JP2894746B2 - Charged beam drawing method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Purpose of the Invention] (Industrial application field) The present invention relates to a charge for drawing a pattern of a semiconductor integrated circuit such as an LSI on a sample such as a mask or a semiconductor wafer with high speed and high accuracy. The present invention relates to a beam writing method, and more particularly, to a charged beam writing method for performing writing processing by decoding writing conditions for writing the pattern from a pattern identification name.

(Prior Art) In recent years, LSI patterns have become increasingly finer and more complex, and electron beam lithography systems have been widely used as devices for forming such patterns. When drawing a desired pattern using this device, design pattern data created using an LSI pattern design tool such as CAD cannot be supplied as the drawing pattern data in the format as it is. Data must be converted into drawing pattern data that can be accepted by the drawing device.

In this data conversion processing, for example, contour processing is performed on the design pattern data to remove multiple exposure portions, and thereafter, a rectangular, trapezoidal, triangular, or the like is applied to each unique unit drawing area (frame area) determined by a beam deflection area. Is divided into the basic figures, so that the figure data is acceptable to the electron beam drawing apparatus. As a result, drawing pattern data relating to the integrated circuit and chip arrangement data defining where the LSI chip is to be drawn on the sample are generated, and these data are stored in a storage medium such as a magnetic disk.

In the drawing processing step, from the data stored in the storage medium, drawing pattern data representing the drawing pattern for each frame area is read out to a pattern data buffer, temporarily stored in the pattern data buffer, and this data is decoded. Is divided into a group of drawing unit figures (a rectangle with a limitation on the figure size and a triangle with a limitation on the shape and size) that can be formed by the beam shaping means. Then, the beam position and beam shape are controlled based on the graphic data obtained as a result, and the table on which the sample is placed is continuously moved in the X direction or the Y direction so that the desired pattern is formed in the frame area. draw.

Next, the table is moved stepwise in the direction orthogonal to the continuous movement direction by the width of the frame area, and the above processing is repeated to perform the drawing processing of the entire desired area. In the two-stage deflection system in which the main deflection unit determines the sub-deflection position and performs drawing by the sub-deflection unit, a frame region is formed by an aggregate of unit drawing regions (sub-fields).
The width of the frame area is defined by the beam deflection width of the main deflection unit. Also in this drawing method, drawing pattern data is read out for each frame area in the same manner as described above, and drawing processing is performed while continuously moving the table.

In the case of a mask, for example, a sample written by such a writing process is mainly used for light exposure and is used for a reduction projection exposure apparatus. In addition to the LSI pattern, the mask is typically represented by an alignment mark or a writing accuracy evaluation mark required for high-precision alignment when transferring the LSI pattern formed on the mask by the reduction projection exposure apparatus to a wafer. It is necessary to form various mark patterns to be formed.

In addition, it is necessary to perform pattern inversion in black and white or mirror inversion based on the process conditions after the pattern drawn on the mask is transferred to the wafer. Further, even when the sample to be drawn is a semiconductor wafer and the pattern is to be drawn directly by an electron beam, it is necessary to reverse the pattern to be drawn in black and white or to perform mirror inversion to perform the drawing process.

Conventionally, in order to set the selective arrangement of a mark pattern on a mask as described above, necessary mark data is arranged at a necessary position in a processing step of converting design pattern data into drawing data. Further, drawing conditions such as black and white reversal and mirror reversal have also been defined as part of the drawing pattern data.

However, this type of method has the following problems. That is, the types and arrangement positions (mark information) of the mark patterns formed on the mask are various depending on the type of the reduction projection exposure apparatus for exposing the mark pattern onto the wafer and the drawing conditions of the mask. The work of selecting and arranging the types of marks in consideration of the above relied on humans. Therefore, errors have occurred in the types and arrangement positions of the marks, and the drawn mask has become a defective product. Further, an operator who converts the design pattern data into drawing pattern data, an operator who draws a pattern on a mask based on the drawing pattern data, and an operator who transfers a mask pattern to a wafer are usually separated. For this reason, when the conditions such as the type and position of the mark to be drawn change, the information is sequentially transmitted from the operator of the pattern transfer to the operator of the data conversion. The efficiency was very poor, and mistakes were often mixed.

Further, drawing conditions such as black-and-white reversal and mirror reversal drawing are also defined by the data conversion operator based on a specific drawing in the same manner as described above.
As a result, the drawn mask or wafer may be defective.

In this situation, the number of masks to be drawn tends to increase exponentially with the spread of automatic design of LSI patterns and conversation design methods using workstations. Has been a major obstacle to the progress of the project.

(Problems to be Solved by the Invention) As described above, conventionally, mark information and drawing conditions are artificially designed by an operator in a process of converting design pattern data into drawing pattern data.
For this reason, the work efficiency is low, and this has been a factor of generating a defective mask or a defective wafer. The above problem is not limited to the electron beam drawing method, but can be similarly applied to an ion beam drawing method using an ion beam.

The present invention has been made in consideration of the above circumstances, and an object of the present invention is to eliminate the occurrence of a defective mask or a defective wafer which has conventionally been caused by manual operation, and to perform drawing by a charged beam drawing apparatus. It is an object of the present invention to provide a charged beam writing method capable of improving the reliability of the LSI pattern and the operation rate of the device.

[Structure of the Invention] (Means for Solving the Problems) The gist of the present invention is to decode mark information, drawing conditions, and the like from an identification name of a pattern to be drawn on a mask or a wafer, thereby forming a peripheral mark necessary for drawing. It is to automatically set a pattern, a drawing condition, and the like.

That is, the present invention provides an LSI for a mask used for pattern transfer.
In a charged beam writing method for writing a pattern, mark information (for example, the type, arrangement position, and position of a mark pattern) of a peripheral mark pattern to be located at a peripheral portion of a main pattern is obtained from a pattern identifier representing a pattern to be written on the mask. In addition to decoding the drawing order), the drawing conditions of the mark pattern and the main pattern (for example, black and white reversal of the drawing pattern, mirror reversal, drawing magnification, beam irradiation amount during drawing, and proximity effect correction procedure) are also decoded.
In this method, a main pattern and a peripheral mark pattern are drawn on the mask based on the result of the decoding.

(Operation) According to the method of the present invention, mark information and drawing conditions are automatically set from the pattern identification name, so that the data conversion operator is determined according to the model of the reduction projection exposure apparatus and the drawing conditions of the mask. A complicated operation of selecting and arranging various types and positions of various marks can be eliminated, and occurrence of a defective mask due to a human error can be eliminated. Further, it is possible to eliminate any setting error of the drawing condition such as black-and-white inversion and mirror inversion drawing from the same background as described above. Further, it is possible to quickly respond to changes in the drawing conditions (including changes and additions of the types and arrangement positions of marks to be drawn). As a result, it becomes possible to increase the operation rate of the charged beam writing apparatus and to improve the productivity of the LSI. Further, the above-described drawing method is expected to exert more effective effects on miniaturization of patterns and improvement in integration degree and increase in the number of drawing patterns accompanying the rapid progress of LSI in the future.

(Examples) Hereinafter, details of the present invention will be described with reference to the illustrated examples.

FIG. 1 is a schematic configuration diagram showing an electron beam writing apparatus used in a method according to one embodiment of the present invention. In the drawing, reference numeral 10 denotes a sample chamber, in which a table 12 on which a sample 11 such as a glass mask or a semiconductor wafer is placed is accommodated. The table 12 is driven by a table driving circuit 13 in the X direction (left and right directions on the paper) and the Y direction (front and back directions on the paper). The moving position of the table 12 is measured by a position circuit 14 using a laser length meter or the like.

An electron beam optical system 20 is arranged above the sample chamber 10. The optical system 20 includes an electron gun 21, various lenses 22 to 26,
It comprises a blanking deflector 31, a beam size varying deflector 32, a beam scanning main deflector 33, a beam scanning sub deflector 34, and beam shaping apertures 35 and 36.
The main deflector 33 is used to position a predetermined unit drawing area (sub-field), the sub-deflector 34 is used to position the figure drawing position in the sub-field, and the beam size changing deflector 32 and the shaping aperture 35 are used. , 36, the table 12 is continuously moved in one direction, and a drawing stripe region is formed by collecting frame regions obtained by dividing the LSI chip pattern region into strips according to the beam deflection width. Further, the table 12 is step-moved in a direction orthogonal to the continuous movement direction, and the above processing is repeated to sequentially draw each drawing stripe region. When drawing only one chip, the frame area is drawn sequentially in the same manner as described above.

On the other hand, the control computer 40 has a magnetic disk (storage medium) 41.
Are connected, and the disk 41 stores the drawing pattern data of the LSI chip. The drawing pattern data of the LSI chip read from the magnetic disk 41 is temporarily stored in a pattern memory (data buffer unit) 42 for each drawing stripe area. The drawing pattern data for each drawing stripe stored in the pattern memory 42, that is, the drawing stripe information including the drawing position and the basic graphic data is decoded by the pattern data decoder 43 and the drawing data decoder 44, which are data decoding units. The signals are sent to a blanking circuit 45, a beam shaper driver 46, a main deflector driver 47 and a sub deflector driver 48.

That is, the pattern data decoder 43 receives the above data, and performs inversion processing on the graphic data included in the drawing stripe area as necessary to generate inverted pattern data. Then, the basic figure data defined as the drawing stripe data is divided into drawing unit figure groups that can be formed by the combination of the shaping apertures 35 and 36, and beam control data is created based on this data. It is sent to the ranking circuit 45. Then, desired beam size data is created, and control data for this beam shaping is sent to the beam shaper driver 46. Next, a predetermined deflection signal is applied to the beam size variable deflector 32 from the beam shaper driver 46 to the optical system 20, and thereby the size of the electron beam is controlled.

In the drawing data decoder 44, positioning data of the subfield is created based on the drawing stripe data, and this data is sent to the main deflector driver 47. Then, a predetermined signal is applied from the main deflector driver 47 to the main deflector 33 of the optical system 20, whereby the electron beam is deflected and scanned to a designated subfield position. Further, the drawing data decoder 44 generates a control signal for scanning the sub deflector, and this signal is sent to the sub deflector driver 48. Then, a predetermined sub-deflection signal is applied from the sub-deflector driver 48 to the sub-deflector 34, whereby drawing processing for each sub-field is performed.

Next, an electron beam drawing method using the above-configured apparatus will be described. FIG. 2 shows a process of generating data for performing a drawing process. LSI pattern is C
The design pattern data is designed and created by an AD system, and the design pattern data is converted into a drawing pattern by a host computer having a high processing capability such as a large computer. Then, the drawing pattern data is read out to perform the electron beam drawing. Here, the data created by the CAD system is usually constituted by a group of figures of a pattern polygon, and has a figure data system in which patterns are allowed to overlap each other. In order to make the LSI pattern data of such a format into a graphic data system that can be accepted by the electron beam drawing apparatus, the host computer performs the following data processing.

(1) A contouring process of a figure is performed to remove a multiple exposure area of a beam.

(2) As shown in FIG. 3A, the area of the LSI chip is divided into frame areas 53a to 53d, which are unit drawing areas which can be deflected by the beam, and a subfield area 54. FIG. 3B shows a drawing figure in a subfield area 54 which is made into polygons 51 and 52 by removing the beam multiple exposure area.

(3) For the drawing figure in the subfield area 54,
As shown in FIG. 3 (c), a graphic dividing process is performed into a basic graphic group 56 composed of rectangular and trapezoidal figures.

The graphic data obtained by such a data generation process is represented by a graphic shape flag, a graphic position, and a graphic size,
The graphic data is defined as a group of graphic data in units of a subfield area and a frame area, and stored in the magnetic disk 41.

When drawing a mask, a mask pattern as shown in FIG. 4 is drawn based on the drawing pattern data generated by the above-described data conversion processing. The drawing mask includes an LSI generated by the data conversion process.
In addition to the pattern portions 60a to 60c, alignment marks for a reduction projection exposure apparatus (an example of a mark shape is shown in FIG. 5A) and drawing around an LSI pattern portion (main pattern) as shown in 61 to 64 A mark for accuracy evaluation (an example of a mark shape is shown in FIG. 5 (b)) and the like are arranged. Further, an identification mark for visually identifying the type of the mask as shown at 65 is drawn and a desired mask is drawn. Created.

Next, a process of performing a drawing process on such a mask will be described. As shown in FIG. 2, the drawing pattern data 60a to 60c relating to the LSI chip pattern data converted by the host computer is stored on the magnetic disk 41 and stored in the control computer.
Put under 40 controls. Then, the control computer 40 selects and arranges a required mark pattern based on a mask name (pattern identification name) representing the type of the converted data, and performs drawing processing under necessary drawing conditions.

Specifically, the control computer 40 determines various drawing conditions with reference to a drawing information table as shown in FIG. 6 and provides the drawing conditions. In the drawing information table shown in FIG. 6, the LSI chip part obtained by converting the data is replaced with the name "MAIN" corresponding to the keyword part from which a part of the mask name is extracted, and drawn before drawing the MAIN pattern. Mark layout information to be drawn and mark layout information to be drawn after drawing the MAIN pattern are defined as mark information.

In this example, when the fifth to seventh digits of the mask name conform to “EFG”, the drawing processing of the mark layout “A” is performed before drawing the MAIN pattern, and the mark layout “B” is drawn after drawing the MAIN pattern. Is indicated by the first definition information in the table. The mark layouts “A” and “B” defined here are the seventh
Mark arrangement data of a system as shown in FIGS. 9A and 9B is stored in advance on the magnetic disk 41 of the control computer together with mark pattern data 61 to 64.

Further, the drawing information table includes a black-and-white inversion definition unit,
A mirror reversing definition unit and a system capable of defining drawing conditions such as scaling (magnification specification), beam irradiation amount at the time of drawing, necessity of proximity effect correction drawing, and the method (type of correction) as other specifications. I have.

Therefore, when a part of the mask name defined second in the above table matches "EDC", the MAIN pattern portion is rendered black and white inverted as shown in FIG. 8 (a). Further, in the case of the thirdly defined mask name "AB... E", the system is subjected to mirror inversion and drawing processing as shown in FIG. 8 (b). In addition, the scaling, the beam irradiation amount, and the proximity effect correction are similarly reflected in the drawing process.

Accordingly, the mask drawing process shown in FIG. 4 is performed by the following procedure.

(1) A mask ID pattern 65 corresponding to the mask name is drawn.

(2) Mark patterns 61 arranged at four corners of the mask
Is drawn.

(3) The data-processed LSI pattern units 60a to 60c are drawn.

(4) Mark pattern 6 arranged around the LSI pattern part
2 to 64 are drawn.

With the above processing steps, LS
With regard to patterns other than the I pattern part and drawing conditions, data can be created without awareness
This eliminates the need for manual work for selecting and arranging marks and setting the drawing conditions required for each mask type, thereby eliminating the occurrence of defective masks due to human error. As a result, it is possible to improve the throughput of the writing process, improve the operation rate of the apparatus, and increase the reliability of the LSI mask for writing. Furthermore, by referring to the ID pattern corresponding to the mask name, the processing conditions after the drawing of the sample can be clearly determined.

Further, as described above, the processing system is also very effective when directly drawing a pattern on a wafer. For example, in the case of wafer drawing, black-and-white inversion, dimensional correction (thickening / thinning of pattern), proximity effect correction, and the like are required depending on the processing process after drawing. In addition, if scaling, mirror inversion, and the like can be designated, there is an advantage that the allowable range of the device is widened and usability is improved. In addition, when manufacturing LSI using a mixture of reduced transfer of a mask and direct writing of a wafer, the pattern writing position is corrected when the pattern is drawn on the wafer in accordance with the pattern transfer distortion of the transfer device to accurately write the pattern. Superposition processing is required.

In the case of performing such processing, it is possible to eliminate defective drawing due to a human error by setting the drawing condition by the pattern identification name in the same manner as the mask drawing, and as a result, it is possible to improve the drawing throughput, the operation rate of the apparatus, and the like. The reliability of the drawing wafer can be improved.

Thus, according to the method of this embodiment, when data is converted,
It is possible to create data without being aware of the pattern and drawing conditions other than the LSI pattern part,
This eliminates the need for manual work for selecting and arranging marks and setting drawing conditions for each type of pattern, thereby eliminating the occurrence of defective masks or defective wafers due to human error. Also, the change or addition of the mark type, the arrangement position, or the drawing condition can be handled only by the mask or wafer drawing processing side, and the influence on the data conversion can be eliminated. As a result, the throughput of the drawing process is improved, the operation rate of the apparatus is improved, and the drawing is performed.
The reliability of the LSI mask can be improved.

The present invention is not limited to the embodiments described above. For example, the means for storing the drawing pattern data is not limited to a magnetic disk, but may be another storage medium such as a magnetic tape or a semiconductor memory. Further, the configuration of the electron beam lithography apparatus is not limited to what is shown in FIG. 1 and can be changed as appropriate. Although the embodiment has been described by taking an electron beam as an example, the invention is not limited to an electron beam but can be applied to an ion beam, a laser beam, and the like. Alternatively, a one-stage deflection system or a three-stage or more deflection system may be used. Further, in addition to a shot method using a variable shaped beam, an apparatus method using a circular beam is also applicable. Further, the drawing pattern data stored in the storage medium is not limited to the figure system of the basic figure, but can be applied to drawing unit figures and polygon figures. In addition, various modifications can be made without departing from the scope of the present invention.

[Effects of the Invention] As described above in detail, according to the present invention, mark information, drawing conditions, and the like are decoded from an identification name of a pattern to be drawn on a mask or a wafer, thereby forming a peripheral mark pattern and drawing required for drawing. Conditions and the like can be set automatically. Therefore, it is possible to eliminate the occurrence of a defective mask or a defective wafer which has conventionally been caused by manual operation, and to improve the reliability of the LSI pattern drawn by the charged beam drawing apparatus and the operation rate of the apparatus.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram showing an electron beam lithography apparatus used in the method of one embodiment of the present invention, FIG. 2 is a schematic diagram showing a generation process of lithography pattern data, and FIG. Schematic diagram showing pattern division and region division of
FIG. 4 is a schematic diagram showing an example of a mask layout to be drawn, FIG. 5 is a schematic diagram showing an example of a mark pattern, and FIG.
FIG. 7 is a schematic diagram showing an internal structure of the drawing information table, FIG. 7 is a schematic diagram showing an example of a mark layout, and FIG. 8 is a schematic diagram showing a drawing system when black-and-white inversion and mirror inversion are designated. 10 ... Sample chamber, 11 ... Sample, 12 ... Table, 20 ... Electronic optical system, 21 ... Electron gun, 22-26 ... Lens, 31-34 ... Deflector, 35,36 ... Beam Forming aperture, 40: Control computer, 41: Magnetic disk (storage medium), 42: Pattern memory, 43: Pattern data decoder, 44: Drawing data decoder, 45: Blanking circuit, 46-48 ... ... deflector driver, 51, 52 ... design pattern, 53a to 53d ... frame area, 54 ... subfield area, 56 ... basic figure, 60a to 60c ... LSI pattern part, 61 ... alignment mark, 62 ~ 64: Evaluation mark, 65: Identification mark.

──────────────────────────────────────────────────続 き Continuation of front page (56) References JP-A-3-104208 (JP, A) JP-A-2-5406 (JP, A) JP-A-1-264222 (JP, A) JP-A-61- 284920 (JP, A) JP-A-60-216547 (JP, A) JP-A-1-238016 (JP, A) JP-A-1-111327 (JP, A) JP-A-62-242526 (JP, A) JP-A-59-87817 (JP, A) JP-A-57-37833 (JP, A) JP-A-56-138925 (JP, A) JP-A-56-90521 (JP, A) (58) (Int.Cl. ⁶ , DB name) H01L 21/027

Claims (4)

(57) [Claims] In a charged beam writing method for writing an LSI pattern on a mask used for pattern transfer, a peripheral area to be located at a peripheral portion of a main pattern is obtained from a pattern identification name indicating a type of a pattern to be written on the mask. A charged beam writing method, comprising decoding mark information relating to a mark pattern, decoding the drawing conditions of the mark pattern and the main pattern, and drawing the main pattern and the mark pattern on the mask based on the decoding result. 2. The mark information includes a mark pattern type,
The decoding of the mark information is performed by defining a keyword portion from which a part of the pattern identifier is extracted in advance, the type of the mark pattern to be arranged for each keyword, the arrangement position, and the drawing order. 2. The charged beam drawing method according to claim 1, wherein the charged beam drawing method is performed by referring to a table including a mark definition unit. 3. The drawing condition includes a black and white reversal of a drawing pattern,
It shows the mirror inversion, the drawing magnification, the beam irradiation amount at the time of drawing, the proximity effect correction procedure, and the like. The decoding of the drawing conditions is performed by a keyword portion in which a part of the pattern identification name is extracted in advance and a keyword portion. 2. The method according to claim 1, wherein the drawing condition is performed by referring to a table in which the drawing condition is defined.
Charged beam drawing method as described. 4. The method according to claim 1, wherein said pattern identification name is simultaneously drawn on a peripheral portion of a main pattern when drawing an LSI pattern on said mask or wafer.
Charged beam drawing method as described.