JP3295855B2 - Charged particle beam exposure method - Google Patents

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 exposure method, and more particularly to a fine line pattern having a width of 0.1 .mu.m or less and having an arbitrary length, which makes it difficult to expose a clear pattern by a variable rectangular exposure method. The present invention relates to a charged particle beam exposure method for exposing a circuit pattern including:

[0002]

2. Description of the Related Art In recent years, with the increasing integration of integrated circuits, a new exposure method using a charged particle beam such as an electron beam has been studied instead of photolithography technology using light, which has been the mainstream of fine pattern formation for many years. It has come to be.

The exposure method using a charged particle beam such as an electron beam has a great feature in that a fine pattern on the order of microns or submicron or less can be formed.

According to the electron beam exposure method, an integrated circuit pattern is exposed on a surface of a resist film formed on a semiconductor wafer by a finely focused electron beam, and is exposed at a position of 0.02 μm or less. Exposure of a pattern having a width smaller than 0.05 μm with alignment accuracy is easily achieved.

On the other hand, in such an electron beam exposure method, an integrated circuit pattern is drawn by a so-called one-stroke technique, in which a focused electron beam is used to completely operate and fill the pattern. Compared to the conventional optical exposure method in which a large area substrate is exposed by light irradiation, there is a disadvantage that the processing capability (throughput) of the exposure process is limited.

In order to overcome such a problem, the inventors of the present invention first converted a focused electron beam into the shape of an element pattern constituting a part of a large and complicated pattern of a semiconductor integrated circuit device. A block exposure method using an electron beam for shaping has been proposed (for example, Japanese Patent Application Laid-Open No. 52-119185).
Reference).

In this block exposure method, an electron having a plurality of openings having a shape such as a rectangle, a triangle, a circle, or a cross, which is the shape of an element pattern of a semiconductor integrated circuit pattern, for shaping a passing electron beam. A beam shaping mask (block mask) is used to shape the electron beam.

By deflecting the electron beam and selectively projecting and passing the electron beam into one of the plurality of openings, the cross-sectional shape of the electron beam is shaped according to the shape of the passed opening, and the selected opening is selected. An object, for example, a photoresist film formed on a semiconductor substrate, can be exposed by passing electrons having a cross-sectional shape of. By continuously and repeatedly exposing the element patterns constituting the integrated circuit pattern, a large area and complex pattern of a desired semiconductor integrated circuit can be exposed.

Such a block exposure method is particularly effective when applied to an exposure step of an integrated circuit pattern which is mostly composed of repeated element patterns, such as a semiconductor memory (DRAM) pattern. 62-26
0322).

[0010] In order to perform the exposure more efficiently by this exposure method, it is necessary to repeat as many types as possible with a high repetition frequency.
For example, it is necessary to prepare an electron beam shaping mask having 20 to 40 kinds of aperture groups, and the present inventors have proposed a block exposure method effective when using an electron beam shaping mask having a large number of aperture shapes. Has been proposed.

Here, prior to the description of the present invention, a block exposure method and a block exposure apparatus which have been proposed by the inventors of the present invention and which form the basis of the present invention will be described.

FIG. 1 is an explanatory view of the configuration of a block exposure apparatus proposed earlier by the present inventors. FIG. 1 shows the configuration of an electron beam exposure apparatus capable of implementing the present invention. In this drawing, 10 is an exposure system, 11 is a cathode electrode, 1
2 is a grid electrode, 13 is an anode electrode, 14 is an electron gun, 15 is a shaping slit, 16 is a first electron lens, 1
7, 21, 22, 23, 24, 33, 34 are deflectors, 1
8 is a second electronic lens, 19 is a third electronic lens, 20
Is a block mask, 25 is a blanking deflector, 26 is a fourth electron lens, 27 is a blanking aperture plate, 27a
Is a blanking aperture, 28 is a refocus coil, 29
Is a fifth electron lens, 30 is a dynamic focusing coil,
31 is a dynamic stig coil, 32 is an objective lens, 35 is a stage, 36, 37, 38, and 39 are adjustment coils, 50 is a control system, 51 is a magnetic storage device, and 52 is a CP.
U and 53 are interface circuits, 54 is a data memory, 55 is a pattern generator, 56, 57, 61, 64,
65 and 69 are D / A converters, 58 is a mask moving mechanism, 5
9 is a clock generator, 60 is a blanking control circuit, 6
2 is a sequence control circuit, 63 is a deflection control circuit, 66 is a stage moving mechanism, 67 is a laser interferometer, and 68 is a stage position calibration circuit.

This block exposure apparatus comprises an exposure system 10 for performing exposure by an electron beam, and a control system 50 for controlling the exposure operation of the exposure system 10.

The exposure system 10 includes a cathode electrode 11 for emitting thermoelectrons, an anode electrode 13 for accelerating electrons to form an electron beam, and a grid for controlling the flow of electrons emitted from the cathode electrode 11 to the anode electrode 13. An electron gun 14 including the electrode 12 is included. In this way,
The electron beam emitted from the electron gun 14 travels as a divergent electron beam in the direction of the target, that is, the target W held on the movable stage 35.

The electron beam thus formed is
It is shaped into a desired cross-sectional shape through a shaping opening formed in the shaping slit 15, and is directed to travel in a direction coincident with an optical axis O extending from the cathode electrode 11 to an object W on which a pattern is formed. . Typically, the shaping aperture is rectangular or square, and the object W is a semiconductor substrate (wafer).

In order to adjust the electron beam to a desired position with respect to the optical axis O, adjusting coils 36, 37, 38 and 3
9 are provided along the optical axis O. The electron beam shaped by the shaping slit 15 in this manner is converged by the first electron lens 16 to a point P ₁ located on the optical axis O. The electron beam is applied to the position corresponding to this point P ₁ in accordance with the control signal HS _1.
Is provided.

Next, the electron beam is converted into a parallel electron beam by a second electron lens 18 having a focal point coincident with the point P _1, and a plurality of beams for shaping the electron beam of a block mask (electron beam shaping mask) 20. Through one of the openings. By passing through the selected opening of the block mask 20, an electron beam having a desired cross-sectional shape can be formed.

For the purpose of projecting the electron beam onto a specific opening of the block mask 20, a pair of deflectors 21 is provided above the block mask 20 in order to deflect the parallel electron beam once off the optical axis O. , 22 are provided,
Further, another pair of deflectors 23 and 24 are provided below the block mask 20 for returning the electron beam to the optical axis O. As described above, the deflectors 21 to 24 deflect the electron beam in response to the supplied deflection control signals PS _{1 to} PS ₄ , and deselect the selected block mask 20.
Projection into opening above.

As described above, the parallel electron beam returned to the optical axis O passes through the third electron lens 19 and converges to a point P ₂ on the optical axis O. This electron beam
, And converges on the object W by the objective lens 32. In this way, the light image having the shape of the opening above the selected block mask 20 is:
It is projected on the upper surface of the object W.

In the objective lens 32, various coils such as a dynamic convergence coil 30 and a dynamic stig coil 31 and a deflector are provided for performing convergence correction or stig correction. Further, deflectors 33 and 34 for moving the converged electron beam on the object W are provided.
The deflector 34 deflects the electron beam in a narrow range of about 100 μm × 100 μm.

Further, a refocus coil 28 is provided above the fifth electron lens 29 in order to perform additional convergence control of the electron beam. Then, in order to control the projection and blocking of the electron beam onto the object W, a blanking for passing the electron beam on the optical axis O is provided between the fourth electron lens 26 and the fifth electron lens 29. A blanking opening plate 27 having an opening 27a is arranged.

The blanking aperture 27a has a small diameter, and the blanking aperture plate 27 blocks the electron beam when the electron beam is deflected off the optical axis O. Therefore, the electron beam disappears from the surface of the object W. In order to project or block the electron beam in this way, a blanking deflector 25 is provided between the third electron lens 19 and the blanking aperture plate 27.
The electron beam is deflected so as to be off the optical axis O in response to the supplied blanking control signal SB.

Next, a control system 50 for controlling the exposure system 10 will be described. Referring again to FIG.
Includes a magnetic storage device 51 that stores various design data of a semiconductor device or an integrated circuit device formed on the object W, and a CPU 52 that controls the exposure system 10. The CPU 52 reads pattern information of the semiconductor device such as pattern position data indicating a position on the object W to which the pattern data is written, mask information indicating an arrangement of openings on the block mask 20, and the like.

The pattern information and the mask information read from the magnetic storage device 51 are transferred to the data memory 54 on the one hand, and to the sequence control circuit 62 via the interface circuit 53 on the other hand. Data memory 5
4 accumulates the pattern information and mask information, in response to the information in the pattern information and mask information, and transfers the pattern generator 55 for generating a deflection control data PD ₁ -PD _4.

Deflection control data PD₁~ PD_FourIs D / A
The signal is sent to the converter 57 where the analog deflection control signal PS
₁~ PS_FourIs converted to Open on the block mask 20
Mouth selection is accomplished as described above. putter
The generator 55 further includes an object W to be exposed.
Position data SD indicating the position_ThreeOccurs. This position data
TA SD_ThreeIs sent to the D / A converter 65, where the objective
Analog for driving the sub deflector 34 in the lens 32
Signal S _ThreeIs converted to

The pattern generator 55 further generates correction data HD indicating the difference between the selected pattern in the desired pattern and the selected pattern on the block mask 20,
This was fed to a D / A converter 56, where the correction data HD is converted into a control signal HS ₁ for driving the deflector 17. Electron beam control signal in response to HS ₁ is moved over the block mask 20, additional shaping is carried out by slightly offset from the selected aperture of the electron beam.

The variable rectangular shaping of the electron beam is performed by the deflector 17.
Deflector 17 is used to deflect the electron beam at high speed within a limited area, typically of a size of 500 μm × 500 μm, while the deflector 2
1 to 24 apply an electron beam, typically 5 mm × 5 mm
Used to deflect at a slow speed over a relatively wide range of

Typically, the deflector 17 is constituted by an electrostatic deflector, while the deflectors 21 to 24 are constituted by electromagnetic deflectors. In addition to this, the pattern generator 55
Is control data M for moving the block mask 20.
KD is generated, and the control data MKD is supplied to the mask moving mechanism 58. The mask moving mechanism 58 moves the block mask 20 substantially in the optical axis O in response to the control data MKD.
Move in a plane perpendicular to.

As described above, by moving the block mask 20 within a range where the selected specific opening can be irradiated by the deflection of the electron beam, all the openings of the block mask 20 can be irradiated with the electron beam. it can.

Further, the pattern generator 55 generates a control signal for driving the refocus coil 28, and transmits this control signal via the D / A converter 69 to the refocus coil 2.
8

Thus, the electron beam is deflected by the deflector 3
The optimum convergence of the electron beam on the surface of the object W is maintained even when the electron beam is deflected by the light sources 3 and 34. Further, the pattern generator 55 generates a timing signal T for instructing the execution of the exposure or the waiting time of the exposure, and this timing signal T is supplied to the clock control circuit or the clock generator 59 to instruct the interruption of the exposure. The blanking control data BC is generated.

The blanking control data BC is supplied to a blanking control circuit 60. The blanking control circuit 60 supplies the blanking control data BC via the D / A converter 61 which generates the blanking control data BC as described above. The ranking deflector 25 is driven. Clock generator 59 further generates a system clock that operates at a predetermined rate that determines the throughput of the exposure, which will be described in detail below.

The sequence control circuit 62 detects the timing information transmitted from the interface circuit 53,
The data memory 54 is controlled so as to instruct the start of the exposure process and to output the main deflection data MD supplied to the deflection control circuit 63 via the pattern generator 55. The deflection control circuit 63 generates main deflection control data SD ₂ in response to the supplied data MD, and this data SD ₂ is supplied to a D / A converter 64 where it is converted into a deflection control signal S _2. You.

The deflection control signal S ₂ is transmitted to the objective lens 32
Is driven in the main deflector 33. Further, the deflection control circuit 63 controls the stage position calibration circuit 68 in response to a signal by the sequence control circuit, the stage position calibration circuit 68, as previously described, generates a drive signal S ₃ D / A The deflector 34 is driven via the converter 65.

In cooperation with the deflection control circuit 63 and the stage position calibration circuit 68, the sequence control circuit 62 monitors the position of the stage 35 by the laser interferometer 67, and the stage moving mechanism 66 for moving the stage 35.
Drive. Thus, the exposure of the pattern selected by the block mask 20 is performed at any desired position on the target object W.

FIG. 2 is an explanatory view of the configuration of a conventional block mask. This figure shows a configuration of a block mask 20 for shaping an electron beam used in the apparatus of FIG. Referring to FIG. 2, the block mask 20 is provided with a number of pattern regions E _{1 to} E ₉ arranged in rows and columns and separated from each other by a pitch EL, and each region has a size of about 5 mm × 5 mm. ing. The size of one region can be set to a size that can deflect the electron beam by the deflectors 21 and 22, for example, a range of 1 to 5 mm □.

FIG. 3 is a diagram for explaining the structure of a conventional pattern area. This figure shows the configuration of a pattern area having a number of pattern blocks formed on the block mask of FIG. Each pattern area shown in FIG.
As shown in, arranged in rows and columns, and has a number of block areas B ₁ .about.B ₃₆ includes an aperture for shaping the electron beam formed apart by a pitch BL to each other. Each block area has a beam size irradiated at one time, and is 100 μm × 100 μm to 500 μm × 50.
It has a size of 0 μm.

FIG. 4 is an explanatory view of the structure of a conventional electron beam shaping aperture. This figure shows an example of an electron beam shaping opening formed in the block mask shown in FIG. In this figure, 20 is a block mask, and 20a to 20a
20g is an opening. This figure shows an example of the electron beam shaping openings formed in the block regions Ba to Bd included in the block regions B _{1 to} B ₃₆ described above. Has 20 g.

FIG. 5 is an explanatory view of a conventional electron beam shaping process. This figure shows a process of shaping an electron beam by an opening formed in a block mask. In this figure, reference numeral 20 denotes a block mask, 71 denotes an electron beam irradiation area, and 72a to 72d denote openings for block exposure.
This figure also shows other shapes of the block exposure openings such as the openings 72a to 72d.

By deflecting the electron beam using the deflectors 21 to 24, it is possible to move the hatched electron beam irradiation area 71 and irradiate one of the block areas with the electron beam. In the illustrated example, the electron beam is shaped according to the shape corresponding to the block exposure opening 72a.

On the other hand, when it is necessary to expose a pattern that is not included in the opening formed in the block mask 20, the electron beam irradiation area 71 has a large, substantially square variable rectangular exposure shown in FIG. To the opening 73a. By adjusting the positional relationship where the electron beam irradiation region 71 and the variable rectangular exposure opening 73a overlap, a shaped electron beam having a rectangular cross-sectional shape of a desired size and shape can be formed. This step is
It is conventionally known as a variable rectangular beam shaping or variable rectangular exposure method. This variable rectangular beam shaping is used less frequently for exposure, and is therefore used when exposing a pattern that is not provided as a shaping aperture in the block mask 20.

By adopting the above-described exposure method, the same rectangular pattern having a relatively low repetition frequency can be generated by using the variable rectangular exposure method while maintaining the same pattern. The use of the block exposure method makes it possible to expose a pattern with a high frequency of pattern repetition as a whole at a high speed.

[0043]

However, in the above-mentioned electron beam exposure method, when the variable rectangular exposure step is applied to a case where an elongated beam shape having a remarkably fine pattern width of less than 0.1 μm is formed. This causes a problem of reducing the throughput of the exposure process.

In order to form such an extremely fine beam, extremely high precision is required for the deflection of the electron beam.
A high-precision D / A converter 57 is required to excite .about.24. In addition, the D / A converter 57 with high precision
And the deflection data PD ₁ -P
The time required to convert the D ₄ corresponding to the analog signal PS ₁ ~PS ₄ is significantly longer.

Further, when the variable rectangular exposure process is applied to form an electron beam having such a fine and long shape, the first edge in the longitudinal direction of the beam is shaped by the opening 73a in FIG. It should be noted that the second edge opposite the first edge is shaped by a shaping aperture provided in the shaping slit 15 at a position separated along the optical axis O.

Considering with reference to FIG. 1, when the convergence operation of the electron lens deviates from the ideal state, the convergence state of the electron beam becomes abnormal, and the first edge of the electron beam having a fine and long shape is formed. It can be seen that the second edge deviates from the ideal convergence state. For this reason, there arises a problem that a fine rectangular pattern as designed cannot be obtained.

Furthermore, due to the fact that the block mask 20 blocks most of the electron beam at the various shaping openings 73a, carbon deposits tend to form on the main surface above the block mask 20 during the exposure process. .

That is, when a pattern of approximately 0.1 μm or less, for example, a pattern of 0.05 μm × 3 μm is generated, assuming that an electron beam having a side of 4 μm is used, the width of the charged particle beam is 1. It will be aligned to allow only 25% to pass.

Accordingly, with respect to the width, 98.75% of the electron beam is blocked by the variable rectangular exposure mask area. If such carbon deposition occurs asymmetrically on one side of the opening 73a for irradiating the electron beam, the deposit changes the distribution of the electric field in the vicinity of the opening having various shapes. Gives a great distortion.

Such a minute distortion of the beam shape is a characteristic of a device to be manufactured, especially when the pattern is an exposure pattern corresponding to an important region such as a gate region of a high-speed MOS transistor or a HEMT. Can have fatal and harmful effects.

In order to solve this drawback, the conditions of the electron lens are changed to form an electron beam having a small cross section. For example, a 0.05 μm-size electron beam is formed so that the overlap between the rectangular aperture and the charged particle beam becomes 50%. An electron beam may be formed. In this case, however, an electron beam having a large cross section cannot be generated, which hinders block exposure of a large aperture.

Accordingly, it is a general object of the present invention to provide a method for exposing a pattern on a new and useful object that eliminates the problems mentioned above. Another object of the present invention is to provide an exposure method capable of accurately and quickly exposing a circuit pattern including a narrow line pattern having a narrow width in which a clear pattern is difficult to be exposed by a variable rectangular exposure method. . Other objects and features of the present invention will become apparent from the detailed description of the specification and the accompanying drawings.

[0053]

According to the present invention, a circuit pattern is exposed on an object by a shaped charged particle beam based on design data representing a circuit pattern to be exposed constituted by a plurality of element patterns. In the charged particle beam exposure method, (b) extracting from the design data block exposure data including exposure data for an element pattern to be repeatedly exposed a plurality of times; Extracting variable rectangle exposure data including exposure data for a rectangular element pattern different from the element pattern from the design data; and (c) extracting the variable rectangle exposure data from the design data with a size less than At two intersecting edges of the aperture formed in the beam shaping mask, the charged particle beam and the aperture of the aperture are formed. -Extract fine line block exposure data including exposure data for a fine line element pattern having a size smaller than a predetermined limit size, which is difficult to expose with a variable rectangular shaped beam shaped by adjusting the burlap state. And (d) the charged particles showing a configuration of the beam shaping mask for shaping the charged particle beam based on the block exposure data, the variable rectangular exposure data, and the fine line block exposure data. Extracting mask data including information on the position, size and shape of an opening formed on the beam shaping mask for shaping the beam; and (e) producing a beam shaping mask based on the mask data When,
(F) converting the charged particle beam into the block exposure data;
The circuit pattern is exposed by selectively passing one of the block exposure opening, the variable rectangular exposure opening and the fine line block exposure opening based on the variable rectangular exposure data and the fine line block exposure data. Process and adopted.

[0054]

According to the present invention, a thin line pattern having a narrow width, in which a block exposure method is not applicable and a clear pattern is difficult to be exposed by a variable rectangular exposure method, is a portion where the repetition frequency of the same pattern is not high. Since the fine line block exposure data is created and the fine line block exposure method is applied, a circuit pattern including an exposure pattern having a width of, for example, 0.1 μm or less, particularly 0.05 μm or less, and an arbitrary length. However, it is possible to quickly and accurately perform exposure.

[0055]

Embodiments of the present invention will be described below with reference to the drawings. (First Embodiment) FIG. 6 is a flow chart of the circuit pattern exposure process of the first embodiment. This figure shows the first embodiment of the present invention.
2 shows an exposure step of a circuit pattern of a semiconductor device by the exposure apparatus of FIG. 1 according to the embodiment.

Referring to FIG. 6, in step 1,
By the CPU 52, the design data of the semiconductor device is
Is read from a storage device such as the magnetic storage device 51 of FIG. First, block exposure data is extracted from the design data. The block exposure data is extracted by converting a pattern having high repeatability into a format in which block exposure is performed. The data format at this time will be briefly described.

FIG. 7 is an explanatory view of the exposure data format after the blocks are extracted in the first embodiment. Exposure data after the block extraction is shown by a broken line in FIG. This figure shows an example of a process of extracting block exposure data for a block exposure step and variable rectangle exposure data for a variable rectangle exposure step from design data of a semiconductor device. Referring to this figure, the data indicated by the first broken line after the data extraction processing (I) (step 2) is the variable rectangle exposure data for the variable rectangle exposure step and the block exposure data for the block exposure step. It has a mixed format with data.

In this example, at least three fields such as a first field (a), a second field (b), and a third field (c) are used to describe each variable rectangular exposure data and block exposure data. Use

The variable rectangular exposure data and the block exposure data are distinguished from each other by the numerical value of the third field (c), as will be described later. In the variable rectangular exposure data, the first field (a) is used for a pattern code for specifying the shape of a pattern to be exposed in the variable rectangular exposure step.

In the example described here, the first field (a) has a numerical value of “0000” corresponding to a rectangular pattern having a variable size and extending in the X direction.
Further, the second field (b) is used to describe the X and Y coordinates of the origin of the pattern on the substrate, while the third field (c) is for specifying the size of each direction of the pattern. Used for

For example, looking at the second line in FIG. 7, the shape is designated by the code "0000" in the first field (a), and 2.5 .mu.m and Y in the X direction specified by the third field (c). Exposure of a pattern on a substrate having a size of 4.5 μm in the direction is specified, and starting from the origin of −100 μm for the X axis and 0 μm for the Y axis by the second field (b). Specified.

On the other hand, in the block exposure data,
The first field (a) has a numerical value indicating a block number, and is used to specify a block formed on the block mask 20. In the example described here, a numerical value of “0081” has been input.

The second field (b) of the block exposure data is used to specify the origin of the exposure pattern as in the case of the variable rectangular exposure data, and the third field (c) is not used.

As described above, by checking the numerical value of the third field (c), it is possible to identify whether the data is variable rectangular exposure data or block exposure data. In the block exposure data, since the irradiation of the electron beam is performed on the entire surface of the block area, it is unnecessary to specify the pattern size.

The data shown in FIG. 7 is obtained by dividing the circuit pattern to be exposed into a number of meshes each having a size corresponding to the block size such as 5 μm × 5 μm, and obtaining each of the meshes obtained by the above-described division into meshes. The frequency of use is obtained by counting up the frequency of use of a pattern, selecting a frequently used pattern from the pattern with the highest frequency of use, and extracting block exposure data.

The data extracting step (I) is executed in step 2 shown in FIG. 6 based on the design data shown in FIG. 7, and the variable rectangular exposure data and the block exposure data are 4, mask data specifying the size and coordinates of the opening formed on the block mask 20 and separated from each other is generated in step 5. Here, Step 3, Step 4 and Step 5 do not necessarily need to be performed sequentially in a specific order, but may be performed simultaneously.

FIG. 8 is an explanatory diagram of the mask data of the first embodiment. This figure shows an example of mask data generated in step 5 of FIG. 6 based on the design data of FIG. 7 obtained in step 2 of FIG.

Referring to FIG. 8, the mask data includes four fields (e) to (h), and the first field (e) is a block corresponding to the data of the first field (a) of the design data. The second field (f) is used to hold the number, and the coordinates of the origin of the block area on the mask such as the block area B _i (i = 1, 2,...) Shown in FIG. Used to hold

The third field (g) is used to hold data for specifying the size of the opening formed in the block mask, and the fourth field (h) is
pattern area _i is included E _i (i = 1, 2, see ... Figure 2) for holding data specifying. In the example described here, the aperture specified by code 0081 has a size of 5 μm × 5 μm, X = 0 μm, Y
It is formed in a block area having an origin of 0 μm.

Of course, the aforementioned value representing the size of the aperture represents the size of the electron beam projected on the surface of the object W after being reduced by the exposure apparatus of FIG.
It does not represent the actual dimensions of the openings formed on the block mask. In actual exposure operation, the address by the electron beam is performed in the normal one pattern region E _i. Thus, the same data is given to the field (h) for those block areas is formed in the same pattern region E _i.

After the variable rectangular exposure data, the block exposure data, and the mask data are extracted in Steps 3, 4, and 5, the variable rectangular exposure data is extracted in one of the X-axis direction and the Y-axis direction. In addition, as previously discussed in connection with the prior art, due to the instability of the electron beam, the predetermined variable exposure is predetermined to be smaller than a limit value at which it is difficult to expose the pattern by the variable rectangular exposure method. Step 6 of extracting data representing a thin line pattern having a width is performed. This critical size is typically around 0.1 μm.

In the example described herein, the predetermined pattern width corresponds to a critical device parameter, such as the gate length of the HEMT being manufactured, of 0.0.
It should be noted that it is set at 8 μm.
Of course, the predetermined pattern width is 0.08μ.
The value is not limited to m, and can be set to another value depending on critical device parameters.

When the variable rectangular exposure data for an extremely fine line pattern having a width of 0.08 μm like the data on the third and fourth rows in FIG. The rectangular exposure data is converted into variable length fine line block exposure data in the next step 9, and the original variable rectangular exposure data is removed. On the other hand, FIG.
In this embodiment shown in (1), the block exposure data remains unchanged.

Further, simultaneously with the extraction of the fine line block exposure data, the block mask data formed in step 5 has a slit having a width of 0.08 μm and an appropriate length set to, for example, 4.0 μm. Modified to incorporate the pattern.

As a result, the corrected block mask data is formed in step 11. The block mask data thus formed has the width of 0.08 μm and the length of 4.0 μm in step 12. It is used for producing a block mask having a configuration similar to the above-described block mask 20 except that a slit pattern having the same is formed. Step 12
It involves a conventional photolithography process applied to the material forming the mask, such as silicon.

Further, in the electron beam exposure apparatus shown in FIG.
The exposure is executed by controlling the electron beam according to the exposure data obtained in step 9.

Next, the step of converting the variable rectangular exposure data into the fine line block exposure data is performed by two steps of converting the variable rectangular exposure data into the fine line block exposure data on the third and fifth lines in FIG. An example will be described in more detail with reference to FIG.

FIG. 9 is an explanatory view of the thin line block exposure data extracting step of the first embodiment. This figure shows an example of extracting thin line block exposure data from variable rectangular exposure data that constitutes a basic part of the present invention. Referring to FIG. 9, the thin line block exposure data converted from the variable rectangular exposure data in the third row of FIG. 7 has a pattern code 0084 corresponding to the thin line block shaping in the first field (a). In the second field (b) of the same fine line block exposure data, there is coordinate data of the origin equal to the origin coordinates of the original variable rectangular exposure data.

The third field (c) contains pattern size data "0.08" for specifying the width of the thin line data.
Is missing for the original variable rectangular exposure data, except that as long as the mask created in step 12 with slits having a width of 0.08 μm is used for exposure, it is no longer needed for exposure and is therefore no longer needed Has pattern size data corresponding to the pattern size data. In other words, the third field (c) contains only the data of the length of the fine line pattern to be exposed.

In the fine line block exposure data converted on the third line in FIG. 7, a length of 1.8 μm is specified. On the other hand, when the fine line block exposure data is converted in the fifth row, it is noted that the length of the fine line exceeds the length of the fine line opening on the block mask set to 4.0 μm in this example. Should be.

In this case, the fine line block exposure data is
The first field having a length of 4.0 μm in the third field (c)
The thin line pattern is divided into two data portions each including a portion and a second portion having a length of 0.8 μm in the third field (c). Exposure of the fine line pattern includes exposure of the first portion and exposure of the second portion. It is performed by performing it successively.

In order to obtain a continuous pattern having a total size of 4.8 μm on the substrate by the two consecutive exposures or shots, the second portion described in the field (b) of the second portion data is used. The origin of the data is
It should be noted that the pattern located at X = 0 μm and Y = 200 μm is shifted by 4 μm in the X direction with respect to the origin. A similar data extraction step is performed for a fine line pattern having another line width such as 0.5 μm.

As a result of the data extraction step (II) in step 6 described above, mask data for a mask having one or more fine line openings for shaping an extra fine line pattern by a fine line pattern exposure step is obtained.

FIG. 10 is an explanatory diagram of the corrected mask data of the first embodiment. This figure shows an example of mask data obtained by modifying the mask data of FIG. As shown in FIG. 10, it should be noted that, in this figure, data for the pattern specified by the code 0084 is added to the data of FIG.

The pattern having the code 0084 is shown in FIG.
The length of 4.0 μ corresponding to the exposure data shown in FIG.
It has a width of 08 μm. The block mask is shown in FIG.
Are created in step 12 together with the openings corresponding to the pattern 0084 based on the data shown in FIG. By adjusting the overlap with the electron beam irradiation area and the block area in which the parameter 0084 is formed, it is possible to expose a very fine thin line pattern having a width of 0.08 μm and an arbitrary length shorter than 4.0 μm. it can.

FIG. 11 is an explanatory diagram showing an example of the configuration of a block mask according to the first embodiment. This figure corresponds to step 1 in FIG.
2 shows an example of the block mask manufactured in FIG. In this figure, 80 is a block mask, 81 is an electron beam irradiation area, 82a to 82d are block exposure openings, 83a is a variable rectangular exposure opening, and 84a and 84b are variable fine line block exposure openings.

The block mask 8 manufactured as described above and used in place of the block mask 20 of the apparatus shown in FIG.
11, the block mask 80 has the same configuration as the block mask 20 of FIG. 5, and the block mask 80 is formed by the block exposure openings 72a to 72 of FIG.
The variable rectangular exposure opening 83a corresponding to the variable rectangular exposure opening 73a is formed, having the block exposure openings 82a to 82d corresponding to d. On the other hand, the block mask 80 and the block mask 20 are different in that variable electron beam block exposure openings 84a, 84b,... Are formed.

Like the variable rectangular exposure opening 83a, the block exposure openings 82a to 82c are used similarly to the conventional mask 20, and the variable fine line block exposure openings 84a and 84a are used.
4b is used to shape a fine linear beam having a desired width and a fine width. As shown in FIG. 9, the shape of the fine line beam, that is, the length of the fine line is determined by the electron beam irradiation region 81 and the variable fine line block exposure opening 84.
It is controlled by adjusting the overlap of a or 84b.

Since the electron beam is shaped by the two opposing main edges, the shape of the thus obtained longitudinally extending fine line beam is stable, and the width of the electron beam is smaller than 0.1 μm. However, it is possible to project an electron beam pattern having a very accurate width onto the surface of an object. As described above, the electron beam exposure method of this embodiment provides a powerful means for exposing a fine critical pattern of a semiconductor device with a large throughput.

(Second Embodiment) The electron beam exposure method according to the second embodiment is an improvement of the electron beam exposure method of the first embodiment. FIG. 12 is a flowchart of the circuit pattern exposure step of the second embodiment. This figure shows a process for exposing a circuit pattern according to a second embodiment of the present invention using the apparatus of FIG. Next, a second embodiment of the present invention, which is an improvement of the first embodiment, will be described with reference to FIG. 12 showing a flowchart of a circuit pattern exposure step corresponding to FIG.

Referring to FIG. 12, this step includes steps 21 to 25 corresponding to steps 1 to 5 for extracting variable rectangular exposure data, block exposure data and block mask data of FIG. . The details of these steps 21 to 25 are the same as the corresponding steps in FIG.

In this embodiment, steps 23 and 2
The variable rectangle exposure data and the block exposure data extracted by step 4 are supplied to step 26, where the variable rectangle exposure data for the thin line having a width smaller than about 0.1 μm and a predetermined length is obtained from the variable rectangle exposure data. The fine line block exposure data is extracted by checking the contents of the data, especially the contents of the third field (c) of the variable rectangular exposure data, and the fine line block exposure data thus extracted is used for the block exposure. Appended to data. In this way, both the variable rectangular exposure data and the block exposure data are modified in steps 27 and 28.

Further, the contents of the mask data formed in step 25 are corrected based on the block exposure data newly extracted in step 29.
Based on the processing in step 29, the mask data corrected in step 30 is formed, and the corrected mask data is stored in the block mask 80 in step 31.
Is used to produce a block mask such as

Based on the variable rectangular exposure data and the block exposure data in step 32, exposure is performed using the block mask prepared in step 31.

FIG. 13 is an explanatory diagram of an example of a block mask according to the second embodiment. This figure shows an example of a block opening formed on the mask 80 as a result of the step 29 in FIG. In this figure, reference numeral 80 denotes a block mask, 81 denotes an electron beam irradiation area, and 82a to 82a.
d is a block exposure opening, 83a is a variable rectangular exposure opening, and 85a to 85f are fine line block exposure openings.

In addition to the conventional block exposure openings 82a to 82d and the variable rectangular path exposure opening 83a, the block mask has a width corresponding to a limit width which cannot be sufficiently shaped by the conventional variable rectangular exposure. .. Openings 85a-85f for exposure to fine line block exposure in the longitudinal direction having a width smaller than 1 μm
Are formed. Fine line block exposure openings 85a-8
5f has various lengths.

Further, these apertures can have several different widths depending on the block exposure data extracted in step 26. By selecting one of the plurality of openings, such as the opening 85b, for the electron beam irradiation area indicated by oblique lines, block exposure can be performed even in a very fine line pattern.

According to the electron beam exposure method of this embodiment,
A more accurate beam shape can be obtained, thereby not only a more accurate exposure pattern can be obtained, but also the time required for data processing can be shortened because the beam deflection margin is increased, and the throughput of the exposure process is improved. In other words, accurate control of beam deflection is not required to realize a desired pattern by the variable rectangular exposure beam exposure method. This advantage can be applied to the first embodiment.

On the other hand, this embodiment does not require data processing or a step of adjusting the length of the beam in order to set the length of the fine line exposure pattern. It is better than the example.

The steps of the first and second embodiments, in particular, step 1 of the first embodiment for correcting mask data
It should be noted that there is a limit to the number of openings that can be formed above block mask 80 at zero, or at step 29 of the second embodiment. Although an opening corresponding to the block exposure data extracted in the extraction step of step 2 or step 22 is already formed in the block mask, for example, the pattern area E ₁ on the block mask (see FIG. 2) Only 48 openings can be formed in a typical pattern area such as).

As described above, the number of beam shaping openings newly formed on the block mask is determined by the maximum value of the openings that can be formed on the block mask and the number of openings already formed on the block mask. Must be set smaller than the difference between On the other hand, the number of block patterns extracted in step 26 can be large. This is because any exposure pattern can be extracted as a block opening. This problem is actually the second
In the embodiment, it is serious when extracting the fine linear block exposure data and the corresponding mask data.

FIG. 14 is a flowchart showing the steps of manufacturing the block mask of the second embodiment. This figure is
13 illustrates a step of manufacturing a block mask used in the step of FIG. With reference to this figure, a method for solving the problem that the number of beam shaping apertures that can be formed on the block mask is insufficient will be described. In order to secure a necessary block opening on the block mask 80, each step in FIG. 14 is applied to the step 26 in FIG.

In the process shown in FIG. 14, exposure data for a pattern having a size that is difficult to expose accurately and clearly according to the variable rectangular exposure method is extracted from the variable rectangular exposure data in step 261. They are classified according to their shape, size, and frequency of use of the exposure pattern.

Next, for each pattern area E _i , the number of block patterns N _BLOCK already formed in the block mask 80 is counted up in step 262, and the maximum number N of block patterns that can be formed in one pattern area E _i is calculated. difference N _MAX -N _BLOCK from _MAX is calculated in step 263. Further, the difference N
Step 263 determines whether _MAX- N _BLOCK is greater than 0 or not.
Is identified.

If the result is NO, indicating that the pattern area E _i is completely occupied by the block patterns arranged in rows and columns, there is no space for additionally providing a block pattern. That is, the process ends. On the other hand, if the result is YES, the pattern with the smallest size is searched in step 264,
Subsequently, the pattern is converted into block data in step 265.

Further, the value of the parameter N _BLOCK is incremented by one in the next step 266, and
The data searched in is removed from the original variable rectangular exposure data in step 267. Further, in the next step 268, it is determined whether or not there is data to be processed. If the result is YES, the process starting from the step 263 is performed for all data included in the variable rectangular exposure data. Is repeatedly executed until the data of (1) is processed.

Further, based on the block exposure data thus corrected, the mask data is corrected by the process of step 29 in FIG. According to the process of FIG. 14, the exposure data for the fine line pattern is preferentially converted into block exposure data because the pattern size is small, and the mask manufactured by the process of FIG. 14 has a fine opening thereon. become. Thereby, accurate beam shaping and exposure is performed even if the exposure pattern has a very small size.

FIG. 15 is a flowchart showing another example of the process of manufacturing the block mask of the second embodiment. The flowchart of this figure shows another step of manufacturing the block mask used in the step of FIG. This figure shows another example in which variable rectangular exposure data is preferentially converted into block exposure data for the most frequently used pattern. To this end, the contents of step 264 are modified as indicated by step 264 '. The other steps are the same as the steps in FIG. 14, and further description is omitted.

FIGS. 16 and 17 are flow charts showing still another step of the process of manufacturing the block mask of the second embodiment. These flowcharts are shown in FIG.
13 shows still another step of the step of manufacturing the block mask used in the step (a). These figures show a process in which variable rectangular exposure data can be converted to block exposure data even when there is no space for a new electron beam shaping opening on the block mask 80.

In this step, steps 271-2
Step 73 is executed when the result of the identification in step 263 is NO. More specifically, if the identification result in step 263 is NO, the search for the least frequently used block exposure data and the corresponding block pattern already formed on the block mask 80 is performed in step 271. Done.

Next, the block exposure data thus found is reconverted into variable rectangular exposure data in step 272, and the block exposure data is simultaneously removed from the mask data. Further, the parameter N
_BLOCK is decremented by one in step 273.

The process shown in FIGS. 16 and 17 is characterized in that a space for a new block pattern is created at the expense of a block exposure pattern which is not frequently used.
Exposure performed using the block exposure pattern removed in this manner is performed by the variable rectangular exposure method in this example. Instead of reducing large block exposure patterns, exposure of very fine patterns can be achieved by a block exposure process.

In the above embodiments, the exposure method using an electron beam has been described, but it goes without saying that the present invention can be applied to exposure using other charged particle beams. In addition, the present invention is not limited to the above-described embodiments, and various modifications and modifications can be made without departing from the scope of the present invention.

[0114]

According to the present invention, a pattern which cannot be applied to the block exposure method due to a low repetition frequency of the same pattern and which cannot be clearly exposed by the variable rectangular exposure method can be accurately exposed. In addition, a specific line width of 0.1 μm or less for performing variable-length fine line block exposure, for example, 0.08 μm
Although it is necessary to determine m and 0.04 μm in advance, the length of the pattern is not restricted, and when performing fine line block exposure, the length of the fine line cannot be changed continuously. Since the desired pattern can be formed by combining thin linear element patterns having different lengths, the burden when designing a semiconductor device is reduced, and the ultra-high capacity of a large-capacity memory, etc. It greatly contributes to the exposure of a fine pattern and, consequently, its manufacturing technology.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a configuration of a block exposure apparatus previously proposed by the present inventors.

FIG. 2 is an explanatory diagram of a configuration of a conventional block mask.

FIG. 3 is a diagram illustrating the configuration of a conventional pattern area.

FIG. 4 is a diagram illustrating the configuration of a conventional electron beam shaping aperture.

FIG. 5 is an explanatory view of a conventional electron beam shaping process.

FIG. 6 is a flowchart of a circuit pattern exposure step of the first embodiment.

FIG. 7 is an explanatory diagram of an exposure data format after block extraction in the first embodiment.

FIG. 8 is an explanatory diagram of mask data according to the first embodiment.

FIG. 9 is an explanatory diagram of a thin line block exposure data extraction step of the first embodiment.

FIG. 10 is an explanatory diagram of corrected mask data according to the first embodiment.

FIG. 11 is a diagram illustrating an example of a block mask according to the first embodiment;

FIG. 12 is a flowchart of a circuit pattern exposure process according to a second embodiment.

FIG. 13 is a diagram illustrating an example of a block mask according to a second embodiment;

FIG. 14 is a flowchart illustrating a process of manufacturing a block mask according to a second embodiment.

FIG. 15 is a flowchart illustrating another example of the process of manufacturing the block mask of the second embodiment.

FIG. 16 is a flowchart (1) showing still another step of the step of manufacturing the block mask of the second embodiment.

FIG. 17 is a flowchart (2) showing still another step of the step of manufacturing the block mask of the second embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Exposure system 11 Cathode electrode 12 Grid electrode 13 Anode electrode 14 Electron gun 15 Shaping slit 16 First electronic lens 17, 21, 22, 24, 33, 34 Deflector 18 Second electronic lens 19 Third electronic lens 20 Block mask 25 Blanking deflector 26 Fourth electron lens 27 Blanking aperture plate 27a Blanking aperture 28 Refocus coil 29 Fifth electron lens 30 Dynamic focusing coil 31 Dynamic stig coil 32 Objective lens 35 Stage 36, 37, 38 , 39 adjusting coil 50 control system 51 magnetic storage device 52 CPU 53 interface circuit 54 data memory 55 pattern generator 56, 57, 61, 64, 65, 69 D / A converter 58 mask moving mechanism 59 clock generator 60 Ranking control circuit 62 Sequence control circuit 63 Deflection control circuit 66 Stage moving mechanism 67 Laser interferometer 68 Stage position calibration circuit 20 Block mask 20a to 20g Opening 71 Electron beam irradiation area 72a to 72d Block exposure opening 80 Block mask 81 Electron beam irradiation Areas 82a to 82d Block exposure openings 83a Variable rectangular exposure openings 84a and 84b Variable fine line block exposure openings 85a to 85f Fine line block exposure openings

──────────────────────────────────────────────────続 き Continued on front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H01L 21/027 G03F 1/16 G03F 7/20 504

Claims (13)

(57) [Claims] (A) extracting, from design data, block exposure data including exposure data for an element pattern repeatedly exposed a plurality of times; and (b) extracting the element pattern from the design data. Extracting variable rectangle exposure data including exposure data for a rectangular element pattern different from the above; and (c) extracting a charged particle beam from the variable rectangle exposure data to a beam shaping mask for a smaller size. At two intersecting edges of the formed aperture, a variable rectangular shaped beam shaped by adjusting the overlap of the charged particle beam and the aperture is more difficult to expose than a predetermined limit size. Extracting thin line block exposure data including exposure data for a thin line element pattern having a small size; (d) Based on the block exposure data, the variable rectangular exposure data, and the fine line block exposure data, for shaping the charged particle beam indicating the configuration of the beam shaping mask for shaping the charged particle beam. (E) extracting mask data including information on the position, size, and shape of the opening formed on the beam shaping mask; (e) producing a beam shaping mask based on the mask data; Charged particle beam, the block exposure data,
The circuit pattern is exposed by selectively passing one of the block exposure opening, the variable rectangular exposure opening and the fine line block exposure opening based on the variable rectangular exposure data and the fine line block exposure data. Charged particle beam exposure, wherein the circuit pattern is exposed on the object by a shaped charged particle beam based on design data representing a circuit pattern to be exposed constituted by a plurality of element patterns having Method. 2. The charged particle beam according to claim 1, wherein the step of extracting the fine line block exposure data further includes the step of removing the fine line block exposure data from the variable rectangular exposure data. Exposure method. 3. The method according to claim 1, wherein extracting the variable-length fine line block exposure data includes extracting exposure data for a fine line pattern having a size smaller than approximately 0.1 μm. Charged particle beam exposure method. 4. The charged particle beam according to claim 3, wherein the fine line block exposure data includes exposure data for a plurality of fine line patterns each having a size smaller than approximately 0.1 μm. Exposure method. 5. The method according to claim 1, wherein the step of extracting the mask data comprises the steps of:
extracting mask data defining an opening extending in a longitudinal direction defined by a pair of main edges opposed to each other at a distance shorter than μm, and exposing the circuit pattern to thin line block exposure data. 4. A charged particle according to claim 3, comprising controlling to shape the longitudinally extending charged particle beam by at least the main edges of the two opposing longitudinally extending openings based on Beam exposure method. 6. The method of claim 5, wherein the step of controlling the charged particle beam is performed such that the longitudinally extending charged particle beam has various lengths determined by the fine line block exposure data. A charged particle beam exposure method described in 1. 7. The method according to claim 1, wherein the step of extracting fine line block exposure data includes the step of extracting fine line block exposure data for a fine line element pattern that is repeatedly exposed a plurality of times. Charged particle beam exposure method. 8. The step of extracting thin line block exposure data, the maximum number of openings that can be formed on the charged particle beam shaping mask, and the openings formed corresponding to the block exposure data and the variable rectangular exposure data. 8. The charged particle beam exposure method according to claim 7, wherein thin line block exposure data for a plurality of different thin line element patterns smaller than the difference in the number is extracted. 9. The step of extracting fine line block exposure data for a plurality of different fine line element patterns starts with the fine line element pattern having the minimum size, and continuously executes the fine line elements with the pattern size increasing. 9. The charged particle beam exposure method according to claim 8, wherein the method is performed. 10. The step of extracting variable-length fine line block exposure data for a plurality of different fine line element patterns,
10. The charged particle beam exposure method according to claim 9, wherein the method is started from a fine line element pattern having the maximum frequency of use and is continuously performed on a fine element pattern with a low frequency of use. 11. The step of extracting thin line block exposure data includes the steps of: determining the maximum number of openings that can be formed on the beam shaping mask; and the number of openings formed corresponding to the block exposure data and the variable rectangular exposure data. When the difference from the block exposure data is zero, the block exposure data extracted in the step of extracting the block exposure data so as to secure an area for the opening corresponding to the fine line block exposure data on the beam shaping mask. The charged particle beam exposure method according to claim 8, further comprising a step of removing data. 12. The method according to claim 11, wherein the step of removing the block exposure data is started from an element pattern having a minimum use frequency and is continuously executed for an element pattern having a higher use frequency. A charged particle beam exposure method described in 1. 13. A variable rectangular shaped beam formed by shaping a charged particle beam at two orthogonal edges of an aperture formed in a beam shaping mask to achieve exposure of a longitudinally extending aperture. Producing a beam shaping mask having a longitudinally extending aperture of a width smaller than a predetermined critical size, which is difficult to perform; and (b) a charged particle beam is shaped by two opposing edges defining said aperture. Directing the charged particle beam through the aperture so as to expose a pattern on the object with the shaped charged particle beam.