JP3461066B2 - Charged particle beam exposure method and apparatus - 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 for shaping a cross section of a charged particle beam with a stencil mask, and irradiating the semiconductor wafer or mask having a resist film with the charged particle beam for exposure. An exposure method and apparatus.

[0002]

2. Description of the Related Art FIG. 11 shows a schematic structure of a conventional charged particle beam exposure apparatus using a stencil mask. The cross section of the charged particle beam, for example, the electron beam emitted from the charged particle gun 10 is shaped by the selected transmission hole pattern on the stencil mask 12, and is reduced and irradiated onto a wafer (not shown). AX is the optical axis, and 13A is the trajectory of the charged particle beam. 21A, 21B, 22A,
22B and 23 are electromagnetic lenses, and 30 and 31 are aperture plates.

The charged particle beam passing through the electromagnetic lens 21B along the optical axis AX passes through the deflector 4 in order to pass the charged particle beam perpendicularly to the selected transmission hole pattern on the stencil mask 12.
It is shaken by 5 and is made parallel to the optical axis AX by the deflector 46. The charged particle beam that has passed through the stencil mask 12 is returned to the optical axis AX side by the deflector 47, and is deflected by the deflector 48 so as to pass on the optical axis AX.

This type of charged particle beam exposure apparatus is
Since the charged particle beam is vertically incident on the stencil mask 12, there is an advantage that the cross-sectional shaping accuracy of the charged particle beam is high and the transmission hole pattern on the stencil mask 12 can be selected in a relatively wide range. However, in any of the deflectors 45 to 48, the change of the charged particle beam position on the aperture plate 31 and the wafer (not shown) with respect to the change of the deflection amount is large (the deflection efficiency is large). This becomes a problem when selecting the pattern of the transmission holes. For example, when the deflection amount by the deflector 46 is slightly increased, the charged particle beam has a locus 13B indicated by a chain double-dashed line. On the other hand, the accuracy required for the charged particle beam irradiation position on the wafer is, for example, 0.01 μm.
It is as small as m. When selecting a transmission hole pattern, the deflector 45
A driving voltage that changes in steps is applied to ~ 48, and the necessary settling waiting time becomes long, for example, ta shown in FIG.

In order to accurately irradiate the charged particle beam on the target position on the wafer, the charged particle beam is blanked by a blanking deflector (not shown) and the aperture plate 31 during the settling time, and this time is set. Beam irradiation on the wafer must be started after the elapse of time, resulting in a dead time and a decrease in exposure throughput. 12
Is a schematic structure of a main part of another conventional charged particle beam exposure apparatus using a stencil mask (Japanese Patent Laid-Open No. 62-206828).
Issue gazette).

In this device, the crossover images CO1 and CO2 of the gun crossover CO of the charged particle gun 10 are present between the electromagnetic lens 21B and the electromagnetic lens 22 and between the electromagnetic lens 22 and the electromagnetic lens 23, respectively. Deflector 4
0 is arranged with its center point in the optical axis direction aligned with the position of the crossover image CO1, and similarly, the deflector 49 is arranged with its center point in the optical axis direction aligned with the position of the crossover image CO2. . 13C is an enlarged view of the vertical cross section of the charged particle beam passing through the electromagnetic lenses 21B, 22 and 23 in a direction perpendicular to the optical axis AX when the deflection amounts of the deflectors 40 and 49 are 0. is there.

The charged particle beam deflected by the deflector 40 to pass the charged particle beam through the selected transmission hole pattern on the stencil mask 12 passes through the position of the crossover image CO2 due to the converging action of the electromagnetic lens 22, Deflector 4
It is deflected by 9 so as to pass on the optical axis AX. In this type of charged particle beam exposure apparatus, the deflector 4 shown in FIG.
Since those corresponding to 6 and 47 are not arranged, the selection range of the transmission hole pattern on the stencil mask 12 becomes narrower than that in the case of FIG.

However, since the deflection efficiency of the deflectors 40 and 49 on the aperture plate 31 and on the wafer is small,
There is an advantage that the required settling waiting time is shorter than that in the case of FIG. 11, for example, tb shown in FIG.

[0009]

In FIG. 12, since the crossover image CO2 is actually located at a position several mm above the upper end of the electromagnetic lens 23, the center point of the deflector 49 in the optical axis direction is actually set. The position of the crossover image CO2 cannot be matched. Therefore, the charged particle beam deflected by the deflector 49 deviates from the optical axis AX, and the accuracy of the charged particle beam irradiation position on the wafer deteriorates. Further, the mask deflectors 40 and 49 on the aperture plate 31 and the wafer. The deflection efficiency above becomes larger than that in the case of FIG. 11 , and the required settling waiting time becomes longer.

In view of these problems, the object of the present invention is to improve the accuracy of the charged particle beam irradiation position on the object to be exposed and to change the driving voltage of the stencil mask deflector stepwise. A charged particle beam exposure method and apparatus capable of shortening the settling waiting time.

[0011]

In the first invention, for example, in FIG. 1, the first electromagnetic lens 21, the second electromagnetic lens 22, the third electromagnetic lens 23 and the aperture 31 coincide with each other in the optical axis AX. And arranged in this order,
A stencil mask 12 having a plurality of transmission hole patterns for shaping the cross section of the charged particle beam is arranged substantially at the second electromagnetic lens position perpendicular to the optical axis AX, and the charged particle beam is in the first electromagnetic lens 21. A selected transmission hole pattern on the stencil mask 12 in the second electromagnetic lens 22;
The exposure target 11 is irradiated through the third electromagnetic lens 23 and the aperture 31 in this order, and the first cross of the gun crossover CO is provided between the first electromagnetic lens 21 and the second electromagnetic lens 22. An over image CO1 is present, and is between the second electromagnetic lens 22 and the third electromagnetic lens 23 and the third electromagnetic lens 23.
The second crossover image CO2 of the gun crossover CO is present on the electromagnetic lens 23 side, and the central axis is made to coincide with the optical axis AX, and the first electromagnetic lens 21 and the second electromagnetic lens 22 are arranged.
And the mask incident-side electrostatic deflector 40 disposed between the second electromagnetic lens 22 and the optical axis AX.
And the second electromagnetic lens 22 between the third electromagnetic lens 23 and
The first electrostatic deflector 41 on the mask emission side disposed on the side of the second electromagnetic lens 22 with its central axis aligned with the optical axis AX.
And the third electromagnetic lens 23 between the third electromagnetic lens 23 and
Charged particle beam having a mask emission side second electrostatic deflector 42 arranged on the side so that the second crossover image CO2 exists between the center point in the optical axis direction and the third electromagnetic lens 23. A charged particle beam exposure method using an exposure apparatus, wherein an output voltage of an amplification circuit 51 is applied to the mask emission side first electrostatic deflector 41 and the mask emission side second electrostatic deflector 42 in common. , A line obtained by extending the charged particle beam with respect to the mask incident side electrostatic deflector 40 to the side opposite to the traveling direction of the charged particle beam emitted from the mask incident side electrostatic deflector 40 is the first cross. The charged particle beam is deflected so as to substantially pass through the position of the overimage CO1, and the second electromagnetic lens 22
Incident obliquely with respect to the mask exit side first electrostatic deflector 41
On the other hand, the charged particle beam is deflected so as to be incident on the mask emitting-side second electrostatic deflector 42 toward a substantially central point of the mask emitting-side second electrostatic deflector 42 in the optical axis direction. The charged particle beam is deflected by the emission-side second electrostatic deflector 42 so as to substantially coincide with the optical axis.

According to the first aspect of the present invention, even if the second crossover image CO2 does not exist at the center point in the optical axis direction of the electrostatic deflector, the charged particle beam that has passed through the second electrostatic deflector on the mask emission side. Is almost on the optical axis regardless of the deflection amount of the electrostatic deflector, and therefore the charged particles on the aperture and on the exposure target are different from the deflection amount which changes stepwise when the transmission hole pattern is selected on the stencil mask. The fluctuation of the beam position is small, the accuracy of the charged particle beam irradiation position on the exposure target is improved, and
It is possible to shorten the settling waiting time required when the step changes.

When output voltages of different amplifier circuits are applied to the first electrostatic deflector on the mask emission side and the second electrostatic deflector on the mask emission side, the transient characteristics of the respective amplifier circuits differ, so that the settling wait time is increased. It is necessary to match the longest settling wait time, and the determination of the settling wait time is complicated. However, according to the first aspect of the present invention, the output voltage of the amplifier circuit causes the mask output side first electrostatic deflector and the mask output side first output voltage. Since the voltage is commonly applied to the two electrostatic deflectors, the settling waiting time is easily determined, and the throughput of the exposure apparatus is improved.

In the first aspect of the first invention, referring to FIG. 1 or 5, for example, in the charged particle beam exposure apparatus, the central point in the optical axis direction of the mask incident side electrostatic deflector 40 is the first point.
The mask incident side electrostatic deflector 40 is arranged so as to substantially coincide with the position of the crossover image CO1, and the mask incident side electrostatic deflector 40 is further aligned with the central axis thereof coincident with the optical axis AX. And the second electromagnetic lens 22 or between the first electromagnetic lens 21 and the mask incident side electrostatic deflector 40, and the length in the optical axis direction is larger than that of the mask incident side electrostatic deflector 40. Between the first correction electrostatic deflector 43 and the mask output side first electrostatic deflector 41 and the mask output side second electrostatic deflector 42 with their central axes aligned with the optical axis AX. Alternatively, it is arranged between the mask emitting side first electrostatic deflector 41 and the second electromagnetic lens 22, and the length in the optical axis direction is the mask emitting side first electrostatic deflector 41 and the mask emitting side second. Second correction electrostatic deflector 4 shorter than any of the electrostatic deflectors 42
4 and the first correction electrostatic deflector 43
The deviation of the trajectory of the charged particle beam emitted from the mask entrance side electrostatic deflector 40 due to the positional deviation of the crossover image CO1 is corrected, and the second correction electrostatic deflector 44 corrects the second crossover image. The deviation of the trajectory of the charged particle beam emitted from the second electrostatic deflector 42 on the mask emission side due to the positional deviation of CO2 is corrected.

According to the first aspect, the stress applied to the electromagnetic lens fluctuates due to fluctuations in thermal expansion due to fluctuations in ambient temperature, fluctuations in atmospheric pressure outside the lens barrel (not shown), and the charged particle gun. Even if the electrode consumption amount of the second crossover image is increased and the positions of the first crossover image and the second crossover image are changed, the charged particle beam that has passed through the second electrostatic deflector on the mask emission side is statically corrected by the above correction. Since the light passes through the optical axis substantially regardless of the deflection amount of the electric deflector, the charged particle beam position on the aperture and the exposure target object is changed with respect to the deflection amount that changes stepwise when selecting the transmission hole pattern on the stencil mask. The fluctuation is small, the accuracy of the charged particle beam irradiation position on the exposure target is improved, and the settling waiting time required when the step is changed can be shortened.

The deflection efficiencies on the aperture and the exposure object are higher than those of the mask entrance side electrostatic deflector, the mask exit side first electrostatic deflector and the mask exit side second electrostatic deflector. The first correction electrostatic deflector and the second correction electrostatic deflector are larger, and the first correction electrostatic deflector and the second correction electrostatic deflector are for correcting a small error. At this time, the first correction electrostatic deflector and the second correction electrostatic deflector can perform high-speed correction, which greatly reduces the settling waiting time.

Further, since the first crossover image is separated from the first correction electrostatic deflector, and the first correction electrostatic deflector is for correcting a small error, the first correction on the stencil mask is performed. The deflection efficiency of the electrostatic deflector is considerably smaller than that of the electrostatic entrance deflector of the mask entrance side, and the influence of the first correction electrostatic deflector on the pattern selection position on the stencil mask is small.

In the second aspect of the first invention, for example, in FIG. 6, in the charged particle beam exposure apparatus, the mask incident side electrostatic deflector makes the central axis coincide with the optical axis, and the first electromagnetic wave is obtained. A mask incident on the first electromagnetic lens side between the lens and the second electromagnetic lens such that the first crossover image exists between the center point in the optical axis direction and the first electromagnetic lens. Side first electrostatic deflector 142, and a mask incidence side first unit disposed on the side of the second electromagnetic lens between the first electromagnetic lens and the second electromagnetic lens with the central axis aligned with the optical axis. 2 electrostatic deflector 141, and further, with the central axis aligned with the optical axis, between the mask incident side first electrostatic deflector and the mask incident side second electrostatic deflector, or It is arranged between the mask incident side second electrostatic deflector and the second electromagnetic electrostatic deflector, and the optical axis direction A first correction electrostatic deflector 144 having a length shorter than that of both the mask incident side first electrostatic deflector and the mask incident side second electrostatic deflector, and a central axis thereof aligned with the optical axis. And is arranged between the mask emitting side first electrostatic deflector and the mask emitting side second electrostatic deflector or between the mask emitting side first electrostatic deflector and the second electromagnetic lens, A second whose length in the direction of the optical axis is shorter than both of the mask emitting side first electrostatic deflector and the mask emitting side second electrostatic deflector
A correction electrostatic deflector 44 is provided, and the output voltage of the amplifier circuit is applied commonly to the mask incident side first electrostatic deflector and the mask incident side second electrostatic deflector, and the mask incident side first electrostatic deflector is applied. The charged particle beam incident through the optical axis is deflected by one electrostatic deflector, and the charged particle beam is emitted by the second electrostatic deflector on the mask incident side. The line extended to the opposite side in the traveling direction is deflected so as to substantially pass through the position of the first crossover image, and the mask is incident by the first correction electrostatic deflector due to the position shift of the first crossover image. The deviation of the locus of the charged particle beam emitted from the second side electrostatic deflector is corrected, and the second correction electrostatic deflector causes the position shift of the second crossover image to cause the mask emission side second static image. The deviation of the trajectory of the charged particle beam emitted from the electric deflector is corrected.

According to the second aspect, the same effect as that of the first aspect can be obtained. In a third aspect of the first invention, the mask emitting side first electrostatic deflector and the mask emitting side second
The intensity of the first electromagnetic lens is changed in a state where the deflection amount of the electrostatic deflector is 0, and the passing current of the aperture with respect to the deflection amount of the mask incident side electrostatic deflector is minimized. Strength A1 of the mask incident side electrostatic deflector, and changing the intensity of the first electromagnetic lens with the deflection amount of the mask incident side electrostatic deflector set to 0, When the same voltage is applied to the two electrostatic deflectors, the passing current of the aperture with respect to the deflection amount becomes the minimum.
The strength A2 of the electromagnetic lens is obtained, and the strength L of the first electromagnetic lens is compared with the ratio M: N of the voltage applied to the first correction electrostatic deflector and the voltage applied to the second correction electrostatic deflector. Is L = (M · A1 + N · A2) / (M + N).

When the intensity of the first electromagnetic lens is A1, the position error of the first crossover image is 0, but there is an error in the position of the second crossover image, and the intensity of the first electromagnetic lens is A.
When it is 2, the position error of the second crossover image is 0, but there is an error in the position of the first crossover image. Therefore, the preferable lens strength L of the first electromagnetic lens for the above correction is A
Neither 1 nor A2. If M / (M + N)> N / (M
+ N), the preferable lens strength L is considered to be a value closer to A1 than A2, and conversely, M / (M + N)
If <N / (M + N), the preferable lens strength L is
It is considered that the value is closer to A2 than A1. From the above, it can be said that the above lens strength L is relatively preferable.

In the third mode, not only the first correction electrostatic deflector and the second correction electrostatic deflector but also these are organically correlated with the strength of the first electromagnetic lens to perform the correction. The correction is more accurate. In a fourth aspect of the first aspect of the invention, an initial value is given to the ratio M: N to obtain the strength L of the first electromagnetic lens by the above equation, the strength of the first electromagnetic lens is set to the L, and the mask emission side first When the deflection amount of the electrostatic deflector and the second electrostatic deflector on the mask exit side is set to 0 and the deflection amount of the electrostatic deflector on the mask entrance side is not set to 0, the first correction electrostatic deflector 1 The applied voltage is changed to obtain the first applied voltage VS1 at which the current passing through the aperture with respect to the first applied voltage is minimized, and the strength of the first electromagnetic lens is set to the L to set the mask incident side electrostatic deflector. To the second correction electrostatic deflector in the state where the same non-zero voltage is applied to the mask emission side first electrostatic deflector and the mask emission side second electrostatic deflector. By changing the applied voltage, the second
The second applied voltage VS2 at which the passing current of the aperture with respect to the applied voltage is minimized is obtained, and the ratio M: N is updated based on the first applied voltage VS1 and the second applied voltage VS2, and the updated value is updated. The intensity L of the first electromagnetic lens is updated based on the ratio M: N and the equation.

According to the fourth aspect, the relatively preferable lens strength L of the first electromagnetic lens can be experimentally obtained. In the fifth aspect of the first invention, for example, in FIG.
The mask emission side first electrostatic deflector 41 or the mask emission side second electrostatic deflector 42 is rotated around the optical axis AX to adjust its rotation angle.

Since the output voltage of the same amplifier circuit is applied to the mask emission side first electrostatic deflector and the mask emission side second electrostatic deflector, it is possible to electrically correct the rotational position error around the optical axis. Although only one of the mask-exiting-side first electrostatic deflector and the mask-exiting-side second electrostatic deflector can be used, according to the fifth aspect, the other can be mechanically corrected. First
In the sixth aspect of the invention, the mask incident side first electrostatic deflector or the mask incident side second electrostatic deflector is rotated around the optical axis to adjust the rotation angle.

According to the sixth aspect, the same effect as the fifth aspect can be obtained. In the seventh aspect of the first invention, for example, in FIG. 8, the output voltage of the amplifier circuit 51 is changed to the mask incident side electrostatic deflector 40, the mask emitting side first electrostatic deflector 41, and the mask emitting side second. The voltage is commonly applied to the electrostatic deflector 42. When output voltages of different amplifier circuits are applied to the mask incident side electrostatic deflector, and the mask emitting side first electrostatic deflector and the mask emitting side second electrostatic deflector, the transient characteristics of each amplifier circuit are different. It is necessary to match the settling wait time with the longest settling wait time, and the setting wait time is complicated to determine. However, according to the seventh aspect, the output voltage of the amplifier circuit is commonly applied to these electrostatic deflectors. Therefore, the settling waiting time can be easily determined, and the throughput of the exposure apparatus can be improved.

In the eighth aspect of the first invention, for example, in FIG. 9, the charged particle beam exposure apparatus further includes storage means 81 and 82, a digital / analog converter 71, and an output of the digital / analog converter 71. And an amplifier circuit 50 that amplifies and applies to the mask incident side electrostatic deflector,
For each transmission hole pattern on the stencil mask 12, a correction value pattern for maximizing a transient charged particle beam irradiation amount at the time of irradiation of the charged particle beam on the exposure object is stored in advance in the storage means. Every time the data D1 supplied to the digital / analog converter 71 for selecting the transmission hole pattern on the stencil mask 12 changes, a correction value pattern corresponding to the data D1 is stored in the storage means. Based on the sum of the value D3 read out from the storage means and the correction value D2 when the irradiation of the charged particle beam onto the exposure object 11 is in a steady state, the first value A drive voltage is applied to the correction electrostatic deflector 43.

The deflection efficiency of the first correction electrostatic deflector on the aperture is larger than that of the mask entrance side electrostatic deflector, and the first correction electrostatic deflector is for correcting a small error. The first correction electrostatic deflector can be driven at a considerably higher speed than the mask incident side electrostatic deflector. According to the eighth aspect, the mask incident side electrostatic deflector can be driven when the step changes. The change in the aperture passing current, which changes substantially in proportion to the cube of the applied voltage, can be reduced, the settling waiting time is shortened, and the exposure throughput is improved.

In a ninth aspect of the first aspect of the invention, in the charged particle beam exposure apparatus, the storage means has a first memory and a second memory, and for each transmission hole pattern on the stencil mask, A pattern of a transient charged particle beam irradiation amount at the time of irradiation of the charged particle beam on the exposure object is stored in the first memory, and the charged particle beam irradiation amount is set for each transmission hole pattern on the stencil mask. The corresponding correction value is stored in the second memory, and the second memory is addressed by the output data of the first memory.

According to the ninth aspect, the first memory in which the pattern of the transient charged particle beam irradiation amount, for example, the secondary electron detection amount pattern when the charged particle beam is irradiated onto the object to be exposed is stored. , The relationship between the charged particle beam irradiation amount and the correction value can be grasped, and the correction value can be easily updated to a more preferable value. In the charged particle beam exposure apparatus according to the second invention, for example, as shown in FIG. 1, the first electromagnetic lens, the second electromagnetic lens, the third electromagnetic lens and the aperture are arranged in this order with their optical axes aligned with each other. A stencil mask having a plurality of transmission hole patterns for shaping the beam cross section is disposed substantially at a position of the second electromagnetic lens perpendicular to the optical axis, and the charged particle beam is in the first electromagnetic lens and the second electromagnetic lens. The selected object is irradiated through the selected transmission hole pattern on the stencil mask, the third electromagnetic lens and the aperture in this order onto the object to be exposed, and between the first electromagnetic lens and the second electromagnetic lens. There is a first crossover image of the gun crossover, and a second crossover image of the gun crossover exists between the second electromagnetic lens and the third electromagnetic lens and on the side of the third electromagnetic lens. In the charged particle beam exposure apparatus, the central axis of the charged particle beam exposure apparatus is arranged between the first electromagnetic lens and the second electromagnetic lens, and the emitted charged particle beam is extended to the opposite side in the traveling direction. A mask incident side electrostatic deflector for deflecting the generated line so as to substantially pass through the position of the first crossover image, a central axis thereof aligned with the optical axis, and the second electromagnetic lens and the third electromagnetic lens. A mask output side first electrostatic deflector for deflecting the charged particle beam to the optical axis side, which is arranged on the second electromagnetic lens side between the electromagnetic lens and the central axis, is aligned with the optical axis. , The second
On the side of the third electromagnetic lens between the electromagnetic lens and the third electromagnetic lens, the charged particle beam is incident toward a substantially central point in the optical axis direction, and the central point in the optical axis direction and the third electromagnetic field. A second electrostatic deflector on the mask emission side for arranging the second crossover image so as to exist between the lens and the second electrostatic deflector for deflecting the charged particle beam so that the charged particle beam substantially coincides with the optical axis. possess an amplification circuit connected in common to the said mask exit side first electrostatic deflector and the mask exit side second electrostatic deflector, the charged
The particle beam obliquely enters the second electromagnetic lens .

In the first aspect of the second invention, for example, as shown in FIG. 1, the mask entrance side electrostatic deflector is arranged so that the center point in the optical axis direction substantially coincides with the position of the first crossover image. Has been done. In the second aspect of the second invention, for example, FIG.
As shown in FIG. 5, the central axis is made to coincide with the optical axis, and between the mask incident side electrostatic deflector and the second electromagnetic lens or between the first electromagnetic lens and the mask incident side electrostatic deflector. And a first correction electrostatic deflector having a length in the optical axis direction shorter than that of the mask entrance side electrostatic deflector, and the mask exit side first electrostatic deflector with its central axis aligned with the optical axis. Between the mask output side second electrostatic deflector or the mask output side first electrostatic deflector and the second electromagnetic lens, and the length in the optical axis direction is the mask output side first electrostatic deflector. 1 electrostatic deflector and a second correction electrostatic deflector that is shorter than either the mask emission side second electrostatic deflector, and the first correction electrostatic deflector is configured to detect the first crossover image. It is for correcting the deviation of the trajectory of the charged particle beam emitted from the electrostatic deflector on the mask entrance side due to the positional deviation. The second correction electrostatic deflector is for correcting the deviation of the trajectory of the charged particle beam emitted from the mask emission side second electrostatic deflector due to the positional deviation of the second crossover image. is there.

In the third aspect of the second invention, as shown in FIG. 6, for example, in the mask incident side electrostatic deflector, the central axis is made to coincide with the optical axis, and the first electromagnetic lens and the second electromagnetic lens are arranged. It is arranged on the side of the first electromagnetic lens between the lens and the lens so that the first crossover image exists between the center point in the optical axis direction and the first electromagnetic lens, and is incident through the optical axis. A mask-incident-side first electrostatic deflector for deflecting the charged particle beam, and a second electromagnetic field between the first electromagnetic lens and the second electromagnetic lens with a central axis aligned with the optical axis. A mask incident side second electrostatic deflector disposed on the lens side for deflecting a line extending the emitted charged particle beam to the opposite side in the traveling direction so as to pass substantially through the position of the first crossover image; Further, the output end is provided with the mask incidence side first electrostatic deflector and the mask incidence side. Having an amplifier circuit which is connected in common to a second electrostatic deflector.

In the fourth aspect of the second invention, as shown in FIG. 6, for example, in the third aspect, the central axis is made to coincide with the optical axis, and the first electrostatic deflector on the mask incidence side and the mask incidence side. Between the second electrostatic deflector on the side and the second side on the mask incidence side
It is arranged between the electrostatic deflector and the second electromagnetic electrostatic deflector, and has a length in the optical axis direction which is larger than that of either the mask incident side first electrostatic deflector or the mask incident side second electrostatic deflector. Between the mask emission side first electrostatic deflector and the mask emission side second electrostatic deflector with the central axis aligned with the optical axis. It is arranged between the mask emission side first electrostatic deflector and the second electromagnetic lens, and has a length in the optical axis direction of the mask emission side first electrostatic deflector and the mask emission side second electrostatic deflector. A second correction electrostatic deflector that is shorter than either of them, and the first correction electrostatic deflector emits the mask incident side second electrostatic deflector due to the position shift of the first crossover image. The second correction electrostatic deflector is for correcting the deviation of the trajectory of the charged particle beam. Deviation due, is used to correct the deviation of the trajectory of the charged particle beam emitted the mask exit side second electrostatic deflector.

In the fifth aspect of the second invention, for example, as shown in FIG. 2, the mask emitting side first electrostatic deflector or the mask emitting side second electrostatic deflector is rotated around the optical axis. It has adjusting means for adjusting the rotation angle. In a sixth aspect of the second invention, there is provided an adjusting means for rotating the mask incident side first electrostatic deflector or the mask incident side second electrostatic deflector around the optical axis and adjusting a rotation angle thereof. .

In the seventh aspect of the second invention, for example, as shown in FIG. 8, the output end of the amplifier circuit is commonly connected to the mask incident side electrostatic deflector. In the eighth aspect of the second invention, the amplifier circuit of the second invention and the amplifier circuit of the third aspect of the second invention are the same. In the ninth aspect of the second invention, for example, as shown in FIG. 9, the transient charged particle beam irradiation amount at the time of irradiation of the charged particle beam on the exposure target is maximized for each of the transmission hole patterns on the stencil mask. Storage means for storing a correction value pattern for
A digital / analog converter, an amplifier circuit for amplifying the output of the digital / analog converter and applying it to the mask incident side electrostatic deflector, and a digital / analog converter for selecting a transmission hole pattern on the stencil mask. Data read control means for reading a correction value pattern corresponding to the data from the storage means each time the data supplied to the analog converter changes, a value read from the storage means, and the exposure object. First correction electrostatic deflector driving means for applying a drive voltage to the first correction electrostatic deflector based on a sum with a correction value when the irradiation of the charged particle beam is in a steady state. Have.

In a tenth aspect of the second aspect of the present invention, the storage means, for each transmission hole pattern on the stencil mask, a transient charged particle beam irradiation amount at the time of irradiation of the charged particle beam on the object to be exposed. The first stored pattern
It has a memory and a second memory in which a correction value corresponding to the charged particle beam irradiation amount is stored for each transmission hole pattern on the stencil mask and which is addressed by the output data of the first memory.

[0035]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. [First Embodiment] FIG. 1 shows the schematic arrangement of a charged particle beam exposure apparatus according to the first embodiment of the present invention. Charged particle gun 10
The resist film on the semiconductor wafer 11 is exposed by a charged particle beam, for example, an electron beam emitted from the semiconductor wafer 11.
A plurality of frequently used transmission hole patterns are formed on the stencil mask 12, and the cross section is shaped by passing the charged particle beam through the selected transmission hole pattern, so that a fine pattern can be formed into a semiconductor in one shot. The image is reduced and projected onto the wafer 11.

Reference numeral 13 shows a longitudinal section of the charged particle beam passing through the electromagnetic lenses 21 to 25 in an enlarged manner in a direction perpendicular to the optical axis AX. The electromagnetic lens 21A includes an electromagnetic lens 21A and an electromagnetic lens 21B, an aperture plate 30 is disposed between the electromagnetic lenses, and the charged particle beam collimated by the electromagnetic lens 21A passes through a square hole of the aperture plate 30. The stencil mask 12 is sized to pass only one through hole pattern. Electromagnetic lens 22
Consists of an electromagnetic lens 22A and an electromagnetic lens 22B,
A stencil mask 12 is arranged between both electromagnetic lenses, and the charged particle beam collimated by the electromagnetic lens 22A passes through a selected transmission hole pattern on the stencil mask 12. Electromagnetic lens 23 and electromagnetic lens 24 for reduced projection
An aperture plate 31 having a circular hole is disposed between and. The electromagnetic lens 25 is an objective lens and focuses the charged particle beam on the semiconductor wafer 11.

Semiconductor wafer 1 for charged particle beam
Deflection to the target position on 1 is performed by the main deflector 32 and the sub deflector 33. The crossover images CO1, CO2, and CO3 of the gun crossover CO are between the electromagnetic lens 21B and the electromagnetic lens 22A, and the electromagnetic lens 22B, respectively.
And the electromagnetic lens 23 and between the electromagnetic lens 23 and the electromagnetic lens 24.

In order to deflect the charged particle beam passing through the optical axis AX to pass through the selected transmission hole pattern on the stencil mask 12, a deflector 40 is provided between the electromagnetic lens 21B and the electromagnetic lens 22A. The center point in the direction is aligned with the position of the crossover image CO1. The charged particle beam that has passed through the electromagnetic lens 22B is deflected so that the optical axis A
Deflectors 41 and 42 are arranged between the electromagnetic lens 22B and the electromagnetic lens 23 so as to pass over X. The deflector 42 is arranged on the electromagnetic lens 23 side, but since the crossover image CO2 is located at a position several mm above the upper end of the electromagnetic lens 23, the center point of the deflector 42 in the optical axis direction is the crossover image CO2. Therefore, the deflector 41 is disposed on the electromagnetic lens 22B side.

The aperture plate 31 is arranged so that the center point of the hole coincides with the position of the crossover image CO3. The stress applied to the electromagnetic lens fluctuates due to the fluctuation of the thermal expansion due to the fluctuation of the ambient temperature and the fluctuation of the atmospheric pressure outside the lens barrel (not shown), and the electrode consumption of the charged particle gun 10 increases, so that the cross The positions of the overimages CO1 and CO2 slightly change. As a result, or due to a deviation from the design value, the charged particle beam that has passed through the electromagnetic lens 23 has an optical axis A.
Since it deviates from above X, in order to correct this, the deflector 4
The correction deflector 43 is arranged between 0 and the electromagnetic lens 22A, and the correction deflector 44 is arranged between the deflector 41 and the deflector 42.

The output terminal of the amplifier 50 is connected to the deflector 40, the output terminal of the amplifier 51 is commonly connected to the deflectors 41 and 42, and the output terminals of the amplifiers 53 and 54 are connected to the correction deflectors 43 and 44, respectively. The ends are connected. Deflector 4
Nos. 0 to 44 are electrostatic types, and potentials having the same absolute value and opposite polarities are applied to the counter electrodes. Since the deflection angles of the deflectors 40 to 44 are small, the trajectory of the charged particle beam has a linear approximation with respect to the deflector applied voltage, and the ratio of the voltages applied to the deflectors 40 to 44 is the pattern selection on the stencil mask 12. It is fixed regardless of the position.

A more detailed schematic structure of the deflectors 41, 42 and 44 is shown in FIG. The deflectors 41, 42 and 44 each have an eight-pole structure in order to make the internal electric field more uniform (the same applies to the deflectors 40 and 43 in FIG. 1). Corresponding to this, the amplifier 51 of FIG.
Consists of. The output terminals of the amplifiers 511 to 518 are respectively
Is connected to the electrode pieces 411 to 418 of the deflector 41, and
It is connected to the electrode pieces 421 to 428 of the deflector 42.

The deflectors 41 and 42 are supporters 6 respectively.
It is supported by 0 and 61. The central axes of the deflectors 41, 42 and 44 coincide with the optical axis AX (see FIG. 1).
The same applies to the deflectors 40 and 43 of the above). The error in the rotational position of the deflector 41 or 42 around the optical axis AX is caused by the amplifiers 511-5.
By performing coordinate conversion on the signal supplied to 18,
It is corrected electrically. However, the deflector 42 and the deflector 41
Since the output voltage of the same amplifier is applied to the electrode pieces corresponding to and, the electrical correction is performed by the deflector 41 and the deflector 42.
Only one can be applied. Therefore, teeth 62 are formed on a part of the peripheral surface of the supporter 60, and a worm 63 is meshed with the teeth 62. By rotating a knob 64 attached to the shaft of the worm 63, the rotation angle of the deflector 41 is changed to the mechanical angle. Will be corrected.

In FIG. 3A, when the voltage applied to the deflectors 41 to 43 is set to 0 and the charged particle beam incident on the deflector 40 through the optical axis AX is deflected by the deflector 40,
A linear approximation of the charged particle beam trajectory is as in 13D. That is, since the center point of the deflector 40 in the optical axis direction coincides with the position of the crossover image CO1, the charged particle beam deflected by the deflector 40 becomes parallel to the optical axis AX when passing through the electromagnetic lens 22A. , Then electromagnetic lens 22B
When it passes through, it is converged to the position of the crossover image CO2,
Then, it is converged to the position of the crossover image CO3.

In FIG. 3B, the voltage applied to the deflectors 40, 43 and 44 is set to 0, and the charged particle beam is made to enter upward from below the aperture plate 31 through the optical axis AX to make the deflector 42 and Assuming that the deflector 41 deflects the light, the linear approximation of the locus is set to 13E. That is, the deflection amount of the deflector 41 is set to the points P4 and P
The chain double-dashed line that extends the straight line connecting 3 and 3 to the opposite side in the traveling direction passes through the position of the crossover image CO2. If this relationship is established when a certain same voltage is applied to the deflectors 41 and 42, this relationship is established for any applied voltage within the range where the linear approximation is established. The charged particle beam that has passed through the electromagnetic lens 22B becomes parallel to the optical axis AX, and when passing through the electromagnetic lens 22A, it is converged at the position of the crossover image CO1.

When the voltage VM1 is applied to the deflector 40 in FIG. 3A and the voltage VM2 is applied to the deflectors 41 and 42 in FIG. 3B, the point P3 in FIG. It is assumed that the point P3 in (B) is the same. This relationship shows that if the voltage ratios VM1: VM2 at this time are kept constant, the voltages VM1 and V1
It holds even if M2 is changed. The voltage VM1 is applied to the deflector 40.
And a voltage VM2 is applied to the deflectors 41 and 42, the linear approximation of the trajectory of the charged particle beam incident on the deflector 40 through the optical axis AX is represented by points P1 and C in FIG.
It passes through O1, P2 and P3 and points P4, P5 and P6 in FIG.

The deflectors 41 and 42 are designed to deflect the charged particle beam as described above when the same voltage is applied to them. The design error in this case can be corrected by adjusting the distance between the deflector 41 and the deflector 42. In FIG. 1, the crossover image CO
Even if the voltages VM1 and VM2 are changed when the error between the positions of 1 and CO2 is 0, the positions of the charged particle beam on the aperture plate 31 and the semiconductor wafer 11 do not change (deflection efficiency is 0). In contrast, the voltages VS1 and VS2
Is changed, the positions of the charged particle beam on the aperture plate 31 and the semiconductor wafer 11 are changed. Deflector 4
Regarding 0 to 44, there are the following properties.

(1) Even if there is an error in the positions of the crossover images CO1 and CO2, the deflection efficiency on the aperture plate 31 and the semiconductor wafer 11 is higher than that of the deflectors 40, 41 and 42 by the correction deflector 43. And 44 are larger and the correction deflectors 43 and 44 are for correcting a small error, the lengths of the correction deflectors 43 and 44 in the optical axis direction are considerably shorter than those of the deflectors 40, 41 and 42. can do.

(2) The length of the correction deflector 43 in the optical axis direction is considerably shorter than that of the deflector 40, and the crossover image C
Since O1 is distant from the correction deflector 43, the deflection efficiency of the deflector 40 on the stencil mask 12 is considerably larger than that of the correction deflector 43, and the correction deflector 43 for the pattern selection position on the stencil mask 12 is large. Is preferable because the influence of is small.

The position error of the crossover images CO1 and CO2 can be corrected by using the electromagnetic lens 21B and the correction deflectors 43 and 44, which will be described below. Figure 4
13A to 13C, 13F to 13H are the deflector 4
The electromagnetic lenses 21 to 21 when the applied voltage to 0 to 44 is 0
The vertical section of the trajectory of the charged particle beam passing through
It is shown enlarged in the direction perpendicular to.

A current I passing through the hole of the aperture plate 31
Is measured by placing a Faraday cup 55 below the hole. As shown in FIG. 4A or 4C, when the position of the crossover image CO1 is displaced from the center position of the deflector 40 in the optical axis direction, the applied voltage to the deflectors 41 to 44 is set to 0.
When the charged particle beam is deflected by the deflector 40, the current I changes according to the deflection amount. The electric current supplied to the electromagnetic lens 21 (to be exact, the electromagnetic lens 21B in FIG. 1, hereinafter the same) is changed to change the lens strength and the position of the crossover image CO1 is changed on the optical axis AX. Figure 4
As shown in (B), the position of the crossover image CO1 is set to the deflector 4
When it coincides with the center position of 0 in the optical axis direction, the current I does not change even if the deflection amount of the deflector 40 is changed.

Therefore, the applied voltage to the deflectors 41 to 44 is set to 0, the strength of the electromagnetic lens 21 is changed, and the change in the current I with respect to the change in the deflection amount of the deflector 40 is measured for each lens strength. Electromagnetic lens 2 that minimizes current I
The strength A1 of 1 is obtained. Next, the deflectors 40, 43 and 4
4 is set to 0, the strength of the electromagnetic lens 21 is changed, and for each lens strength, the current I with respect to the change in the deflection amount in the deflectors 41 and 42 to which the same voltage is applied.
Is measured, and the intensity A2 of the electromagnetic lens 21 at which the current I is minimized is obtained.

When the intensity of the electromagnetic lens 21 is A1, the position error of the crossover image CO1 is 0, but there is an error in the position of the crossover image CO2, and when the intensity of the electromagnetic lens 21 is A2, the position of the crossover image CO2. Error is 0
However, since there is an error in the position of the crossover image CO1, the preferable lens strength L of the electromagnetic lens 21 is neither A1 nor A2. Assuming that the preferable lens strength L is obtained, the lens strength of the electromagnetic lens 21 is set to L.

The voltage applied to the deflectors 41, 42 and 44 is set to 0, a voltage VS1 proportional to the voltage VM1 applied to the deflector 40 is applied to the correction deflector 43, and its proportional constant K1.
For each proportional constant K1, the change in the current I with respect to the change in the deflection amount of the deflector 40 is measured, and the proportional constant number K1 at which the current I becomes the minimum is obtained. Next, the voltage applied to the deflector 40 and the correction deflector 43 is set to 0, a voltage VS2 proportional to the voltage VM2 applied to the deflectors 41 and 42 is applied to the correction deflector 44, and the proportional constant K2 is changed. ,
For each proportionality constant K2, the change in the current I with respect to the change in the deflection amount of the deflectors 41 and 42 is measured, and the proportionality constant K2 at which the current I becomes the minimum is obtained. From such K1 and K2, the applied voltage ratio VS1: VS2 = M: N is obtained. The ratio M: N is constant regardless of the selected position on the stencil mask 12.

Next, the relationship between the ratio M: N and the preferable lens strength L of the electromagnetic lens 21 will be considered. If M /
If (M + N)> N / (M + N), lens strength L
Is considered to be a value closer to A1 than A2. On the contrary, if M / (M + N) <N / (M + N), the lens strength L is considered to be a value closer to A2 than A1. Therefore, the applied voltage to the deflectors 41 to 44 is set to 0,
The strength of the electromagnetic lens 21 is changed, and the change in the current I with respect to the change in the deflection amount of the deflector 40 is measured for each lens strength, and the strength A1 of the electromagnetic lens 21 at which the current I is minimized is obtained.

Next, the applied voltage to the deflectors 40, 43 and 44 is set to 0, the strength of the electromagnetic lens 21 is changed, and the same voltage is applied to each of the lens strengths of the deflector 4.
The change in the current I with respect to the change in the deflection amount at 1 and 42 is measured, and the strength A2 of the electromagnetic lens 21 at which the current I becomes the minimum is obtained. L = (M · A1 + N · A2) / (M + N) (1) is determined. The ratio M: N has an initial value of, for example, 1: 1
The lens strength L at this time is obtained by the above equation (1), and under this lens strength L, M: N is obtained as described above, and M: N is obtained by performing such processing once or a plurality of times. Make N close to the optimum value.

M: N can also be obtained as follows. The current I and the charged particle beam irradiation point P on the semiconductor wafer 11 when the Faraday cup 55 is removed are
It depends on the voltages VS1 and VS2. Therefore, the initial value of M: N is set to, for example, 1: 1 and the lens strength L at this time is obtained by the above equation (1).
An empirical formula of the relationship between 1 and VS2 and the current I and the irradiation point P is obtained, and M: N where the current I is maximum and the irradiation point P is on the optical axis AX is obtained. By performing such processing once or a plurality of times, M: N is brought close to the optimum value.

The above-mentioned proportional constants K1 and K2 and the lens strength L are determined at the time of calibration of the charged particle beam exposure apparatus. In the first embodiment, not only the correction deflectors 43 and 44, but also these are deflected by the electromagnetic lens 21B.
Since the correction is performed in association with the intensity of, the correction can be performed more accurately.

The relationship between the settling waiting time when the voltages VM1 and VM2 are stepwise changed and this correction is as follows. When selecting the transmission hole pattern on the stencil mask 12, the voltages VM1 and VM2 are set as shown in FIG.
The step changes as shown in FIG. The accuracy required for the charged particle beam irradiation position on the semiconductor wafer 11 is as small as 0.01 μm, and it is necessary to blank the charged particle beam until this accuracy is ensured.
Since the charged particle beam is continuously emitted, this blanking is performed by the aperture plate 31 and the deflector (not shown) arranged above the aperture plate 31. Even if the step changes of the voltages VM1 and VM2 are the same, the necessary settling waiting time (blanking time) is equal to the voltages VM1 and VM2.
The larger the change (deflection efficiency) of the charged particle beam position on the semiconductor wafer 11 with respect to the change of 2, the longer the length.

If the above correction is not made, this deflection efficiency becomes large, so the settling waiting time is, for example, t in FIG.
However, this correction reduces the deflection efficiency and shortens it to t1 = 500 ns, for example. This allows
The exposure throughput is improved. In the first embodiment,
Since the correction is performed more accurately as described above, the effect of reducing the settling waiting time is great.

[Second Embodiment] FIG. 5 shows a schematic structure of a main part of a charged particle beam exposure apparatus according to the second embodiment of the present invention. In this device, a correction deflector 43 is arranged between the deflector 40 and the electromagnetic lens 21B, and a correction deflector 44 is arranged between the deflector 41 and the electromagnetic lens 22B. The other points are the same as those in FIG.

Also in the second embodiment, the properties (1) and (2) described in the first embodiment are secured. [Third Embodiment] FIG. 6 shows a schematic structure of a main part of a charged particle beam exposure apparatus according to the third embodiment of the present invention.

In this device, the deflectors 40 and 43 of FIG.
Instead of the above, the deflectors 41, 42 and 44 are arranged on the opposite side in the optical axis direction. That is,
The deflector 1 is provided between the electromagnetic lens 21B and the electromagnetic lens 22A.
42, 144 and 141 are arranged corresponding to the deflectors 42, 44 and 41, respectively. Deflectors 141 and 1
The output voltage VM1 of the amplifier 151 is commonly applied to 42, and the output voltage V1 of the amplifier 154 is applied to the correction deflector 144.
S1 is applied. The crossover image CO1 exists between the center point on the optical axis of the deflector 142 and the electromagnetic lens 21B. Others are the same as the configuration of FIG.

FIGS. 7A and 7B are respectively shown in FIG.
It corresponds to (A) and (B). In FIG. 7 (A),
The voltage applied to the deflectors 144, 41, 42 and 44 is 0
When the charged particle beam incident on the deflector 142 through the optical axis is deflected by the deflector 142 and further deflected by the deflector 141, the linear approximation of the trajectory of the charged particle beam is 13
Try to be like I.

In FIG. 7B, the deflectors 141 and 14
When the voltage applied to Nos. 2, 144 and 44 is set to 0, the charged particle beam is incident upward from below the aperture plate 31 through the optical axis AX, and is deflected by the deflectors 42 and 41, a linear approximation of its locus is obtained. So that it looks like 13J. In FIG. 7A, the voltage VM1 is applied to the deflectors 141 and 142.
Is applied, the voltage VM is applied to the deflectors 41 and 42 in FIG.
2 is applied, point P3 in FIG. 7 (A) and FIG. 7 (B)
It is assumed that the point P3 is the same as the point P3. This condition holds even if the voltages VM1 and VM2 are changed if the voltage ratios VM1: VM2 at this time are kept constant.

When the voltage VM1 is applied to the deflectors 141 and 142 and the voltage VM2 is applied to the deflectors 41 and 42,
The linear approximation of the trajectory of the charged particle beam that has entered the deflector 142 through the optical axis AX is as shown by points P1 and P5 in FIG. 7A.
A, P4A, P2 and P3, and the point P4 in FIG.
Go through P5 and P6. Others are the same as in the case of the first embodiment.

[Fourth Embodiment] FIG. 8 shows a schematic structure of a main part of a charged particle beam exposure apparatus according to the fourth embodiment of the present invention. The voltage applied to the deflector 40 is equal to that of the deflectors 41 and 4
In this device, the size of the deflector 40 is determined so that the proportional constant is 1, and the output voltage VM2 of the amplifier 51 is applied to the deflector 40 in common. ing. 1 of the deflectors 40, 41 and 42
One can electrically adjust the rotation angle around the optical axis AX, and the other two can be mechanically adjusted as in the case of FIG. Others are the same as the configuration of FIG.

In the fourth embodiment, since the output voltage VM2 of the amplifier 51 is commonly applied to the deflectors 41, 42 and 40, when the output voltage VM2 of the amplifier 51 changes stepwise, the deflectors 41 and 42 are changed. And 40 have the same transient characteristics of the electric field, facilitating the determination of the settling time. That is, the deflector 40 and the deflector 4
If different amplifiers are used for 1 and 42, the transient characteristics of each amplifier are different, so it is necessary to match the settling wait time with the longest settling wait time, and it is complicated to determine the settling wait time. The embodiment eliminates the need, and facilitates the determination of the settling wait time.

[Fifth Embodiment] FIG. 9 shows the arrangement of an aperture passing current adjusting unit of a charged particle beam exposure apparatus according to the fifth embodiment of the present invention. As described in the first embodiment, since the trajectory of the charged particle beam has a relatively small deflection amount, a linear approximation is established regarding the voltage applied to the deflectors 40 to 43. The second-order term of the applied voltage causes aberration but does not affect the locus of the center line of the charged particle beam. The third-order term of the applied voltage affects the locus of the center line of the charged particle beam,
When the applied voltage changes stepwise, it causes the position of the charged particle beam on the aperture plate 31 to change. Therefore, the current passing through the aperture plate 31 is equal to the applied voltage of 3
It changes in almost proportion to the power. At this time, the semiconductor wafer 11
Although the fluctuation of the above charged particle beam irradiation point can be ignored, it affects the exposure amount, so it is necessary to consider the settling waiting time for the third-order term of the applied voltage.

For example, the accuracy required for the charged particle beam irradiation point on the semiconductor wafer 11 is 0.01 μm. The maximum value of the deflection amount (distance from the optical axis AX) of the charged particle beam on the stencil mask 12 is 5 mm. When the reduction ratio of the projection of the through hole pattern on the semiconductor wafer 11 on the semiconductor wafer 11 is set to 1/100, the step change of 5 mm on the stencil mask 12 corresponds to the step change of 50 μm on the semiconductor wafer 11, 0.01 μm change above
To achieve the following, in FIG. 10A, it is necessary to set the difference ΔV of the voltage VM1 to the stable value V0 to ΔV / V0 ≦ 0.01 / 50 = 1/5000 (2). The time t1 when ΔV / V0 = 1/5000 is the settling waiting time, which is actually 5.00 × 10
^-7 sec = about 500 nsec.

It is approximated as VM1 (t) = V0 {1-exp (−t / a)}, and ΔV / V0 = at t = 5.00 × 10 ⁻⁷ sec.
Assuming 1/5000, a = 6.03 × 10 ⁻⁸ s
It becomes ec. Current I passing through the hole in the aperture plate 31
Since (t) is approximately proportional to {VM1 (t)} ³ , it is approximated as I (t) = I0 {1-exp (-t / a)} ³ . In order to make the exposure amount accurate, the change of the irradiation amount of the charged particle beam onto the semiconductor wafer 11 is a stable value of 99.
To blank the charged particle beam until
Considering the projection reduction ratio 1/100, in FIG. 10B, the difference ΔI0 of the current I (t) with respect to the stable value I0 is expressed as ΔI0 / I0 ≦ (1/100) · (1/100) = 1/10000 ... (3) is required. The time t3 when ΔI0 / I0 = 1/10000 is the settling waiting time, which is about 620 ns. In this case, the necessary settling waiting time is determined by the above equation (3) instead of the above equation (2).

The settling waiting time t3 in FIG.
10A corresponds to the settling wait time t2 in FIG. 10A, and is the case where the correction described in the first embodiment is not performed, and the fluctuation of the charged particle beam position on the aperture plate 31 is It becomes larger and longer than the settling waiting time t 1 . The deflection efficiency on the aperture plate 31 and on the semiconductor wafer 11 is larger in the correction deflector 43 than in the deflector 40, and the correction deflector 43 is for correcting a small error. VS1 is considerably smaller than the output voltage VM1 of the amplifier 50. Therefore, the settling time t5 of the voltage VS1 is equal to the voltage VM1 as shown in FIG.
It is considerably smaller than that. Due to this property, the correction deflector 43 can be used to correct the fluctuation of the charged particle beam position on the aperture plate 31.

In the apparatus shown in FIG. 9, when the voltage applied to the deflectors 40 to 44 changes stepwise, the position of the charged particle beam on the aperture plate 31 changes due to the third term of the applied voltage. The correction is performed by the device 43, and the components 70 to 84 are added to the configuration of FIG. The data D1 for selecting the transmission hole pattern on the stencil mask 12 is stored in the data generator 70, and the pattern selection data D1 sequentially read from the data generator 70 is analogized by the D / A converter 71. Amplifier 50
Is amplified to a voltage VM1 and applied to the deflector 40. The voltage VS1 which is analogized by the D / A converter 72 and amplified by the amplifier 53 is applied to the correction deflector 43. D /
In the A converter 72, the correction value D2 = K1 · D1 and the correction value D
The value obtained by adding 3 and 3 in the adder 73 is supplied. K1 is the proportional constant described above.

At the stage of creating the correction value for the correction deflector 43, D3 = 0 is set, and when the pattern selection data D1 is stepwise changed, the value substantially proportional to the current I passing through the hole of the aperture plate 31 is 2 It is detected by the secondary electron detector 74, amplified by the amplifier 75, converted into a digital value SE by the A / D converter 76, and supplied to the control circuit 77. At the time of this step change, a voltage proportional to the deflector 40 is supplied to the deflectors 41 to 43.
The applied voltage to 43 also changes stepwise.

At the start of step change of the pattern selection data D1, the load signal supplied from the data generator 70 to the load control input terminal L of the counter 78 becomes active, and the pattern selection data D1 becomes the data input terminal D of the counter 78. And is loaded into the counter 78. Immediately after this, the data generator 70 sets the start signal ST to the high level. As a result, the AND gate 79 is opened, and the clock CLK from the clock generator 80 is supplied to the clock input terminal CK of the counter 78 via the AND gate 79. The memory 81 is addressed by the count value fetched from the data output terminal Q of the counter 78, and the control circuit 77 writes the secondary electron detection amount SE to this address. This writing is performed for each data D1 corresponding to the transmission hole pattern on the stencil mask 12.

The memory 82 stores 2 for each data D1 corresponding to the transmission hole pattern on the stencil mask 12.
A correction value D3 for maximizing the secondary electron detection amount SE is written. The correction value D3 is determined as follows, for example. For a certain pattern selection data D1 and a certain secondary electron detection amount SE before correction (at a certain time t after the start of step change), the correction value D3 is changed, and the secondary electron detection amount SE for each correction value D3 is measured. The empirical formula SE
(D3) is obtained, and the correction value D3 for maximizing the secondary electron detection amount SE is determined. This is performed for each of the plurality of points of the secondary electron detection amount SE before correction, and by the interpolation method,
Correction value D3 for arbitrary secondary electron detection amount SE before correction
Ask for. This is performed for each data D1 corresponding to the transmission hole pattern on the stencil mask 12.

The correction value D3 (D1,
SE) is supplied to the control circuit 77, the control circuit 77
The address (D1, S
E) is supplied, and the correction value D3 is written in this address.
At the time of exposure, the output of the memory 81 is used instead of the output of the A / D converter 76. This is because the output of the secondary electron detector 74 depends on the unevenness on the semiconductor wafer 11, so that the memory 8 for the flat portion on the semiconductor wafer 11 is
This is because the secondary electron detection amount SE written in 1 is used.

In the exposure, the memories 81 and 82 are put in the read state, and the input of the selector 83 is set to the memory 81 side. When the data D1 is step-changed to select the transmission hole pattern on the stencil mask 12, the load control input terminal L of the counter 78 is activated and the pattern selection data D1 is loaded into the counter 78. Immediately thereafter, the start signal ST Is set to a high level, the clock CLK from the clock generator 80 is counted by the counter 78, and the memory 81 is addressed by the counted value,
The data SE is read from the memory 81, and the selector 83
The correction value D3 is read from the memory 82 and supplied to the adder 73.

The correction value D3 for certain pattern selection data D1 is shown in FIG. With this correction value D3,
The aperture passing current I becomes as shown by the alternate long and short dash line in FIG. 10B, and the settling waiting time is shortened from t3 to t5 which is shorter than t1. This allows the stencil mask 1
The settling waiting time required when selecting the 2 through hole pattern is t
Therefore, the exposure throughput is improved.

Although it is possible to directly address the memory 82 with the counter 78, in the fifth embodiment,
Since the memory 81 in which the secondary electron detection amount SE is stored is used, the relationship between the secondary electron detection amount SE and the correction value D3 can be grasped, and the correction value D3 can be easily updated to a more preferable value.

[Brief description of drawings]

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

FIG. 2 is a perspective view showing a schematic configuration of a mask emission side deflector in FIG.

FIG. 3 is an explanatory diagram of a charged particle beam trajectory before and after a stencil mask in the apparatus of FIG.

4 is an explanatory diagram of a method of adjusting a crossover image position of the apparatus of FIG.

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

FIG. 6 is a schematic configuration diagram of a main part of a charged particle beam exposure apparatus according to a third embodiment of the present invention.

7 is an explanatory diagram of a charged particle beam trajectory before and after a stencil mask in the apparatus of FIG.

FIG. 8 is a schematic configuration diagram of a main part of a charged particle beam exposure apparatus according to a fourth embodiment of the present invention.

FIG. 9 is a configuration diagram of an aperture passing current adjusting unit of a charged particle beam exposure apparatus according to a fifth embodiment of the present invention.

FIG. 10 is a waveform diagram showing an operation of the aperture passing current adjusting unit in FIG.

FIG. 11 is a schematic configuration diagram of a main part of a conventional charged particle beam exposure apparatus.

FIG. 12 is a schematic configuration diagram of a main part of another conventional charged particle beam exposure apparatus.

FIG. 13 is a diagram showing a settling waiting time that differs depending on the magnitude of the deflection efficiency.

[Explanation of symbols]

10 charged particle gun 11 semiconductor wafer 12 stencil mask CO gun crossover CO1 to CO3 crossover images 21 to 25 electromagnetic lens 30, 31 aperture plates 40 to 42, 141 to 144 deflectors 411 to 418, 421 to 428 electrode piece 43, 44 Correction deflectors 50 to 54, 75, 511 to 518, 151, 154
Amplifier 55 Faraday cup 70 Data generator 74 Secondary electron detector 78 Counter 80 Clock generator 81, 82 Memory

─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Takamasa Sato 1015 Kamiodanaka, Nakahara-ku, Kawasaki, Kanagawa Prefecture, Fujitsu Limited (72) Inventor, Hiroshi Yasuda 1015, Kamedotachu, Nakahara-ku, Kawasaki, Kanagawa Prefecture, Fujitsu Limited ( 56) References JP-A-7-130622 (JP, A) JP-A-7-78742 (JP, A) JP-A-5-36592 (JP, A) JP-A-6-326012 (JP, A) JP HEI 5-90140 (JP, A) JP HEI 6-84769 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H01L 21/027

Claims (21)

(57) [Claims] 1. A first electromagnetic lens, a second electromagnetic lens, and a third electromagnetic lens.
The electromagnetic lens and the aperture are arranged in this order with their optical axes aligned with each other, and a stencil mask having a plurality of transmission hole patterns for shaping the cross section of the charged particle beam is provided substantially perpendicular to the optical axis at the second electromagnetic lens position. An object to be exposed is arranged such that the charged particle beam passes through the first electromagnetic lens, the second electromagnetic lens, the selected transmission hole pattern on the stencil mask, the third electromagnetic lens, and the aperture in this order. There is a first crossover image of the gun crossover between the first electromagnetic lens and the second electromagnetic lens, and the first crossover image is present between the second electromagnetic lens and the third electromagnetic lens. 3 A second crossover image of the gun crossover exists on the electromagnetic lens side, a central axis is made to coincide with the optical axis, and a mask is incident between the first electromagnetic lens and the second electromagnetic lens. Side electrostatic A deflector, and a mask emission-side first electrostatic deflector arranged on the side of the second electromagnetic lens between the second electromagnetic lens and the third electromagnetic lens, with the central axis aligned with the optical axis. , With the central axis aligned with the optical axis, on the side of the third electromagnetic lens between the second electromagnetic lens and the third electromagnetic lens, between the center point in the optical axis direction and the third electromagnetic lens. A charged particle beam exposure method using a charged particle beam exposure apparatus having a mask emission side second electrostatic deflector arranged so that the second crossover image exists, wherein an output voltage of an amplification circuit is The charged particle beam is applied to the mask incident side electrostatic deflector and the mask incident side electrostatic deflector in common, and the charged particle beam is applied to the mask incident side electrostatic deflector. A line obtained by extending the charged particle beam emitted from the deflector to the opposite side in the traveling direction is the first cross-over line. The charged particle beam is deflected so that it almost passes through the position of the image.
The beam obliquely enters the second electromagnetic lens , and the charged particle beam is directed to the first electrostatic deflector on the mask emission side, at a substantially central point in the optical axis direction of the second electrostatic deflector on the mask emission side. Toward the mask output side second electrostatic deflector, and the mask output side second electrostatic deflector is deflected so that the charged particle beam substantially coincides with the optical axis. And a charged particle beam exposure method. 2. The charged particle beam exposure apparatus according to claim 1, wherein the mask incident side electrostatic deflector is arranged so that a center point in the optical axis direction of the mask incident side electrostatic deflector substantially coincides with a position of the first crossover image. And the central axis is aligned with the optical axis so as to be located between the mask incident side electrostatic deflector and the second electromagnetic lens or between the first electromagnetic lens and the mask incident side electrostatic deflector. Is disposed between the mask and the mask and has a length in the optical axis direction shorter than that of the electrostatic entrance deflector on the mask entrance side.
A correction electrostatic deflector and a central axis thereof aligned with the optical axis, between the mask emitting side first electrostatic deflector and the mask emitting side second electrostatic deflector, or the mask emitting side first static deflector. A second electrostatic deflector disposed between the electrostatic deflector and the second electromagnetic lens and having a length in the optical axis direction shorter than that of either the mask emitting side first electrostatic deflector or the mask emitting side second electrostatic deflector. A correction electrostatic deflector, wherein the trajectory of the charged particle beam emitted from the mask incident side electrostatic deflector is deviated due to the positional deviation of the first crossover image by the first correction electrostatic deflector. And the second correction electrostatic deflector corrects the deviation of the trajectory of the charged particle beam emitted from the mask emission side second electrostatic deflector due to the position deviation of the second crossover image. The charged particle beam exposure method according to claim 1. 3. The charged particle beam exposure apparatus according to claim 1, wherein the mask incident side electrostatic deflector is arranged between the first electromagnetic lens and the second electromagnetic lens with the central axis aligned with the optical axis. A mask incident side first electrostatic deflector arranged on the first electromagnetic lens side such that the first crossover image exists between the center point in the optical axis direction and the first electromagnetic lens; A mask incident side second electrostatic deflector arranged on the side of the second electromagnetic lens between the first electromagnetic lens and the second electromagnetic lens so as to match the optical axis, and further, The axis is aligned with the optical axis, and the mask incidence side first
It is arranged between the electrostatic deflector and the mask incident-side second electrostatic deflector or between the mask incident-side second electrostatic deflector and the second electromagnetic electrostatic deflector, and has a length in the optical axis direction. Is a first correction electrostatic deflector that is shorter than both the mask incident side first electrostatic deflector and the mask incident side second electrostatic deflector, and has the central axis aligned with the optical axis, It is arranged between the mask emission side first electrostatic deflector and the mask emission side second electrostatic deflector or between the mask emission side first electrostatic deflector and the second electromagnetic lens, and the optical axis direction A second correction electrostatic deflector having a length shorter than either of the mask emission side first electrostatic deflector and the mask emission side second electrostatic deflector, and the output voltage of the amplifier circuit is incident on the mask. Applied to both the first electrostatic deflector on the mask side and the second electrostatic deflector on the mask entrance side, and The charged particle beam incident therethrough is deflected, and a line obtained by extending the charged particle beam to the opposite side in the traveling direction with respect to the mask incident side second electrostatic deflector is The light is deflected so as to substantially pass through the position of the first crossover image, and the first correction electrostatic deflector outputs the second electrostatic deflector on the mask entrance side due to the position shift of the first crossover image. The deviation of the trajectory of the charged particle beam is corrected, and the second correction electrostatic deflector causes the charged particle beam emitted from the mask emission side second electrostatic deflector due to the positional deviation of the second crossover image. 2. The charged particle beam exposure method according to claim 1, wherein the deviation of the trajectory is corrected. 4. The intensity of the first electromagnetic lens is changed in a state in which the deflection amounts of the mask emission side first electrostatic deflector and the mask emission side second electrostatic deflector are set to 0, and the mask incidence side is changed. The intensity A1 of the first electromagnetic lens at which the current passing through the aperture is minimized with respect to the deflection amount of the electrostatic deflector is obtained, and the first electromagnetic lens is set with the deflection amount of the mask incident side electrostatic deflector set to 0. Change the intensity of the
The same voltage is applied to the electrostatic deflector and the second electrostatic deflector on the mask emission side, and the intensity A2 of the first electromagnetic lens that minimizes the passing current of the aperture with respect to the deflection amount is obtained. To the ratio M: N of the voltage applied to the electrostatic deflector and the voltage applied to the second correction electrostatic deflector.
4. The charged particle beam exposure method according to claim 2, wherein the intensity L of the electromagnetic lens is L = (M · A1 + N · A2) / (M + N). 5. An initial value is given to the ratio M: N to obtain the strength L of the first electromagnetic lens by the above equation, and the strength of the first electromagnetic lens is set to the L, and the mask output side first electrostatic deflector is set. And setting the deflection amount of the mask output side second electrostatic deflector to 0 and setting the deflection amount of the mask incident side electrostatic deflector to 0.
The first applied voltage VS1 for changing the first applied voltage to the first correction electrostatic deflector is calculated to obtain the first applied voltage VS1 at which the passing current of the aperture with respect to the first applied voltage is minimized. The strength of the electromagnetic lens is set to L, the deflection amount of the mask incident side electrostatic deflector is set to 0, and the same non-zero voltage is applied to the mask emitting side first electrostatic deflector and the mask emitting side second electrostatic deflector. In this state, the second applied voltage to the second correction electrostatic deflector is changed to obtain the second applied voltage VS2 at which the passing current of the aperture with respect to the second applied voltage is minimized, and the first applied voltage is applied. Characterized in that the ratio M: N is updated based on the voltage VS1 and the second applied voltage VS2, and the intensity L of the first electromagnetic lens is updated based on the updated ratio M: N and the formula. 5. The charged particle beam exposure method according to claim 4. 6. The mask output side first electrostatic deflector or the mask output side second electrostatic deflector is rotated around the optical axis to adjust a rotation angle thereof.
6. The charged particle beam exposure method according to any one of items 1 to 5. 7. The mask entrance side first electrostatic deflector or the mask entrance side second electrostatic deflector is rotated around the optical axis to adjust the rotation angle thereof.
The charged particle beam exposure method described. 8. An output voltage of an amplifier circuit is commonly applied to the mask entrance side electrostatic deflector, the mask exit side first electrostatic deflector, and the mask exit side second electrostatic deflector. 8. The charged particle beam exposure method according to claim 1, wherein the method is the charged particle beam exposure method. 9. The charged particle beam exposure apparatus further comprises a storage means, a digital / analog converter, and an amplifier circuit for amplifying an output of the digital / analog converter and applying the amplified output to the mask incident side electrostatic deflector. And, for each transmission hole pattern on the stencil mask,
A correction value pattern for maximizing the transient charged particle beam irradiation amount at the time of irradiation of the charged particle beam on the exposure object is stored in advance in the storage means, and the transmission hole pattern on the stencil mask is stored. Each time the data supplied to the digital / analog converter for selection changes, a correction value pattern corresponding to the data is read from the storage means, and the value read from the storage means and the exposure The driving voltage is applied to the first correction electrostatic deflector based on the sum of the correction value when the irradiation of the charged particle beam onto the object reaches a steady state. The charged particle beam exposure method described. 10. The charged particle beam exposure apparatus, wherein the storage means has a first memory and a second memory, and for each transmission hole pattern on the stencil mask,
A pattern of the transient charged particle beam irradiation amount at the time of irradiation of the charged particle beam on the exposure object is stored in the first memory, and the charged particle beam irradiation amount is stored for each transmission hole pattern on the stencil mask. 10. The charged particle beam exposure method according to claim 9, wherein a correction value corresponding to is stored in the second memory, and the second memory is addressed by the output data of the first memory. 11. A first electromagnetic lens, a second electromagnetic lens, a third electromagnetic lens and an aperture are arranged in this order with their optical axes aligned with each other, and a plurality of transmission hole patterns for shaping the cross section of the charged particle beam are formed. A stencil mask is arranged substantially at a position of the second electromagnetic lens perpendicular to the optical axis, and the charged particle beam is in the first electromagnetic lens, in the second electromagnetic lens,
The object to be exposed is irradiated through the selected transmission hole pattern on the stencil mask, the third electromagnetic lens, and the aperture in this order, and a gun is provided between the first electromagnetic lens and the second electromagnetic lens. Charged particle beam in which a first crossover image of a crossover exists and a second crossover image of the gun crossover exists between the second electromagnetic lens and the third electromagnetic lens and on the side of the third electromagnetic lens In the exposure apparatus, a line formed by extending the charged particle beam emitted to the opposite side in the traveling direction is disposed between the first electromagnetic lens and the second electromagnetic lens with the central axis aligned with the optical axis. A mask incident side electrostatic deflector for deflecting the first crossover image so as to substantially pass through the position, and a second electromagnetic lens and a third electromagnetic lens whose central axis coincides with the optical axis. Located on the side of the second electromagnetic lens between It is,
A mask electrostatic discharge side first electrostatic deflector for deflecting the charged particle beam toward the optical axis, and a central axis of the electrostatic deflector between the second electromagnetic lens and the third electromagnetic lens. On the side of the third electromagnetic lens, the charged particle beam is incident toward a substantially central point in the optical axis direction, and the second particle is disposed between the central point in the optical axis direction and the third electromagnetic lens.
A mask output side second electrostatic deflector for deflecting the charged particle beam so that the charged particle beam is substantially aligned with the optical axis and an output end for the mask output side first electrostatic device are arranged so that a crossover image exists. A deflector and an amplifier circuit commonly connected to the mask emission side second electrostatic deflector.
Then, the charged particle beam is oblique to the second electromagnetic lens.
Charged particle beam exposure apparatus characterized by incidence . 12. The charging device according to claim 11, wherein the mask incident side electrostatic deflector is arranged such that a central point in the optical axis direction substantially coincides with a position of the first crossover image. Particle beam exposure system. 13. A center axis is made to coincide with the optical axis, and between the mask incident side electrostatic deflector and the second electromagnetic lens or between the first electromagnetic lens and the mask incident side electrostatic deflector. A first correction electrostatic deflector disposed between the first and second electrostatic deflectors whose length in the optical axis direction is shorter than that of the mask incident side electrostatic deflector; It is arranged between the deflector and the mask emission side second electrostatic deflector or between the mask emission side first electrostatic deflector and the second electromagnetic lens, and the length in the optical axis direction is the mask emission side. A second correction electrostatic deflector that is shorter than either the first electrostatic deflector or the mask emission-side second electrostatic deflector, wherein the first correction electrostatic deflector is the first crossover image. Due to the misalignment of
The second correction electrostatic deflector is for correcting the deviation of the trajectory of the charged particle beam emitted from the electrostatic entrance deflector on the mask entrance side, and the second correction electrostatic deflector detects the position deviation of the second crossover image. 13. The charged particle beam exposure apparatus according to claim 12, wherein the charged particle beam exposure apparatus is for correcting the deviation of the trajectory of the charged particle beam emitted from the second electrostatic deflector on the mask emission side. 14. The mask incident-side electrostatic deflector has a central axis aligned with the optical axis, and an optical deflector is provided on a side of the first electromagnetic lens between the first electromagnetic lens and the second electromagnetic lens. A mask entrance side first for arranging the first crossover image between an axial center point and the first electromagnetic lens for deflecting the charged particle beam incident through the optical axis. An electrostatic deflector, the central axis of which is aligned with the optical axis, and the electrostatic deflector is disposed on the second electromagnetic lens side between the first electromagnetic lens and the second electromagnetic lens,
A mask incident side second electrostatic deflector for deflecting a line extending the emitted charged particle beam to the opposite side in the traveling direction so as to pass substantially through the position of the first crossover image, and 12. The charged particle beam exposure apparatus according to claim 11, wherein the output end has an amplifier circuit commonly connected to the mask entrance side first electrostatic deflector and the mask entrance side second electrostatic deflector. . 15. A center axis is made to coincide with the optical axis so as to be between the mask incident side first electrostatic deflector and the mask incident side second electrostatic deflector or the mask incident side second electrostatic deflector. Disposed between the second electromagnetic electrostatic deflector and the second electromagnetic electrostatic deflector, and has a length in the optical axis direction larger than that of either the mask incident side first electrostatic deflector or the mask incident side second electrostatic deflector. A short first correction electrostatic deflector, with its central axis aligned with the optical axis, between the mask emission side first electrostatic deflector and the mask emission side second electrostatic deflector, or the mask emission side It is arranged between the first electrostatic deflector and the second electromagnetic lens and has a length in the optical axis direction larger than that of either the mask emitting side first electrostatic deflector or the mask emitting side second electrostatic deflector. A short second correction electrostatic deflector, wherein the first correction electrostatic deflector is caused by the position shift of the first crossover image,
This is for correcting the deviation of the trajectory of the charged particle beam emitted from the second electrostatic deflector on the mask entrance side.
The correction electrostatic deflector is for correcting the deviation of the trajectory of the charged particle beam emitted from the mask emission side second electrostatic deflector due to the positional deviation of the second crossover image. 15. The charged particle beam exposure apparatus according to claim 14. 16. An adjusting means for rotating the mask emitting side first electrostatic deflector or the mask emitting side second electrostatic deflector around the optical axis to adjust a rotation angle thereof. The charged particle beam exposure apparatus according to any one of claims 11 to 15. 17. An adjusting means for rotating the mask incident side first electrostatic deflector or the mask incident side second electrostatic deflector around the optical axis to adjust a rotation angle thereof. The charged particle beam exposure apparatus according to claim 14 or 15. 18. The charged particle beam exposure according to claim 11, wherein the output end of the amplification circuit is also commonly connected to the mask incident side electrostatic deflector. apparatus. 19. The amplifier circuit according to claim 11 and claim 1.
16. The charged particle beam exposure apparatus according to claim 14 or 15, wherein the amplifier circuit according to claim 4 is the same as the amplifier circuit. 20. For each transmission hole pattern on the stencil mask, a correction value pattern for maximizing a transient charged particle beam irradiation amount at the time of irradiation of the charged particle beam on the exposure object is stored. Storage means, a digital / analog converter, an amplifier circuit for amplifying the output of the digital / analog converter and applying it to the mask incident side electrostatic deflector, and for selecting a transmission hole pattern on the stencil mask Data read control means for reading a correction value pattern corresponding to the data from the storage means each time the data supplied to the digital / analog converter changes, and a value read from the storage means. A drive voltage is applied to the first correction electrostatic deflector based on the sum with a correction value when the irradiation of the charged particle beam onto the exposure object is in a steady state. The charged particle beam exposure device according to claim 11, characterized in that it comprises a first correction electrostatic deflector drive means. 21. The storage means stores, for each transmission hole pattern on the stencil mask,
A first memory in which a pattern of a transient charged particle beam irradiation amount upon irradiation of the charged particle beam onto the exposure object is stored, and the charged particle beam irradiation amount for each transmission hole pattern on the stencil mask. A second correction value stored in the first memory and addressed by the output data of the first memory.
21. The charged particle beam exposure apparatus according to claim 20, further comprising a memory.