JPH07312338A - Method and apparatus for charged-particle-beam exposure - Google Patents

Abstract

PURPOSE:To offset the influence on the charged particle beam of a transient waveform contained in a driving signal and to enhance the throughput of a charged-particle-beam exposure by a method wherein the driving signal which is supplied to a main deflector or a subdeflector contains a predetermined correction signal. CONSTITUTION:A charged particle beam EB which is incident on an object 21 to be exposed, e.g. a semiconductor wafer or a mask, is scanned on the object 21 to be exposed. For this, the charged particle beam EB is deflected by a main deflector 25 and by a subdeflector 26 whose scanning range is narrower than that of the main deflector 25. Then, the object 21 to be exposed is exposed to the charged particle beam EB. In this exposure method, a driving signal is supplied to the main deflector 25 or the subdeflector 26 on the basis of the output of a D/A converter 33 or 43. In order to offset the influence on the charged particle beam EB of a transient waveform contained in the driving signal at a time when the output of the D/A converter 33 or 43 is changed, a predetermined correction signal is contained in the driving signal which is supplied to the main deflector 25 or the subdeflector 26.

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 apparatus.

[0002]

2. Description of the Related Art With the miniaturization of semiconductor integrated circuit devices, electron beam exposure apparatuses have been used. According to this apparatus, fine processing of 0.05 μm or less can be performed with alignment accuracy of 0.02 μm or less. However, since the exposure is performed by scanning the electron beam on the semiconductor wafer, the exposure time becomes longer than the light exposure. Therefore, it is necessary to improve the throughput of electron beam exposure.

One of the causes of the decrease in throughput is a settling waiting time for waiting for the vibration of the drive signal to settle when the drive signal supplied to the deflector is changed stepwise. For example, when digital deflection data is analogized by a D / A converter, amplified by an amplifier and supplied to a deflector, a settling wait time is required due to a transient waveform included in the output of the D / A converter. Become. This transient waveform is due to the D / A converter being configured, for example, as shown in FIG. In this D / A converter, the output terminals of the constant current sources 10 to 13, both input terminals of which are connected to the power supply line VCC, have operational amplifiers 1 through analog switches 14 to 17, respectively.
8 is connected to the input terminal of the operational amplifier 18, and the resistor 19 is connected between the input and output terminals of the operational amplifier 18. The constant current sources 10 to 13 output constant currents I, 2I, 4I and 8I, respectively.

For example, when all the analog switches 14 to 17 are turned off at the same time, the analog switches 14 to 17 are turned on at the same time due to a difference in operating speed of the analog switches 14 to 17. The output of the operational amplifier 18 is shown in FIG.
2 (B) includes a transient waveform (glitch waveform). Also, regarding the amplifier as the driver for the deflector, when the input changes stepwise, a transient waveform is included in the output due to the frequency characteristic of the amplifier. During such a transient waveform settling waiting time, exposure cannot be performed, which causes a decrease in the electron beam exposure throughput.

[0005]

SUMMARY OF THE INVENTION In view of such problems, an object of the present invention is to provide a charged particle beam exposure method and apparatus capable of improving the throughput of charged particle beam exposure.

[0006]

The charged particle beam exposure method and apparatus according to the present invention will be described with reference to the reference numerals of the corresponding components in the drawings. In the first invention, for example, in FIG. 1, a charged particle beam EB incident on an exposure object 21, for example, a semiconductor wafer or a mask.
The main deflector 25 in order to scan the object 21 to be exposed.
Further, in the charged particle beam exposure method of deflecting the charged particle beam EB by the sub deflector 26 having a narrower scanning range than the main deflector 25 and exposing the exposure object 21 with the charged particle beam EB, the D / A converter 33 or A drive signal is supplied to the main deflector 25 or the sub-deflector 26 based on the output of 43, and D /
When the output of the A converter 33 or 43 changes, a predetermined drive signal is supplied to the main deflector 25 or the sub deflector 26 in order to cancel the influence of the transient waveform included in the drive signal on the charged particle beam EB. Included correction signal.

According to the first aspect of the invention, when the output of the D / A converter changes, the influence of the transient waveform included in the drive signal on the charged particle beam is canceled by the correction signal, so that the transient waveform settling waits. The time can be eliminated or shortened to improve the throughput of charged particle beam exposure. Since the correction signal is predetermined, this cancellation is effectively performed.

In the first aspect of the first invention, the correction signal is D
The signal is a signal predetermined according to the change of the input of the / A converter. Since the transient waveform is determined according to the change of the input of the D / A converter, the cancellation is more effectively performed according to the first aspect. In the second aspect of the first invention, the correction signal is generated, for example, by sequentially reading the correction data stored in the storage means 41 in FIG.

According to the second aspect, it is possible to perform the above cancellation more completely by merely rewriting the correction data stored in the storage means 41, and the correction waveform is independent of changes in ambient temperature and aging. Is. For example, D /
The drive signal is sampled in accordance with the change of the input of the A converter 33, and the data is multiplied by a constant to obtain the storage means 41.
, The gain of the amplifier circuit 44 is adjusted, and the time axis of the data stored in the storage means 41 is shifted so that the cancellation can be performed more completely.

In the third aspect of the first invention, for example, in FIG. 1, each time the input value of the D / A converter 33 changes, a value corresponding to the input value is loaded into the counter 40 as an initial value and the counter 40 is loaded. The clock CLK2 is supplied to 40 to change the count value of the counter 40, and the storage means 41 is addressed by the count value. According to this third aspect, for transient waveforms that differ according to changes in the input value of the D / A converter 33,
It is possible to cancel the above.

In the fourth aspect of the first invention, for example, in FIGS. 5 and 6, the charged particle beam EB is continuously deflected in the predetermined direction B by the main deflector 25, and the charged particle beam EB is opposite to the predetermined direction B. A correction signal is included in the drive signal supplied to the main deflector 25 or the sub-deflector 26 when swinging back to the side.
In the fourth aspect, since the charged particle beam is continuously deflected in the predetermined direction B by the main deflector, the correction signal may be included in the drive signal only when the charged particle beam is swung back in the direction opposite to the predetermined direction B. .

In the fifth aspect of the first invention, for example, in FIGS. 5 and 6, in addition to the fourth aspect, the exposure object 21
Is mounted on the moving stage 22, the moving stage 22 is continuously moved in a direction C perpendicular to the deflection direction of the main deflector 25, and the main deflector 25 charges the sub-deflector 26 for each scanning. The particle beam EB is deflected following the movement of the moving stage 22.

According to the fifth aspect, since a problem such as a connection shift between subfields as shown in FIG. 10 does not occur, it becomes possible to accurately expose a fine pattern. Further, since the scanning width of the charged particle beam by the sub-deflector is relatively narrow, for example, 10 μm, the irradiation position accuracy of the charged particle beam EB on the exposure object 21 in the sub-deflection direction is high, and the number of bits is small and high speed. And high precision D /
Since the A converter 43 can be used, high speed and high precision scanning of the charged particle beam EB by the sub deflector 26 is possible.

In the sixth aspect of the first invention, for example, in FIG. 5, the data in the second storage means 31 is read by the count value of the second counter 30 in the fifth aspect, and the data is D / A. The signal is supplied to the converter 33 to be converted into an analog signal, the drive signal is supplied to the main deflector 25 via the low-pass filter 36 and the amplifier circuit 34, and the clock CLK1 is supplied to the clock input terminal of the second counter 30 to output the drive signal. , And a sawtooth wave whose cycle is equal to the cycle of the count value and the frequency of the clock CLK1 is changed to make the scanning speed of the main deflector 25 variable.

According to the sixth aspect, the clock CLK2
The scanning speed can be adjusted more accurately and more easily than the analog sawtooth wave generator circuit using the time constant circuit, and the change of the scanning speed due to the ambient temperature and aging change like the analog sawtooth wave generator circuit. Therefore, even if the scanning speed is changed according to the sensitivity of the resist, it becomes possible to expose a fine pattern that requires high accuracy.

In the seventh aspect of the first invention, for example, FIGS.
In addition to the sixth aspect, the blanking aperture array 28 in which a plurality of openings 60 are formed in a zigzag pattern in the substrate and a pair of electrodes 61 and 62 are formed at the edge of each opening 60 of the substrate is shown in FIG. The charged particle beam EB is supplied to the main deflector 25.
And a position before passing through the sub-deflector 26, the charged particle beam EB is passed through the blanking aperture array 28 to form a multi-beam, and a voltage is applied between the pair of electrodes 61 and 62. The exposure target 21 is exposed with a pattern according to the voltage applied to the electrodes 61 and 62 by controlling whether the exposure target 21 is irradiated with the charged particle beam EB passing through the opening 60.

According to the seventh aspect, it becomes possible to expose a finer and arbitrary pattern. In the second invention, for example, as shown in FIG. 1, in order to scan the charged particle beam EB incident on the exposure target 21 on the exposure target 21, the main deflector 25 and the scanning range of the main deflector 25 are smaller than the main deflector 25. The narrow sub-deflector 26 deflects the charged particle beam EB,
In the charged particle beam exposure apparatus for exposing the exposure object 21 with the charged particle beam EB, the first D / A converter 33
, A first amplifier circuit 34 coupled between the output terminal of the first D / A converter 33 and the input terminal of the main deflector 25, and a second amplifier circuit 44 coupled to the input terminal of the sub-deflector 26. And 1D
In order to cancel the influence of the transient waveform included in the output of the first amplifier on the charged particle beam EB when the input of the A / A converter 33 changes, the output signal of the first amplifier or the second amplifier is previously Waveform correction circuits 40, 41, 50, and 51 that include a determined correction signal.

According to the first aspect of the invention, when the output of the first D / A converter changes, the influence of the transient waveform contained in the drive signal on the charged particle beam is canceled by the correction signal, so that the transient waveform is settled. Latency can be eliminated or reduced to improve charged particle beam exposure throughput. Since the correction signal is predetermined, this cancellation is effectively performed.

In the first aspect of the second aspect of the invention, for example, as shown in FIG. 1, when the input value of the first D / A converter 33 changes, the waveform correction circuit sets an initial value corresponding to the input value. ,
A counter 40 that counts the clock CLK2, a second D / A converter 43 that is coupled to a front stage of the second amplifier circuit 44,
The count value of the counter 40 is supplied to the address input terminal, the correction waveform data is stored, and the read data is stored in the second D
Storage means 41 included in the input of the / A converter 43.

According to this first aspect, the D / A converter 33
It is possible to cancel the transient waveform that differs depending on the change of the input value of. Further, the above cancellation can be performed more completely by simply rewriting the correction data stored in the storage means 41, and the correction waveform is independent of changes in ambient temperature and aging. In the second aspect of the second invention, for example, as shown in FIG. 5, the waveform correction circuit sets the initial value when the input value of the first D / A converter 33 changes from one of the maximum value and the minimum value to the other. The counter 40 for counting the clock CLK2 and the second amplifier circuit 4
The second D / A converter 43 connected to the preceding stage of 4 and the count value of the counter 40 are supplied to the address input terminal, the correction waveform data is stored, and the read data is stored in the second D / A converter 43. And storage means 41 included in the input.

According to the second aspect, the method of the fifth aspect of the first invention can be realized, and the effect thereof can be obtained. In the third aspect of the second invention, for example, as shown in FIG. 5, the second counter 30 and the data for driving the main deflector which are addressed by the count value of the second counter 30 are stored and read. Second storage means 31, which supplies the obtained data to the first D / A converter 33, the first D / A converter 33, and the first amplifier circuit 3.
4 and a low pass filter 36 coupled between

According to the third aspect, the method of the sixth aspect of the first invention can be realized, and its effect can be obtained. In the third invention, in order to scan the charged particle beam incident on the exposure target object on the exposure target object, the exposure target object is mounted on a moving stage, and the main deflector and the main deflector are used for scanning. In a charged particle beam exposure method of deflecting the charged particle beam with a sub-deflector having a narrow range and exposing the object to be exposed with the charged particle beam, the charged particle beam is continuously directed in a predetermined direction by the main deflector. Deflection is performed and the moving stage is continuously moved in a direction perpendicular to the predetermined direction, and an error in the detected position of the moving stage with respect to the target position is corrected by the sub deflector.

According to the third aspect of the present invention, since the problem such as the connection shift between the subfields as shown in FIG. 10 does not occur, it is possible to accurately expose the fine pattern. Therefore, it is suitable for exposure of a fine pattern using a charged particle beam, particularly exposure of a fine pattern using a blanking aperture array as in the seventh aspect of the first invention.

[0024]

Embodiments of the present invention will be described below with reference to the drawings. In different drawings, the same or similar components corresponding to each other are denoted by the same or similar reference numerals. [First Embodiment] FIG. 1 shows a main structure of a charged particle beam exposure apparatus according to the first embodiment. A semiconductor wafer 21 as an exposure object is placed on a moving stage 22.
The movement of the moving stage 22 is controlled by the stage control circuit 23, and the position of the moving stage 22 is determined by the stage position detector 2.
4, detected by, for example, a laser interferometer, and the detected position is fed back to the stage control circuit 23.

A resist film is deposited on the semiconductor wafer 21 and is exposed by irradiating it with an electron beam EB as a charged particle beam. The scanning of the electron beam EB on the semiconductor wafer 21 is performed by moving the moving stage 22 and above the moving stage 22.
Main deflector 25 and sub-deflector 26 for the electron beam EB
And by. Usually, the main deflector 25 is an electromagnetic type and the sub deflector 26 is an electrostatic type.

The movable stage 22, the main deflector 25, and the sub-deflector 26 are in the descending order of the scannable range with the required accuracy.
However, the sub-deflector 26, the main deflector 25, and the moving stage 22 are arranged in order of increasing scanning speed. The electron beam E on the semiconductor wafer 21 is utilized by utilizing such a scanning property.
The scanning of B is performed as shown in FIG. 2, for example. That is, the electron beam EB is scanned in the subfield F in the direction A by the sub-deflector 26, and the one subfield F is scanned.
Each time the scanning inside is completed, the main deflector 25 performs step scanning in the longitudinal direction B of the stripe ST (direction perpendicular to the direction A) by the width of the subfield F. Further, the movable stage 22 continuously scans in the direction C perpendicular to the direction B. For example, the length of the stripe ST is 2 mm, and one side of the subfield F is 100 μm. Such scanning method itself is well known to those skilled in the art.

There are various kinds of exposure patterns in the area D1 indicated by diagonal lines, for example, a rectangular pattern which is exposed by one shot of the electron beam EB, a stencil which is arranged above the main deflector 25 and is well known to those skilled in the art. It is a fine pattern formed by passing the electron beam EB through the opening pattern of the mask, or an arbitrary fine pattern obtained by using the blanking aperture array 28 described later shown in FIG.

In FIG. 1, the clock CLK1 is supplied to the clock input terminal CK of the n-ary counter 30 to be counted, the main deflector scanning memory 31 is addressed by the count value i, and read from the main deflector scanning memory 31. The output data g (i) of the deflection amount of the main deflector is converted into an analog signal by the D / A converter 33, and then amplified by the amplifier circuit 34.
The drive current XA as shown in (4) is supplied to the main deflector 25.

The drive current XA contains a transient waveform. As the transient waveform, when the output of the D / A converter 33 changes, the transient waveform according to the value before and after the change is amplified by the amplifier circuit 3.
4 and the frequency characteristics of the amplifier circuit 34. This transient waveform is determined by the values before and after the change of the input value of the D / A converter 33, and the D / A converter 3
The values before and after the change of the input value of 3 are determined by the count value i of the n-ary counter 30. The count value of the n-ary counter 30 is i-
Hereinafter, the transient waveform when changing from 1 to the count value i (however, when i = n, changes from n to 0) is referred to as an i-th transient waveform. When the count value i changes from n to 0 in FIG.
It corresponds to returning from the end point of the scan in the middle direction B to the start point.

On the other hand, the corrected waveform data f (i, j) is read from the waveform correction memory 41 by addressing with the data g (i) as the upper side and the count value j of the m-ary counter 40 as the lower side. The adder 42 adds the corrected waveform data and the deflection signal Xs corresponding to the original deflection amount of the sub deflector 26, and supplies the result as the deflection signal Xsa to the D / A converter 43. The output of the D / A converter 43 is amplified by the amplifier circuit 44, and a pair of sub-deflections in which a drive voltage proportional to the output of the D / A converter 43 and a drive voltage having an opposite polarity and an equal absolute value are opposed to each other. Applied to the device 26.

The deflection signal Xs is after the coordinate conversion described later in the third embodiment, and the deflection signal Xs contains the target position of the moving stage 22.
A signal for correcting the position error (stage feedback amount) detected by the stage position detector 24 by the sub deflector 26 is also included. The clock CLK1 is supplied to the set input terminal S of the RS flip-flop 50, and when the clock CLK1 rises, the RS flip-flop 50 is set, its output terminal Q becomes high level, and the AND gate 51 is opened. As a result, the clock CLK2
Is supplied to the clock input terminal CK of the m-ary counter 40 through the AND gate 51, and the waveform correction memory 41 outputs the data f (i, i) of the i-th corrected waveform corresponding to the i-th transient waveform.
j) and j = 0 to m are output. The frequency of the clock CLK2 is at least m times the frequency of the clock CLK1.

When the count value j of the m-ary counter 40 changes from m to 0 and the carry changes from 0 to 1, the RS flip-flop 50 is reset, the output terminal Q thereof becomes low level, and the AND gate 51. Is closed. As a result, the supply of the i-th correction waveform f to the adder 42 is stopped. In this way, the deflection signal Xsa includes the i-th correction waveform data f (i, j), j = 0 to m corresponding to the i-th transient waveform included in the drive current XA, and the drive current XA and The correction waveform f is as shown in FIG. Then, the deflection of the electron beam EB by the main deflector 25 based on the transient waveform is
It is canceled by the deflection by the sub-deflector 26 based on the correction waveform f. Therefore, the settling waiting time based on the transient waveform is omitted, and the throughput of exposure by the electron beam EB is improved.

[Second Embodiment] FIG. 4 shows the essential structure of a charged particle beam exposure apparatus according to the second embodiment. The device of FIG.
While the correction of the transient waveform included in the output of the D / A converter 33 is executed by the sub-deflector 26, the correction of the main deflector 25 is executed by the apparatus of FIG.

That is, the output of the waveform correction memory 41 is set to D
After being analogized by the A / A converter 43, this is added to the adder 35.
At the same time, it is added to the output of the D / A converter 33 and is supplied to the amplifier circuit 34. The analog signal Xsb corresponding to the signal Xs in FIG. 1 is supplied to the amplifier circuit 44. Also, the clock C
The frequency of LK2 is set to m times the frequency of the clock CLK1 or less, that is, the count value of the counter 40 is set to be m or less in one cycle of the clock CLK1, and the count value i of the counter 40 at the rising timing of the clock CLK1. Is cleared to zero, the configuration is simplified without using the RS flip-flop 50 and the AND gate 51 of FIG.

According to the second embodiment, even if there is a deviation in the coordinate system between the main deflector 25 and the sub-deflector 26, it is not necessary to correct the correction waveform by the deviation. [Third Embodiment] FIG. 5 shows a main structure of a charged particle beam exposure apparatus according to the third embodiment, and FIG. 6 shows a novel electron beam scanning method according to the third embodiment.

In this apparatus, a blanking aperture array 28 is provided above the main deflector 25 via an aperture 27.
Are arranged. Blanking aperture array 28
6, the openings 60 are arranged in a zigzag pattern on a thin substrate, and each opening 60 is surrounded by a common electrode 61 and a blanking electrode 62 formed on the substrate. For example, the openings 60 are squares each side of which is 25 μm, and the number of openings is 16 × 64. An electron beam EB having a substantially uniform current density is projected onto the blanking aperture array 28 in a region indicated by a chain line.

By setting the voltage between the common electrode 61 and the blanking electrode 62 to 0 or Vs, the common electrode 61
The electron beam passing through the opening 60 surrounded by the blanking electrode 62 and the blanking electrode 62 passes through the aperture 27 and is irradiated onto the semiconductor wafer 21 or is blocked by the aperture 27. This action is shown in FIG. In FIG. 7, reference numeral 64 denotes an electromagnetic lens, and the directions A to C are added to the numbers of different components of the same type. The electrode potential is, for example, common electrode 61A to
61C, blanking electrodes 62A and 62C are 0, and blanking electrode 62B is Vs.

In FIG. 6, the openings 60 are arranged in a staggered pattern so that, for example, the gaps in the column direction of the exposure pattern by the electron beam that have passed through the openings in the second row 632 of the blanking aperture array 28 are mainly formed. After the electron beam EB is scanned by the deflector 25 in the direction B for one pitch, the first column 631 is scanned.
This is for filling with the exposure pattern of the electron beam that has passed through the opening of. This also applies to other columns. The blanking control circuit 7 has a voltage pattern corresponding to the pattern to be exposed on all the blanking electrodes 62.
Supplied from 0.

With this structure, it is possible to expose the semiconductor wafer 21 with an arbitrary fine pattern according to the voltage pattern applied to all the blanking electrodes 62. The scanning of the electron beam EB on the semiconductor wafer 21
This is performed as shown in FIG. That is, the electron beam EB
Is continuously scanned in the direction B by the main deflector 25, and in parallel with this, is continuously scanned in the direction C perpendicular to the direction B by the moving stage 22. For example, the length and width of the band BND are 2 mm and 10 μm, respectively, and the scanning time for one band is 100 μs. In this case, the moving speed of the moving stage 22 is 10 μm / 100 μs = 10
It is set to 0 mm / s. The sub-deflector 26 is used for deflecting the electron beam EB by following the movement of the moving stage 22 for each scanning by the main deflector 25, and its scanning width is equal to the width of the band BND, for example, 10 μm, Relatively narrow.

FIG. 9 shows the relationship between the stage detection position Yd, the Y-direction position Yj of the j-th band being scanned, and the stage feedback amount Yd-Yj for the sub deflector 26 in the stage movement direction Y. The stage feedback amount Yd-Yj is the position of the band being scanned with respect to the electron beam irradiation position when the Y-direction deflection amount of the sub deflector 26 is zero. Time point t1 corresponds to the beginning of the jth band being scanned and time point t3 corresponds to the end of this band. At time t2, the Y-direction deflection amount of the sub-deflector 26 is 0, and scanning is performed on the j-th band.

In FIG. 5, the change in the input value of the D / A converter 33 is small, and the output of the D / A converter 33 is supplied to the amplifier circuit 34 via a low-pass filter 36, for example, a CR integrator circuit. Therefore, the output XA of the amplifier circuit 34
Between the start end and the end of one scanning line by the main deflector 25, as shown in FIG. 8, continuously changes without waveform correction, and the settling waiting time is unnecessary. Therefore, when the main deflector 25 turns the scanning line back from the end to the start, the transient waveform as shown in FIG. 8 included in the output of the amplifier circuit 34 is corrected by the waveform data stored in the waveform correction memory 41A. Good.

When the count value i of the counter 30 changes from n to 0, that is, when the main deflector 25 swings back from the end to the start of one scanning line, the carry output of the counter 30 transits from 0 to 1. , RS flip-flop 50 is set, AND gate 51 is opened, and clock CL
K2 is supplied to the clock input terminal CK of the counter 40. The count value j of the counter 40 is initially zero-cleared. Corrected waveform data f from the waveform correction memory 41A
(J), j = 0 to m is output, the count value j of the m-ary counter 40 changes from m to 0, and the carry is 0 to 1.
When transitioning to, the RS flip-flop 50 is reset and the AND gate 51 is closed.

The corrected waveform data f (j) is supplied to the adder 42, is added to the stage feedback amount Xso, and its sum Xsoa and the similar sum Ysoa for the Y axis are supplied to the coordinate conversion circuit 45. It The coordinate conversion circuit 45
The difference between the designed exposure position and the actual exposure position is corrected. This difference is due to the arrangement accuracy of the sub-deflector 26, the difference between the deflection sensitivity in the X-axis direction and the deflection sensitivity in the Y-axis direction, the coordinate system of the main deflector 25 and the sub-deflector 26, and the coordinates on the semiconductor wafer 21. It is caused by the positional deviation from the system. The output Xsa of the coordinate conversion circuit 45 is, for example, Xsa = g.
It becomes Xsoa + rYsoa + o. Here, g, r and o are constants.

In this way, the deflection signal Xsa includes the correction waveform data f (j), j = 0 to m corresponding to the transient waveform included in the drive current XA, and the drive current XA and the correction waveform f are included. Is as shown in FIG. And the electron beam EB
The deflection by the main deflector 25 based on the transient waveform is canceled by the deflection by the sub deflector 26 based on the correction waveform f. Therefore, the settling waiting time based on the transient waveform is omitted, and the throughput of exposure by the electron beam EB is improved.

In the first embodiment, the main deflector 25
Between the sub-fields F as shown in FIG. 10 due to an error in the deflection direction (deviation of the XY deflection coordinate system) between the sub-deflector 26 and the sub-deflector 26 and an error in the deflection sensitivity (deflection intensity with respect to the input of the deflector). A connection deviation may occur, and the connection deviation increases in proportion to the length of one side of the subfield F. However, the continuous scanning method according to the third embodiment does not cause such a problem.

Further, since the saw waveform is generated by the counter 30, the main deflector scanning memories 31 and 32, and the low-pass filter 36, the scanning speed depends on the frequency of the clock CLK1 and the analog sawtooth generating circuit using the time constant circuit. The adjustment can be performed more accurately and easily, and the change of the scanning speed due to the ambient temperature or the change over time as in the analog sawtooth wave generation circuit can be prevented.

Although the irradiation position accuracy of the electron beam EB on the semiconductor wafer 21 is inversely proportional to the scanning width by the deflector, the scanning width by the sub-deflector 26 is, for example, 1 as described above.
Since the width is relatively narrow at 0 μm, the electron beam irradiation position accuracy in the sub-deflection direction becomes high. Further, since this scanning width is narrow, the high-speed and high-precision D / A converter 4 has a small number of bits.
3 can be used. Therefore, high-speed and high-precision scanning of the electron beam EB by the sub-deflector 26 becomes possible.

A coordinate conversion circuit similar to the above for correcting the arrangement error of the main deflector 26 and the difference between the deflection sensitivity in the X-axis direction and the deflection sensitivity in the Y-axis direction is, for example, the D / A converter 3.
3 and the low-pass filter 36, but this is omitted in FIG. 5 for simplification. This point is the same for FIGS. 1 and 4 and the following examples. [Fourth Embodiment] FIG. 11 shows the essential structure of a charged particle beam exposure apparatus according to the fourth embodiment. The fourth embodiment is a modification of the third embodiment and has the same relationship as the second embodiment with respect to the first embodiment.

That is, the output of the waveform correction memory 41A is analogized by the D / A converter 43, and the low-pass filter 4 is used.
After passing through 6, it is added to the low pass filter 3 by the adder 35.
It is added to the output of 6 and supplied to the amplifier circuit 34. The analog signal Xsb corresponding to Xsa when f (j) = 0 in FIG. 5 is supplied to the amplifier circuit 44. In addition,
The present invention includes various modifications.

For example, in the above embodiment, the case where the correction waveform data for the transient waveform included in the output of the amplifier circuit 34 is stored in the waveform correction memory 41 has been described, but in order to facilitate the creation of the correction waveform data, Data for correcting the transient waveform based on the characteristic of only one of the D / A converter 33 and the amplifier circuit 34 is stored in the waveform correction memory 41, and the transient waveform based on the other characteristic is corrected by another method. It may be.

Although the transient waveform included in the output of the amplifier circuit 44 has not been discussed in the above embodiment, the transient waveform is corrected by another method or by correcting the data stored in the waveform correction memory 41. May be Also,
Since the output g (i) of the main deflector scanning memory 31 is determined by the input i, the count value i may be supplied as the upper address of the waveform correction memory.

Alternatively, the output of the counter 30 may be supplied to the D / A converter 33 without using the main deflector scanning memory 31.

[0053]

As described above, both the charged particle beam exposure method according to the first aspect of the invention and the charged particle beam exposure apparatus according to the second aspect of the invention drive when the output of the D / A converter changes. Since the effect of the transient waveform included in the signal on the charged particle beam is canceled by the correction signal, the settling waiting time of the transient waveform can be eliminated or shortened, and the throughput of the charged particle beam exposure can be improved.
Since the correction signal is determined in advance, it has an excellent effect that the cancellation is effectively performed.

Since the transient waveform is determined according to the change of the input of the D / A converter, the first aspect of the first aspect of the invention has an effect that the above-mentioned cancellation is performed more effectively. According to the second aspect of the first aspect of the present invention, the above cancellation can be performed more completely by simply rewriting the correction data stored in the storage means, and the correction waveform is independent of changes in ambient temperature and aging. Has the effect of being.

According to the third aspect of the first aspect of the present invention, it is possible to cancel the transient waveform that differs depending on the change in the input value of the D / A converter.
In the fourth aspect of the first invention, since the charged particle beam is continuously deflected in the predetermined direction by the main deflector, if the drive signal includes the correction signal only when the charged particle beam is swung back in the direction opposite to the predetermined direction. Has the effect of being good.

According to the fifth aspect of the first aspect of the present invention, since a problem such as connection shift between subfields does not occur, it is possible to accurately expose a fine pattern. Further, since the scanning width of the sub-deflector is relatively narrow, the irradiation position accuracy of the charged particle beam on the exposure target in the sub-deflection direction is high, and the high-speed and high-precision D / A converter with a small number of bits is used. Can be used, and therefore, an effect that the charged particle beam can be scanned at high speed and with high accuracy by the sub-deflector is obtained.

According to the sixth aspect of the first invention, the scanning speed can be adjusted more accurately and easily than the analog sawtooth wave generation circuit using the time constant circuit by the clock frequency.
Moreover, it is possible to prevent the change of the scanning speed due to the ambient temperature and aging like an analog sawtooth wave generation circuit, and even if the scanning speed is changed according to the sensitivity of the resist, a fine pattern that requires high precision. There is an effect that it becomes possible to expose.

According to the seventh aspect of the first invention, it is possible to expose a finer and more arbitrary pattern. According to the first aspect of the second aspect of the present invention, it becomes possible to cancel the transient waveform that differs depending on the change in the input value of the D / A converter, and rewrite the correction data stored in the storage means. Only by doing so, it is possible to perform the cancellation more completely, and further, it is possible to obtain an effect that the correction waveform is irrelevant to changes in the ambient temperature and changes over time.

According to the second aspect of the second invention, the method of the fifth aspect of the first invention can be realized, and the effect can be obtained. According to the third aspect of the second invention, there is an effect that the method of the sixth aspect of the first invention can be realized and the effect thereof can be obtained.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a main part of a charged particle beam exposure apparatus according to a first embodiment of the present invention.

FIG. 2 is an explanatory diagram of an electron beam scanning direction.

FIG. 3 is a diagram showing a D / A converter output waveform and a correction waveform.

FIG. 4 is a configuration diagram of a main part of a charged particle beam exposure apparatus according to a second embodiment of the present invention.

FIG. 5 is a configuration diagram of a main part of a charged particle beam exposure apparatus according to a third embodiment of the present invention.

FIG. 6 is an explanatory diagram of another electron beam scanning direction.

FIG. 7 is a diagram for explaining the operation of the blanking aperture array.

8 is a diagram showing an output waveform and a correction waveform of the amplifier circuit 34 of FIG.

FIG. 9 is a diagram showing a relationship between a stage detection position, a scanning band position, and a stage feedback amount for a sub deflector with respect to a stage moving direction.

FIG. 10 is a diagram illustrating a problem of the first embodiment.

FIG. 11 is a configuration diagram of a main part of a charged particle beam exposure apparatus according to a fourth embodiment of the present invention.

FIG. 12 is a diagram for explaining the problems of the prior art, (A) showing D
A circuit diagram of the / A converter, and (B) is an output waveform diagram of this circuit.

[Explanation of symbols]

 21 semiconductor wafer 22 moving stage 25 main deflector 26 sub-deflector 28 blanking aperture array 30, 40 counter 31 main deflector scanning memory 33, 43 D / A converter 34, 44 amplifying circuit 35, 42 adder 36, 46 Low-pass filter 41, 41A Waveform correction memory 45 Coordinate conversion circuit

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Hiroshi Yasuda 1015 Kamiodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa Fujitsu Limited

Claims (12)

[Claims] 1. A main deflector (25) and a narrower scanning range than the main deflector for scanning a charged particle beam (EB) incident on an exposure object (21) on the exposure object. The sub-deflector (26) deflects the charged particle beam,
In the charged particle beam exposure method of exposing the exposure object with the charged particle beam, a drive signal is supplied to the main deflector or the sub deflector based on the output of a D / A converter (33, 43), In order to cancel the influence of the transient waveform included in the drive signal on the charged particle beam when the output of the D / A converter changes, the drive signal supplied to the main deflector or the sub-deflector is previously A charged particle beam exposure method, characterized in that it includes a defined correction signal. 2. The correction signal is supplied to the D / A converter (3
The charged particle beam exposure method according to claim 1, wherein the signal is a signal predetermined according to a change in the input of 3). 3. The charged particle beam exposure method according to claim 2, wherein the correction signal is generated by sequentially reading the correction data stored in the storage means (41). 4. Each time the input value of the D / A converter (33) changes, a value corresponding to the input value is loaded into a counter (40) as an initial value and a clock (CLK
4. The charged particle beam exposure method according to claim 3, further comprising the step of supplying 2) to change the count value of the counter, and to address the storage means (41) with the count value. 5. The main deflection is performed when the charged particle beam (EB) is continuously deflected by the main deflector (25) in a predetermined direction and the charged particle beam is swung back to the opposite side to the predetermined direction. 5. The charged particle beam exposure method according to claim 1, wherein the correction signal is included in a drive signal supplied to the device or the sub deflector (26). 6. The object to be exposed (21) is mounted on a moving stage (22), the moving stage is continuously moved in a direction perpendicular to a deflection direction of the main deflector (25), and 6. The charged particle beam exposure method according to claim 5, wherein the charged particle beam is deflected by the sub deflector in accordance with the movement of the moving stage for each scanning by the main deflector. 7. A low-pass filter that causes the data in the second storage means (31) to be read by the count value of the second counter (30) and supplies the data to the D / A converter (33) for analogization. A drive signal is supplied to the main deflector (25) through (36) and an amplifier circuit, and a clock (CLK1) is supplied to the clock input terminal of the second counter to supply the drive signal with a cycle of the count value. 7. The charged particle beam exposure method according to claim 6, wherein the scanning speed of the main deflector is made variable by setting the sawtooth wave equal to the cycle of the above and changing the frequency of the clock. 8. A plurality of openings (60) are formed in a substrate in a zigzag pattern, and a pair of electrodes (61, 61) is formed at the edge of each opening of the substrate.
62) forming a blanking aperture array (2
8) is arranged at a position before the charged particle beam (EB) passes through the main deflector (25) and the sub deflector (26), and the charged particle beam is passed through the blanking aperture array. It is controlled whether or not the charged particle beam passing through the opening is irradiated onto the exposure object (21) depending on whether or not a multi-beam is applied and a voltage is applied between the pair of electrodes. The charged particle beam exposure method according to claim 6 or 7, wherein a pattern according to the applied voltage is exposed on the exposure target. 9. A main deflector (25) and a narrower scanning range than said main deflector for scanning a charged particle beam (EB) incident on an exposure object (21) on said exposure object (21). The sub-deflector (26) deflects the charged particle beam,
In a charged particle beam exposure apparatus for exposing the exposure object with the charged particle beam, a first D / A converter (33), an output end of the first D / A converter and an input end of the main deflector are provided. A first amplifier circuit (34) coupled between the second amplifier circuit (44) and an input terminal of the sub-deflector (44)
And an output of the first amplifier or the second amplifier for canceling the influence of the transient waveform included in the output of the first amplifier on the charged particle beam when the input of the first D / A converter changes. A charged particle beam exposure apparatus, comprising: a waveform correction circuit (40, 41, 50, 51) for including a predetermined correction signal in a signal. 10. The waveform correction circuit is configured such that when an input value of the first D / A converter (33) changes, an initial value corresponding to the input value is set and a clock (CLK
A counter (40) for counting 2), and a second D / connected to the front stage of the second amplifier circuit (44).
The count value of the counter is supplied to the A converter (43) and the address input terminal, the correction waveform data is stored, and the read data is stored in the second converter.
The charged particle beam exposure apparatus according to claim 9, further comprising a storage unit (41) included in an input of the D / A converter. 11. The waveform correction circuit sets an initial value when the input value of the first D / A converter (33) changes from one of the maximum value and the minimum value to the other and sets a clock (CLK2). A counter (40) for counting, and a second D / connected to the front stage of the second amplifier circuit (44).
The count value of the counter is supplied to the A converter (43) and the address input terminal, the correction waveform data is stored, and the read data is stored in the second converter.
The charged particle beam exposure apparatus according to claim 9, further comprising a storage unit (41) included in an input of the D / A converter. 12. A second counter (30), data addressed to a count value of the second counter, for driving a main deflector is stored, and the read data is stored in the first D / A converter. And a low-pass filter (36) coupled between the first D / A converter and the first amplifier circuit (34). The charged particle beam exposure apparatus according to claim 9.