increase the scale of the storage device, and the risk of overflow is also reduced. Drawing data can be significantly compressed, data transfer efficiency increases, and drawing throughput improves as a result. Furthermore, since the subfield data can arbitrarily specify the sub-2 field 1 position, drawing can be performed without changing the data arrangement in the storage device according to the drawing data and arrangement conditions.

[Embodiments of the invention]

Hereinafter, details of the present invention will be explained with reference to illustrated embodiments.

First, when the structure of the cell data used in the method of this embodiment is described in pseudo-Backus notation, it becomes as shown in 6 [: below.

■ (Cell Library>::=(<Ce1
l DatFL＞)”END■ (Cell Dat
a)::=(Cell ID)(Cell S Gallery e>(
(Patts+rn Data))”■ (Ce11
10):: =BCELL CNAME■ (Cel
l 5ize) :: =CX CY■ (Patte
rn Data〉: : = (R section TI TRIA
NG I LINE) However, BCELL... ID mark indicating the start υ of cell data CNAME... Cell data name CX... Cell size in the X direction CY.
... Cell size in the Y direction RFJCT Curved/ovoid data TR Konichi ... Triangle data LINE ... Indicates line segment data. Also, in the above pseudo-Backus notation,
<> indicates a defined element, and -21 indicates that the left side is defined by the right side. Further, ()'' indicates that the element occurs one or more times, and [] indicates that the element occurs 0 or 1 time. () indicates that any one is selected.

In (2) to (2), ■ indicates that the cell data library is composed of one or more pieces of cell data, and an END mark is attached at the end. '■ indicates that the cell data is composed of a cell ID, a cell size, and one or more gloss turn data. ■B indicates cell IO and cell data start υ
Indicates that it is composed of a CELL mark and a cell data name. (2) indicates that the cell size consists of the X-direction size CX and the Y-direction size CY. ■ indicates that the mother turn data is either rectangle data RECT, triangle data TRTANG, or line segment data LI■. Here, CX and CY are i'? X direction of the turn (or pattern group),
Set it equal to the pitch in the Y direction.

On the other hand, when the structure of the drawing data used in the method of this embodiment is described in pseudo-Backus notation, it becomes as follows.

■ (Draw Data):: =((Subf I
eld ID) ((Cell Data)) )
END■ (Subfield ID)::==BSU
B 5UBFNO-〇(Cell Data)::==
CNAME (Yff5et')) [(Repeat))
■ (Offset)::=OFFSgT OX OY
@) (Repeat:＞::=RgPT IX I
YHowever, BSUB...ID indicating the start of the subfield
Mark 5UBFNO--Positive integer representing the order of arrangement of subfields in the X direction CNAME--Cell data name 0FFSET--Indicates the offset amount -jID mark OX--In the X direction Offset amount, integer OY
...offsera) amount in the Y direction, integer REPT・
...10 marks IX indicating repeated data...
...Number of repetitions in the X direction, integer IY ...Y
Indicates the number of repetitions in the direction, an integer. Therefore, the drawing data in ■ is composed of a collection of one or more subfield data consisting of subfield 9 ID and one or more cell data, and finally the END Indicates that a mark is attached. 1 indicates that the subfield TD consists of a BSUB mark indicating the start of the subfield and a subfield number. The subfield number is 1.2, where n is the number of subfields in the horizontal direction within the field. ... is an integer expressed as 5n. ■ indicates that the cell data consists of a cell data name, offset data, and repeat data. Both the offset data and the first data can be omitted. ◎ indicates that the offset data is composed of the offset data mark 0FFSET, the offset amount ox in the X direction, and the offset amount ox in the y direction) ioY. This indicates that the number of edges is IY.

FIG. 3 is a schematic diagram showing the structure of cell data used in the method of this embodiment. Three cell data cell ], c
ell 2 + cell 3, each cell size is cxl* cyl + cel
l 2 is (!X21cy2, cell 3 is CX31e
)'3. Each cell data is shown in FIG. 7, which will be explained later, and is data in which a turn group used for subfield drawing is defined by subfield coordinates.

FIG. 4 is a schematic diagram showing the structure of drawing data used in the method of this embodiment, and describes subfield 1 and subfield 2. ■ indicates sashi field number 1, 0 indicates cell data c referenced by subfield number 1
ell 1 is shown.

Offset data and repeat data are omitted.

In this case, the default value is offset data of 0.0,
Repeat data is 1,1. ◎ indicates subfield number 1, and indicates that cell data cell 2 is further referenced by offset data OXI + o71 and repeat data 3.3.

0 indicates subfield number 2, and [phase] indicates cell data cell 36 referenced by subfield number 2.

Offset data and repeat data are omitted,
They are 0.0 and 1,1, respectively. O is subfield number 2, and indicates that cell data cell 2 is further referenced by offset data OX2 + 572 and repeat data 3 and 5.

Subfield number 3 and subsequent subfields are omitted. FIG. 5 is a block diagram showing the main structure of an electron beam exposure apparatus used in an embodiment of the method of the present invention. In the figure, 51 is a cell data buffer that stores the cell data A. Reference numeral 52 denotes a drawing data decoder to which the drawing 7'-event B is supplied, and this decoder 52 interprets the data B to obtain subfield designation data C1 offset/repeat data D and cell 3H1 constant data f. .

Then, the subfield designation data C is sent to the main deflection controller 53, and the offset, tsuto, and repeat data D
is sent to the sub-deflection controller 54, and cell designation data E is sent to the seventh data pad 7751.

The cell data buffer 5 reads the pattern data F specified by the cell finger>12 data E, and stores the data F.
is sent to the pattern data decoder 55. At the same time, the buffer 5I sends cell size data G representing the size of the cell to the controller 54. The gloss turn data decoder 55 interprets the above data F and determines whether it is possible.
Signals are sent to the variable beamformer driver 56 to generate rectangles, triangles, and line segments. Then, according to the signal from the driver 56, variable beam shaping is performed to form designated graphic data. Further, the main deflection controller 53 interprets the Sarafeel 1 number based on the input subfield designation data C and sends subfield positioning data to the main deflection driver 58.

Based on the signal from the driver 58, the main deflector 59 deflects and scans the electron beam to a designated subfield position.

Further, the sub-deflection controller 54 generates a sub-deflector scanning control signal based on the input data D and G, and sends the signal to the sub-deflector driver 60. Based on the signal from the driver 60, the sub-deflector 11 determines the graphic drawing position within the sub-field.

Note that the sample 70 to be drawn is placed on a stage 71 that is movable in the X direction and the Y direction.
1 remains in a stopped state until the drawing of one field is completed, and each time the drawing of the field is finished, it is stepped to the position of the next field. .

Next, regarding the drawing method using the above device, 70 in FIG.
- Explain with reference to the chart.

First, cell data A as shown in FIG. 3 is stored in the cell data buffer 51 in advance. In this state, when drawing data B as shown in FIG. 4 is input to the drawing data decoder 52, the decoder 52- inputs subfield designation data (sub number) C, offset/repeat data D, and cell designation data (cell Data name) E is obtained. That is, the decoder 52 obtains the sub number, the first cell data name, and the offset/repeat data corresponding to the cell data name from the first drawing data. decoder 5
2, when the system number is obtained, the main deflection controller 5
3, sub-positioning data corresponding to the sub-number is generated, and the data is sent to the main deflection driver 58.

Further, when the first cell data name is obtained by the decoder 52, second turn data corresponding to the cell data name is read from the cell data buffer 51, and the data is sent to the pattern decoder 55. At the same time, the cell size data of the cell in question is sent from the par/7a 5 to the sub-deflection controller 54. The decoder 55 generates a signal for generating a graphic corresponding to the pattern data, and sends the signal to the variable beamformer driver 56.

Furthermore, when the decoder 52 obtains offset/repeat data corresponding to the above cell data name, the sub-deflection controller 54 generates a sub-deflection scanning controller signal, and this signal is sent to the sub-deflection driver 60. . and a variable beam shaper 57. connected to each driver 56.58.60. The main deflector 59 and the sub-deflector 6 draw /4' turn data corresponding to the cell data name in the subfield corresponding to the sub number.

Next, when the drawing of the pattern data is completed, it is determined whether or not the drawing data includes the next cell data name. Here, if there is a next cell data name, the next cell data name is read out by the decoder 52, and the main turn data corresponding to the cell data name is drawn in the same manner as before. On the other hand, if there is no next cell data name, it is determined whether there is next drawing data. If there is next drawing data, the decoder 52 reads out a sub number from the next drawing data and shifts it, and draws the subfield corresponding to the sub number in the same way as before. Furthermore, if there is no next drawing data, it is assumed that the drawing of this field has ended, and the drawing of the next field is started.

FIG. 7 is a diagrammatic representation of the cell data shown in FIG.
rcell 2 + cell 3 is shown. FIG. 8 shows drawing 1 when drawing is performed using the apparatus shown in FIG. 5 using the above cell data and the drawing data shown in FIG. 4.
An example is shown. Figure 8 (-) shows the cell data after drawing the cell data of cell 1 shown in Figure 7 (-).
The p4 turn of Zabu field 1 is obtained by drawing the cell data of 2 in a 3×3 array from the position (oxl*oyt). Figure 8(b) shows the cel shown in Figure 7(c).
After drawing the cell data of cell 13, the cell data of cell 2 is drawn in a 5×3 array starting from the position (ox2 + o72') to form the pattern t[ of subfield 2.

Here, it can be seen that Q10<O, and the drawing start position of cell2 in subfield 2 has jumped out of the subfield in the X direction. As described above, both the x and y coordinates of the drawing start position can take negative values as necessary. As a result of the above, drawing/9 turns as shown in FIG. 8(c) is obtained.

Thus, according to this embodiment, by configuring the cell data as shown in FIG. 3 and the drawing data as shown in FIG. 4, it is possible to refer to common cell data for drawing a plurality of subfields. It becomes possible. ! Furthermore, the drawing start position and the number of repetitions can be arbitrarily set in each subfield, and these can be determined based on the periodicity of the arm turn without being constrained by subfield boundaries specific to the apparatus. Therefore, as described in the [Effects of the Invention] section, there are advantages such as easy and speedy data processing, a small amount of data, and improved data transfer efficiency, resulting in a significant increase in throughput. It is possible to make significant improvements.

Note that the present invention is not limited to the embodiments described above. For example, by adding subfield repetition information to the drawing data, compression in subfield units becomes possible. Further, in the embodiment, an electromagnetic deflector is used for large deflection and an electrostatic deflector is used for small deflection, but it is also possible to use an electrostatic deflector for both large deflection and small deflection.
Furthermore, the combination and positional relationship of the deflectors are arbitrary. In addition, it is not limited to the method in which the stage on which the sample is placed is stopped when drawing the field; for example,
It is also possible to apply this method to a method in which the stage is continuously moved when drawing a field, as shown in this issue. In this case, as a bias voltage for large deflection, a deflection voltage proportional to the periodic movement that can follow the movement of the stage is required. Furthermore, it is also possible to apply the present invention to an ion beam drawing method that uses an ion beam instead of an electron beam. In addition, various modifications can be made without departing from the gist of the present invention.

[Brief explanation of drawings]

Fig. 1 is a schematic diagram showing a conventional drawing method, Fig. 2 is a schematic diagram illustrating problems with the conventional drawing method, and Figs. 3 to 8 each illustrate an embodiment of the present invention. Fig. 3 is a schematic diagram showing an example of the structure of cell data, Fig. 4 is a schematic diagram showing an example of the structure of drawing data, and Fig. 5 is the main part configuration of the electron beam exposure apparatus used in the example method. FIG. 6 is a flowchart showing how to draw both culms using the above device, FIG. 7 is a schematic diagram showing the cell data described above, and FIG. 8 is a schematic diagram showing drawn subfields. It is. 5 No. Cell data buffer, 52 Drawing r-ta decoder, 53 Main deflection controller, 54...
Sub-deflection controller, 55... Pattern data decoder, 56... Variable beam former driver, 57...
Variable beam shaper, 58... Main deflector driver, 59
...Main deflector, 60...Sub-deflector driver, 6-
...Auxiliary flattening organ. Applicant's agent Patent attorney Takehiko Suzue Figure 1 Figure 2 (b) (c) Figure 61'53 Figure Otsu'54

Claims (2)

[Claims] (1) Divide the chip data of the integrated circuit into field regions determined by the charged beam deflection width of the charged beam optical system, divide this field into multiple minute subfield regions, and write the subfields sequentially to create the field. In a charged beam lithography method in which a pattern is drawn on the entire sample by drawing a plurality of fields, a group of pattern data used for subfield lithography is a group of predetermined patterns defined in the coordinate system of the subfield. Subfield position specification data that is stored in a storage device as cell data in units and specifies the position of a subfield in which drawing data should be drawn; cell specification data that specifies the cell data to be drawn in that subfield; and the subfield. It consists of offset data representing the drawing start position of the cell data in , and repeat data representing repetition information in the X and Y directions.When drawing, positioning is performed in the subfield specified by the subfield position designation data, and the above A charged beam drawing method, characterized in that the cell data specified by cell designation data is read from the cell data storage device, and the cell data in the subfield is drawn in accordance with the offset data and repeat data. (2) The charged beam drawing method according to claim 1, characterized in that when drawing cell data within the subfield, the cell data is drawn so as to overlap the subfield boundary as necessary.