JPH06177023A - Adjusting method for charged particle beam aligner - Google Patents

Abstract

PURPOSE:To shorten adjustment time of a charged particle beam aligner using a charged particle beam. CONSTITUTION:Deflection phase and quantity of the first and second mask deflectors 13 and 14 are obtained, and based upon this, the deflection difference and quantity ratio that make deflection efficiency of crossover image zero, when the deflectors 13 and 14 are driven at the same time, are obtained. Then deflection phase and quantity of the third and fourth mask deflectors 15 and 16 are obtained, and based upon this, the deflection difference and quantity ratio that make deflection efficiency of crossover image zero, when the deflectors 15 and 16 are driven at the same time, are obtained. Then, the first- fourth mask deflectors 13-16 are driven so that charged particle beam 3 is made incident, as vertically as possible, to a round aperture 20, and at the same time, without changing the said deflection phase difference and quantity ratio, relative angle and intensity between the first and second mask deflectors 13 and 14 and the third and fourth mask deflectors 15 and 16 are obtained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for adjusting a charged particle beam exposure apparatus using a so-called charged particle beam such as an ion beam or an electron beam. More specifically, the exposure processing is performed using a transmission mask plate (block mask). The present invention relates to a method of adjusting a charged particle beam exposure apparatus.

In recent years, with the increase in the density of integrated circuits, a new exposure method using charged particle beams, particularly electron beam exposure or X-rays has been investigated in place of photolithography, which has been the mainstream for fine pattern formation for many years, and has actually been studied. Has come to be used for.

The electron beam exposure is characterized in that it can form a pattern using an electron beam to form a fine pattern of about a micron or less. However, since the electron beam exposure method is a so-called "one-stroke writing" exposure apparatus, its processing capacity is limited.

Therefore, in recent years, a method has been proposed in which a pattern for forming a variable rectangle such as a rectangle, a square, or a triangle is prepared in advance, and an electron beam is selected and irradiated / exposed when necessary. For example, Japanese Patent Application No. 52-119185 proposes exposure using a mask in which patterns are arranged in blocks. In addition, JP-A-62-260
No. 322 proposes a method in which a repetitive pattern required for the shape of a memory cell or the like and a quadrilateral opening for a general-purpose rectangular pattern are provided on a diaphragm plate for exposure. This exposure method is effective for exposing a memory type in which the repeated cell pattern occupies most of the exposure area. In this case, in order to perform the exposure efficiently, it is necessary to have a plurality of pattern groups on the transmission mask. Then, when performing exposure, the electron beam is changed / selected for each pattern.

Under the circumstances, the present applicant has developed a so-called block exposure method in which about 20 to 40 blocks are arranged on a transmission mask and an arbitrary position is selected by a mask deflector for exposure.

[0006]

2. Description of the Related Art Conventionally, there is known a charged particle beam exposure apparatus, for example, an electron beam exposure apparatus whose conceptual diagram is shown in FIG.

In the drawing, 1 is an electron gun, 2 is an optical axis, 3 is an electron beam emitted from the electron gun 1, and 4 is an electron beam 3 emitted from the electron gun 1 for shaping the cross-sectional shape of the electron beam into a rectangular shape. A rectangular shaping aperture 5 is an electromagnetic lens for focusing the electron beam 3 rectangularly shaped by the rectangular shaping aperture 4.

Denoted at 6 is a deflector for variable rectangular shaping used for variable-rectangular shaping of the electron beam 3 subjected to rectangular shaping, and 7 is a parallel beam of the electron beam 3 having passed through the deflector 6 for variable rectangular shaping. It is an electromagnetic lens for converting.

A block mask 8 is formed by forming a plurality of block patterns. The block mask 8 is constructed, for example, as shown in the plan view of FIG.

In FIG. 35, 9 is a substrate, 10 is a region called an area arranged in a matrix at a predetermined pitch, and 11 is a region called a block arranged in a matrix in the area. A block pattern is formed as a unit.

The area 10 is sized so that the electron beam 3 can be deflected to the maximum on the block mask 8. As shown in FIG.
Six blocks 11 are set. 37 is a plan view showing an example of a block pattern, and 12 is a block pattern.

The block mask 8 has a mask area 120 of a calibration mask pattern in one area. As shown in FIG. 25, the mask area 120 has a configuration in which mask patterns having rectangular openings are arranged in a grid pattern at a predetermined pitch.

Further, in FIG. 34, reference numerals 13 to 16 designate the case where the cross-sectional shape of the electron beam 3 which is collimated by the electromagnetic lens 7 is shaped into the shape of the block pattern 12 formed on the block mask 8, that is, block shaping. This is a so-called mask deflector used for the case.

The mask deflector 13 is a mask deflector for deflecting the electron beam 3 in the direction of the selected block pattern. Further, the mask deflector 14 is
This is a mask deflector for deflecting the electron beam 3 deflected by the mask deflector 13 in the direction perpendicular to the block mask 8 on the selected block pattern.

Further, the mask deflector 15 is a mask deflector for deflecting the electron beam 3 block-shaped by the selected block pattern in the direction of the optical axis 2 in the non-deflected state. Further, the mask deflector 16 is a deflector for returning the electron beam 3 deflected by the mask deflector 15 onto the optical axis 2 in the non-deflected state.

The mask deflector 13 and the mask deflector 1
6 and the mask deflector 14 and the mask deflector 15 are arranged at mirror-symmetrical positions with the block mask 8 in between.

Numeral 17 is an electromagnetic lens for focusing the electron beam 3 which is converted into a parallel beam by the electromagnetic lens 7. The electromagnetic lens 17 and the electromagnetic lens 7 are mirror-symmetrical with the block mask 8 in between. It is arranged.

Further, 18 is a blanking electrode for controlling passage of the electron beam 3, 19 is a reduction lens for reducing the electron beam image, 20 is a diaphragm aperture 21, 21.
Reference numeral 22 is an image forming lens for forming an electron beam image on the sample.

Reference numeral 23 is a main deflector for deflecting the electron beam 3 to a sub-deflection area, and 24 is a sub-deflector for deflecting the electron beam 3 to an exposure position in the sub-deflection area. Deflector), 25 is a sample, and 26 is a sample stand for mounting the sample 25.

The control unit 50 shown in FIG. 38 controls the exposure unit shown in FIG.

The control unit 50 stores a storage medium 51 in which design data of the integrated circuit device is stored, a CPU 52 for controlling the entire charged particle beam, for example, drawing information fetched by the CPU 52, and a wafer W on which the pattern is to be drawn. Interface 53 for transferring various kinds of information such as the drawing information and mask information of the block mask 8, a data memory 54 for holding the drawing pattern information and the mask information transferred from the interface 53, and according to the drawing pattern information and the mask information, for example, 1 of the transparent hole of the transparent mask
Generating mask irradiation position data P1 to P4 indicating the position of the specified transmission hole on the transmission mask, and generating wafer exposure position data S3 indicating the position on the wafer where the pattern is exposed, In addition, a pattern generating unit 55 as a designating unit, a holding unit, a computing unit, and an output unit that performs various processes including a process of computing a correction value H according to the shape difference between the pattern shape to be drawn and the designated transmission hole shape. , an amplifier 56 for generating a corrected deflection signal HS1 from the correction value H, the voltage value from the data P ₁ ~P ₄ PS1~PS
It has a DAC / AMP unit 57 for generating 4 and a mask moving mechanism 58 for moving the block mask 8 as necessary.

Further, a clock control circuit 59 for generating a system clock and a blanking clock for operating the entire exposure apparatus in response to the exposure time / exposure waiting time from the pattern generator 55, and a clock control circuit 59. It has a blanking control circuit 60 which receives an output and generates a blanking timing, and an amplifier section 61 which generates a blanking signal SB.

Further, based on the exposure start / end information transferred from the interface 53, the main differential deflection information is output to the data memory 54 through the pattern generator and the desired stage is given to the stage controller. Command to move to the position, control the stage correction unit so as to correct the difference between the stage movement position and the main differential deflection, and cause the clock control unit to generate / stop the clock in the pattern generation unit. , A deflection controller 63 for generating a main differential deflection signal S2 based on the main differential deflection information from the data memory 54, a pattern generation section 55, and a deflection control. Amplifier units 64 and 65 that generate exposure position determination signals S2 and S3 based on the output from the circuit 63; Receiving a stage control mechanism 66 for moving the stage in accordance with the requirements, a laser interferometer 67 for detecting the stage position, the Meindefu deflection amount from the deflection control circuit, and a stage moving position from the stage control unit,
Stage correction unit 68 for correcting the difference from the main differential deflection
And are equipped with.

The electron beam emitted from the electron gun 1 is shaped into a rectangular shape by the first slit, and then the lenses 5 and 7 are formed.
It is converged by and is irradiated on the transmission mask. Deflection in a relatively large range (within about 5 mm) by the transmission mask is performed by the first to fourth deflectors 13 to 16, and a deflection in a relatively small range after being selected by the deflectors (within about 500 μm). ) Is performed by the slit deflector 6. In the case of variable rectangular exposure, the slit deflector 6 is used to shape into an arbitrary shape size (for example, an arbitrary rectangular size of 3 μm □ or less).

The beam transmitted through the transmission mask 8 passes through the blanking 18. And it is reduced by the 4th lens,
Then, the light is deflected by the sub deflector 24 in a small deflection area of about 100 μm. Also, the main differential coil 23
As a result, the sub-deflector deflection area is largely deflected in the exposure field in the range of about 2 mm.

The data to be exposed is read from the storage medium 51 by the CPU 52 and stored in the data memory 54. When exposure is started by a signal to the sequence controller 62, the stored main diff deflection position is first sent to the deflection control circuit 63, the deflection amount data S2 is output, and the main diff coil 23 is sent via the DAC / AMP 64.
Is output to. Then, after the output value is stabilized, the sequence controller commands the clock control circuit to generate the system clock, and as a result, the data memory 54 sends the stored pattern data to the pattern generating section. The pattern generation section generates shot data based on the read pattern data.

More specifically, this shot data shows the signals P1 to P4 indicating the beam irradiation position on the mask, the H signal indicating the deflection amount of the beam irradiation position on the mask, and the beam passing through these masks. Shows the S3 signal for deflecting the beam shaped in step 3 to the desired position on the wafer, shot time data, and how long it will take to wait for the electrostatic and electromagnetic deflectors to settle when these signals are applied. It consists of shot waiting time data. These signals enter the pattern correction unit after being generated by the pattern generation unit. The pattern correction unit corrects a wafer rotation or the like that occurs when the exposure sample W is set on the stage. The output signals for generating shots are D
It is input to an AC (digital converter), converted into analog, and applied to the electrodes and coils through an amplifier provided as necessary.

Such an electron beam exposure apparatus is a DRAM
When manufacturing an LSI having a repeating pattern such as (dynamic random access memory), the electron beam 3
Since the block exposure can be performed by shaping the cross-sectional shape of the above into a repetitive pattern, the exposure time can be shortened.

Here, in order to perform block exposure with high accuracy, the electron beam 3 is accurately deflected to the position of the selected block pattern, and the deflected electron beam 3 is again deflected to the optical axis in the non-deflected state. Accurately turn back to Aperture 2
It must be controlled to pass through the aperture 20 without being shaded by zero.

In order to perform such control, the relationship between the intensity of the deflection signal to be applied to the mask deflectors 13 to 16 and the position (coordinates) on the block mask 8 is determined. It is necessary to check in advance and determine the intensity of the deflection signal to be applied to each of the mask deflectors 13 to 16 according to the position of the selected block pattern.

FIG. 39 shows an example of the trajectory of the electron beam 3 from the mask deflector 13 to the aperture 20, in which 12A is a selected block pattern.
x _m and y _m are two-dimensional orthogonal axes on the block mask 8.

Further, the shape 28 such as the "floating device" shown by the broken line is such that when the electron beam 3 is deflected at a position equidistant from the intersection 27 of the optical axis 2A and the block mask 8 in the non-deflected state. This shows that the trajectory of the electron beam 3 is on the surface of the shape 28.

FIG. 40 shows the electron beam 3 shown in FIG.
Is a view of the orbit of FIG. 6 when viewed from above the block mask 8, in which the arrow 29 indicates the deflection direction of the mask deflector 13 and θ.
1 indicates the deflection direction of the mask deflector 13 by an angle formed by the x _m axis, and arrow 30 indicates the deflection direction of the mask deflector 14.
θ2 indicates the deflection direction of the mask deflector 14 as an angle with the x _m axis.

Further, arrow 31 indicates the deflection direction of the mask deflector 15, θ3 indicates the deflection direction of the mask deflector 15 by the angle formed by the x _m axis, arrow 32 indicates the deflection direction of the mask deflector 16, and θ4 indicates. The deflection direction of the mask deflector 16 is shown by an angle formed by the x _m axis.

FIG. 41 is a diagram showing changes in the trajectory of the electron beam 3 when the deflection direction θ1 of the mask deflector 13 is changed. When θ1 is increased, the trajectory of the electron beam 3 is changed to the solid line 33. Changes to a broken line 34.

FIG. 42 is a diagram showing changes in the trajectory of the electron beam 3 when the deflection direction θ2 of the mask deflector 14 is changed. When θ2 is increased, the trajectory of the electron beam 3 is changed from the solid line 35. It changes like a broken line 36.

FIG. 43 is a diagram showing changes in the trajectory of the electron beam 3 when the deflection directions θ3 and θ4 of the mask deflectors 15 and 16 are changed. When θ3 is increased, the trajectory of the electron beam 3 changes. , And changes from the solid line 37 to the broken line 38, and when θ4 is increased, the trajectory of the electron beam 3 changes from the solid line 39 to the broken line 40.

FIG. 44 is a diagram showing changes in the plane on which the trajectory of the electron beam 3 exists when the deflection intensity G1 of the mask deflector 13 is changed. When G1 is increased, the trajectory of the electron beam 3 changes. The existing surface changes from the shape surface indicated by the solid line 41 to the shape surface indicated by the alternate long and short dash line 42.

FIG. 45 is a diagram showing changes in the plane on which the trajectory of the electron beam 3 exists when the deflection intensity G2 of the mask deflector 14 is changed. When G2 is increased, the trajectory of the electron beam 3 changes. The existing surface changes from the shape surface indicated by the solid line 43 to the shape surface indicated by the alternate long and short dash line 44.

Further, FIG. 46 shows a change in the surface on which the trajectory of the electron beam 3 exists and the deflection intensity G4 of the mask deflector 16 when the deflection intensity G3 of the mask deflector 15 is changed.
FIG. 9 is a diagram showing a plane on which the orbit of the electron beam 3 exists when V is changed. When G3 is increased, the plane on which the orbit of the electron beam 3 is present is changed from the shape indicated by the solid line 45 to the shape indicated by a dashed line 46 By changing to a surface and increasing G4, the trajectory of the electron beam 3 exists on the surface of the shape indicated by the alternate long and short dash line 47.

Here, in the intensity of the deflection signal to be applied to the mask deflectors 13 to 16, the component in the X _m axis direction is the BSX.
If the i, y _m axis direction components are BSYi, for example, it is sufficient to consider only the offset term, the gain term, the low tension term, and the orthogonality correction term, and if the higher order terms are not considered, the mask deflection Deflection signal intensities BSXi and BSYi to be applied to the detectors 13 to 16 and the deflected electron beam 3
The relationship between the position (x _m , y _m ) on the block mask 8 which should pass can be expressed as shown in Equation 1.

[0042]

[Equation 1]

That is, when the deflection signal intensities BSXi and BSYi are determined according to the relationship represented by the equation (1), the electron beam 3 is accurately deflected to the position of the selected block pattern, and then the non-deflection is performed. It can be swung back to the optical axis and passed through the aperture 20 without being obscured by the aperture 20.

In the equation (1), the matrix on the right side and the matrix on the left side are called the correction coefficient matrix,
O _Xi and O _Yi are correction factors for offset terms, G _Xi and C _Yi are correction factors for gain terms, R _Xi and R _Yi are correction factors for low tension terms, and H _Xi and H _Yi are correction factors for orthogonality correction terms. Is.

Further, although i takes a value of 1 to 4, when i = 1, the equation (1) shows that the intensity BSX1, BSY1 of the deflection signal to be applied to the mask deflector 13 and the electron beam 3
Position on the block mask 8 to be deflected (x _m , y _m )
Will show the relationship between and.

When i = 2, the equation (1) is
Deflection signal strength BSX to be applied to the mask deflector 14
2, a BSY2, position on the block mask 8 should deflect the electron beam 3 (x _m, y _m) will indicate the relationship between the.

When i = 3, the equation (1) is
Deflection signal strength BSX to be applied to the mask deflector 15
3, and BSY3, position on the block mask 8 should deflect the electron beam 3 (x _m, y _m) will indicate the relationship between the.

When i = 4, the equation (1) is
Deflection signal strength BSX to be applied to the mask deflector 16
4, a BSY4, position on the block mask 8 should deflect the electron beam 3 (x _m, y _m) will indicate the relationship between the.

Therefore, if the correction coefficient matrix is obtained for each of the mask deflectors 13 to 16, the electron beam 3
After being accurately deflected to the position of the selected block pattern, it can be swung back to the optical axis in the non-deflected state and passed through the aperture 20 without being obscured by the aperture 20.

When obtaining the correction coefficient matrix, the electron beam 3 is deflected to a proper position on the plane where the block mask 8 is originally arranged without using the block mask 8 as the first step. Then, again considering the case where the electron beam 3 is swung back to the optical axis in the non-deflected state and passed through the aperture 20, the deflection condition (deflection intensity and deflection direction) in this case needs to be obtained.

This is carried out by simulation. If the deflection condition is not appropriate, the electron beam 3 will be shaded by the aperture 20, so whether the deflection condition is appropriate depends on the sample current value, specifically, the sample current value. Is sample 2
It can be determined by whether or not the output current value of the Faraday cup placed in place of 5 is the maximum value. Therefore, in this simulation, the procedure for searching for the deflection condition that maximizes the sample current value is executed.

In this case, since the mask deflector 13 and the mask deflector 16 and the mask deflector 14 and the mask deflector 15 are respectively arranged at mirror-symmetrical positions with the block mask 8 in between, θ1 = θ4. θ2 = θ3, G1 = G
4, it is appropriate to set the condition of G2 = G3.

Here, in FIGS. 47 and 48, the electron beam 3 is originally deflected to an appropriate position on the plane where the block mask 8 is arranged without using the block mask 8, and again,
FIG. 49 and FIG. 50 show a subroutine for obtaining a deflection condition when the electron beam 3 is returned to the optical axis in the non-deflected state and passed through the aperture 20.

In this simulation, it is premised that the position on the block mask 8 before calibration is displayed as Xm and Ym as deflection parameters. First, as shown in FIG. = Ym =
0, that is, the electron beam 3 is directed to the mask deflector 13
At ~ 16, the sample current value when passing through the aperture 20 in the undeflected state is measured and set to I ₀ (step P1).

Next, initial setting is performed. This initial setting is Xm = ΔXm, Ym = ΔYm, θ1 _min = θ2
_min = 0, θ1 _max = θ2 _max = 2π, G2 _min = 0,
G2 _{ma x} = {[BSX2 _max) ² + (BSY
2 _max ) ² ] / (Xm ² + Ym ² )} ^1/2 (step P2).

Here, ΔXm is a value smaller than the size of the electron beam 3 on the block mask 8 in the X direction, and ΔYm is a value smaller than the size of the electron beam 3 on the block mask 8 in the Y direction.

Next, according to step P2, the positions Xm, Y
After the deflection position was changed to m, the sample current value was measured,
The value is set to I (step P3), and it is determined whether the absolute value of the difference between the current value I and the current value I ₀ is smaller than the predetermined convergence condition E (step P4). This convergence condition E
Is specified by the user, but is specified within the range of the accuracy of the ammeter that measures the sample current value.

Here, when the absolute value of the difference between the current value I and the current value I ₀ is larger than the convergence condition E (N in step P4).
In the case of O), I _max = I, Δθ1 = (θ1 _max −θ1
_min ) / N, ΔG2 = (G2 _max −G2 _min ) / N, Δ
θ2 = (θ2 _max −θ2 _min ) / N, and the subroutines shown in FIGS. 23 and 24 are executed. Note that N is a value of 4 or more designated by the user.

In the subroutine, Xm, Ym, θ
1_min, Θ1_max, G2_min, G2 _max, Θ2_min, Θ
Two_max, Δθ1, ΔG2, Δθ2, I_maxIs entered
(Step N1), G1 = G4 = 1.0, G2 = G3 =
G2_min, Θ1 = θ4 = θ1_{mi n}, Θ2 = θ3 = θ2
_minAs a result, initial setting is performed (step N2). G
The fixed values of 1 = G4 = 1.0 are G1, G2, and G.
For 3 and G4, it is enough to know the relative values.
Because.

Next, the deflection condition (θ
1, G2, θ2) is changed, the sample current value is measured,
The value is set to I (step N3). Then, it is determined whether or not this current value I is larger than the current value I _max stored in step P5 of the main routine shown in FIG. 21 (step N4).

As a result, if the current value I is larger than the current value I _max stored in step P5 of the main routine shown in FIG. 47 (YES in step N4),
Since a more appropriate deflection condition is obtained, I _max = I, θ1 _out = θ1, G2 _out = G2, θ as data I _max , θ1 _out , G2 _out , and θ2 _out to be output.
2 _out = θ2 is stored (step N5).

On the other hand, if the current value I measured in step N3 is smaller than the current value I _max stored in step P5 of the main routine shown in FIG. 47 (step N).
(NO in step 4) and when step N5 ends, G1 = G4 = 1.0, G2 = G3 = G2 _min , θ2
= Θ3 = θ2 _min , in order to search a better deflection condition for θ1 and θ4, θ1 ← θ1 + Δθ1, θ4
= Θ1 (FIG. 24, step N6).

Subsequently, it is judged whether or not θ1 is larger than θ1 _max (step N7), and while θ1 is smaller than θ _max (NO in step N7), step N3 is executed.
If N1 is larger than θ1 _max due to the repetition of the processes up to N6 (YES in step N7), G1 = G
4 = 1.0, θ2 = θ3 = θ2 _min , θ1, θ
4, θ2 is varied to maximize the sample current value θ1,
To search θ4 and G2, θ1 = θ4 = θ1 _min ,
G2 ← G2 + ΔG2, G3 = G2 (step N
8).

Subsequently, it is judged whether or not G2 is larger than G2 _max (step N9), and while G2 is smaller than G2 _max (NO in step N9), step N is executed.
3 to N8 are repeated and G2 becomes larger than G2 _max (YES in step N9), G1 =
Under G4 = 1.0, in order to search θ1, G2, θ2 where the sample current value becomes maximum by varying θ1, G2, θ2, G2 = G3 = G2 _min , θ2 ← θ2 + Δθ2, θ
3 = θ2 (step N10).

Subsequently, it is judged whether or not θ2 is larger than θ2 _max (step N11), and while θ2 is smaller than θ2 _max (NO in step N11), steps N3 to N10 are repeated, When θ2 becomes larger than θ2 _max (YES in step N11), θ1 _out , G2 _out , and θ stored in step N5 are stored.
2 _out and I _max are returned to the main routine, and the subroutine ends (step N12).

Therefore, in the main routine, I _max , θ1 _out , G2 _out , and θ2 _out returned from the subroutine.
Based on I = I _max , θ1 _max = θ1 _out + Δθ
1, θ1 _min = θ1 _out −Δθ1, G2 _max = G2 _out
+ ΔG2, G2 _min = G2 _out −ΔG2, θ2 _max = θ
2 _out + Δθ2, θ2 _min = θ2 _out −Δθ2,
Returning to step P3, steps P3 to P6 are repeated until Xm and Ym reach the maximum deflection range, and the optimum deflection condition is searched while gradually narrowing the search range of the deflection condition.

When Xm and Ym exceed the maximum deflection range (YES in step P8), the latest θ1 stored in step N5 of the subroutine shown in FIG.
_out , G2 _out , and θ2 _out are the optimum deflection conditions, and θ1
= Θ4 = θ1 _out , G2 = G3 = G2 _out , θ2 = θ3
= Θ2 _out and the process ends (step P9).

Here, when the deflection intensities G1 and G4 of the mask deflectors 13 and 16 are set to 1.0 and the electron beam is deflected without using a block mask and returned to the optical axis in the non-deflected state. Deflection strengths G2 and G3 of the mask deflectors 14 and 15
And the deflection direction θ of the mask deflectors 13, 14, 15, 16
1, θ2, θ3, θ4 can be obtained.

Then, next, in order to obtain the relationship between the deflection parameters Xm and Ym and the intensities BSXi and BSYi of the deflection signals to be applied to the mask deflectors 13 to 16, the deflection intensities G1 to G1 of the mask deflectors 13 to 16 are calculated. G4 and deflection direction θ1
~ Θ4 is used to perform the calculations of Expressions (2) to (17) to obtain the correction coefficients G _xi 'and G _Yi ' of each gain term and the correction coefficients R _xi 'and R _Yi ' of the low tension term. Ask.

G_xi′ = G1 × cos θ1 (2) R_xi′ = G1 × sin θ1 (3) G_Yi′ = G1 × cos θ1 (4) R_Yi′ = −G1 × sin θ1 (5) G_x2′ = −G2 × cos θ2 (6) R_x2′ = −G2 × sin θ2 (7) G_Y2′ = −G2 × cos θ2 (8) R_Y2′ = G2 × sin θ2 (9) G_X3′ = −G3 × cos θ3 (10) R_x3′ = G3 × sin θ3 (11) G_Y3′ = −G3 × cos θ3 (12) R_Y3′ = −G3 × sin θ3 (13) G_x4’= G4 × cos θ4 (14) R_x4′ = −G4 × sin θ4 (15) G_Y4’= G4 × cos θ4 (16) R_Y4′ = G4 × sin θ4 (17) However, in this case, the correction coefficient for obtaining the offset term
O_Xi’, O_Yi'And a correction factor for obtaining the orthogonality correction term
Number H_Xi, H_YiIs O for convenience_Xi’= O_Yi’= H _Xi’= H
_Yi’= 0.

From the expressions (2) to (17), the intensity of the deflection signal to be applied to the mask deflectors 13 to 16 BSXi,
The relationship between BSYi and the deflection parameters Xm and Ym can be displayed as shown in Expression (18).

[0072]

[Equation 2]

Therefore, next, a block mask 52 having rectangular calibration patterns 48 to 51 for calibration as shown in FIG. 51 is prepared and calibration is performed. Note that 53 to 56 indicate the positions (reference points) of the transmission patterns 48 to 51.

Here, the position of the calibration transmission pattern k (k = 48 to 51) is x _m (k), y
_When displaying with _m (k), the deflection parameter X necessary for deflecting the electron beam 3 to the transmission pattern k
When m and Ym are represented by Xm (k) and Ym (k), the correction coefficient matrix shown in Expression (1) can be obtained by the relational expression shown in Expression (19). Note that BSXi (k) and BSYi (k) in equation (19) can be obtained by the relational expression shown in equation 20.

[0075]

[Equation 3]

In this way, the correction coefficient matrix shown in equation (1) is obtained for each mask deflector 13-16, and the intensity of the deflection signal to be applied to each mask deflector 13-16 according to equation (1). When determining, the electron beam 3 is accurately deflected to the position of the selected block pattern, and again, the deflected electron beam 3 is returned to the optical axis in the non-deflected state and passed through the aperture 20. be able to.

[0077]

The operation of the conventional calibration will now be described with reference to, for example, the transmission pattern 48. As shown in FIG. 52 (A), the electron beam 3 is thrown into the transmission pattern 48. Electron beam 3
The deflection parameters Xm and Ym are adjusted while determining how the electron beam 3 is transmitted, and the electron beam 3 is moved as shown in FIG.
As shown in FIG. 5, the reference point (origin) 57 of the electron beam 3 is deflected so as to overlap the reference point 53 of the transmission pattern 48, and the deflection parameters Xm and Ym at this time cause the electron beam 3 to be located at the position x of the transmission pattern 48. _{m (48), y m (} 4
8) Deflection parameter Xm (4
8) and Ym (48).

In this case, when the correction coefficient matrix shown in the equation (18) is close to the true correction coefficient matrix, Xm (k) = x _m (k) + Δx _m , Ym (k) = y
_{As m} (k) + Δy _m , by moving Δx _m and Δy _m in a narrow area, Xm (k) and Ym (k) can be quickly searched.

However, there is no guarantee that the correction coefficient matrix shown in equation (18) is close to the true correction coefficient matrix. This is because the Xm-axis and Ym-axis directions and the scale used in obtaining the correction coefficient matrix shown in the equation (18) have not been added yet, and this work is performed for the first time by calibration. .

Therefore, the work of deflecting the electron beam 3 to the calibration transmission patterns 48 to 51 cannot be performed easily and quickly, and the mask deflectors 13 to
There is a problem that the correction coefficient matrix for adjusting the intensities BSXi and BSYi of the deflection signal given to 16 cannot be obtained easily and quickly.

That is, ΔX _m, ΔY _m must be such an amount that the aperture passing current is not lost _, and in order to satisfy the predetermined maximum deflection range X _m, Y _m , the values shown in FIGS.
X _m / ΔX for all steps of the flowchart shown in
It has to be repeated _m times, resulting in a huge amount of calculation and a long adjustment time. Also, Δθ1 to Δθ4, ΔG
The same applies to 1 to ΔG4, and each value Δθ1 to Δθ
4, .DELTA.G1 to .DELTA.G4 must also be small enough not to lose sight of the aperture passing current, which also increases the calculation amount and lengthens the adjustment time.

On the other hand, G1 = G4, G2 = G3, θ1 = θ4, θ2 = θ3 used in the above-mentioned conventional adjusting method.
The adjustment assumptions such as the above are based on the assumption that the upper and lower lens optical systems and the deflection system arranged with the mask block sandwiched are mirror-symmetrical. When the symmetry is broken due to the occurrence of the above, the above assumption is no longer satisfied, and there is a problem that the optimum deflection cannot be obtained in the actual system.

Furthermore, adjusting the current passing through the aperture to the maximum does not satisfy the vertical incidence of the beam on the round aperture. Therefore, when a different pattern on the mask block is selected, the irradiation position on the sample is adjusted. There was a problem that a gap would occur.

In the above method, if the trajectory of the beam near the block mask is as simulated,
That is, it is effective when the effective arrangement of the mask deflector and the magnetic field generated by the lens are mirror-symmetric with respect to the block mask. However, in reality, this symmetry is not strictly satisfied due to the mechanical precision of the lens barrel fabrication and the interaction with other optical systems located below the block mask. Further, the offset amount of the pattern position generated when the block mask is moved to replace the block mask or to use a block pattern of a different mask area is not considered.

Therefore, the effective arrangement of the mask deflector,
If the adjustment is performed under the assumption that the magnetic field produced by the lens is mirror-symmetric with respect to the block mask, the correct adjustment result cannot be obtained, and even if the block pattern is selected with the mask deflector, the current loss at the aperture is large. In addition, there is a problem that the beam is displaced on the sample surface.

When the block mask is moved,
There is a problem that the output value of the mask deflector must be calculated again in order to compensate for the offset amount of the pattern position.

FIG. 53 shows the pattern generating means 55 of the conventional control section.

The pattern generating means 55 has a pattern generating section 80 and a mask memory 81.

The mask memory 81 stores the deflection data for each pattern consisting of the deflection signals BSXi and BSYi to the deflectors 13 to 16 which are calculated by the above-mentioned method. It is assigned to each pattern, and its address is calculated and referred to by the pattern data code.

The pattern generating section 80 has a data memory 54.
From the pattern data code included in the exposure data, the predetermined deflection data in the mask memory 81 is read.

The read deflection data is added to the amplifier unit 57, and the deflection signal from the amplifier 57 is output to each of the deflectors 13 and 1.
4,15,16.

Here, the intensity of the deflection signal BSXi, BSY
Reference is made to the above formula (1) representing i.

Matrix on the right side of equation (1)

[0094]

[Equation 4]

Is a correction coefficient matrix for representing the relationship between the position of the block mask 8 at that time and the deflection signal to be applied to each of the mask deflectors 13 to 16.

Therefore, by moving the block mask,
When the mask is positioned in another mask area on the block mask, it is necessary to recalculate all the above correction coefficients. The correction coefficient exists separately for each deflector, and the number thereof is huge, and it takes about half a long time to recalculate the correction coefficient.

If a block mask is used, then FIG.
As shown in step 83 _-1 to 83 _-4 4, sometimes the block mask in order to select a different mask area is moved, each time re-adjustment.

During the readjustment, the exposure position is stopped, and the stop time is about half a day.

As a result, it takes about half a day to readjust each time the block mask is moved,
The exposure apparatus is stopped for a long time, the operation efficiency of the exposure apparatus is stopped, and as a result, the throughput of the semiconductor device is reduced.

The present invention has been made in view of the above points, and adjustment of a mask deflector of a charged particle beam exposure apparatus configured to perform block exposure can be performed simply and quickly. An object of the present invention is to provide a method for adjusting a charged particle beam exposure apparatus.

[0101]

Means for Solving the Problems The above-mentioned problems are solved by a charged particle beam generating means for generating a charged particle beam and a polygonal shaping plate which is arranged at a position where the charged particle beam passes and which shapes the charged particle beam. And a transparent hole mask plate having at least one transparent hole pattern having an arbitrary shape,
When the charged particle beam generating means is arranged at the upper part, the charged particle beam is deflected so that the charged particle beam is deflected in pairs one above the other with the mask of the transmission hole interposed therebetween.
In a method for adjusting a charged particle beam exposure apparatus, which comprises: a through fourth deflecting means and a round aperture disposed under the first through fourth deflecting means, the charged particle beam exposure apparatus is arranged at a position closest to the polygon shaping plate. A first step of driving only the provided first deflecting means to obtain a deflection phase and a deflection amount of the crossover image on the round aperture by the first deflecting means, and a position lower than the first deflecting means. In the second step, only the second deflecting means arranged above the transparent hole mask plate is driven to determine the deflection phase and the deflection amount of the crossover image on the round aperture by the second deflecting means. And when the first and second deflecting means are simultaneously driven, the deflection difference between the first and second deflecting means is set so that the deflection efficiency of the crossover image on the round aperture becomes zero. The third step of obtaining the deflection amount ratio and the driving of only the third deflecting means arranged at the lower position of the transmission hole mask plate, and the deflection of the crossover image on the round aperture by the third deflecting means. The fourth step of obtaining the phase and the deflection amount, and driving only the fourth deflecting means arranged at the lower position of the third deflecting means, and the crossover image on the round aperture by the fourth deflecting means. 5th step of obtaining the deflection phase and the deflection amount of 3rd and 4th so as to make the deflection efficiency of the crossover image on the round aperture zero when the 3rd and 4th deflection means are simultaneously driven. The sixth step of obtaining the deflection difference and the deflection amount ratio of the deflecting means, the first to fourth deflecting means are driven, the charged particle beam is made to enter the round aperture most vertically, and the deflection phase difference and The seventh step of obtaining the relative angle between the first and second deflecting means and the third and fourth deflecting means without changing the direction ratio, and the first to fourth steps after obtaining the relative angle. The deflecting means is driven so that the charged particle beam enters the round aperture most vertically, and the first and second deflecting means and the third and fourth deflecting means are maintained without changing the deflection phase difference and the deflection amount ratio. And the relative phase difference of the first to fourth deflecting means are used to match the deflection amount and the deflection direction of the deflection coordinate on the transmission hole mask plate, and the transmission is performed. This can be solved by a method for adjusting a charged particle beam exposure apparatus, which comprises a ninth step of selecting a transmission hole pattern on a hole mask plate.

[0102]

In the above adjusting method, the first step is the first step.
The deflection phase and the deflection amount of the deflection means are obtained, and the deflection phase and the deflection amount of the second deflection means are obtained in the second step. The deflection phase and the deflection amount of the deflection means obtained in each step are defined as the deviation amount from the normal beam passage position on the round aperture, and are the deviation amounts including error factors such as an assembly error of the apparatus.

Therefore, by adjusting the first and second deflecting means so that the charged particle beam comes to the normal beam passage position on the round aperture, it becomes possible to collectively correct the error factors. .

In the present invention, the first and second steps are performed in the third step.
Are simultaneously driven, and the deflection phase and the deflection amount of the first deflection unit obtained in the first step and the deflection phase and the deflection amount of the second deflection unit obtained in the second step are mutually By adjusting the first and second deflecting means so as to cancel each other (ie, the deflection efficiency of the crossover image on the round aperture is set to zero), the charged particle beam is moved to the normal beam passing position on the round aperture. It is configured to adjust to come to.

As described above, by performing the first to third steps, how to operate the first and second deflecting means under the condition that the third and fourth deflecting means are not driven. Even if it is driven, the charged particle beam is always irradiated onto the regular beam passing position on the round aperture.

In the subsequent fourth step, the deflection phase and the deflection amount of the fourth deflecting means are obtained, and in the fifth step, the deflection phase and the deflection amount of the fifth deflecting means are obtained. The deflection phase and the deflection amount of the deflection means obtained in each step are also defined as the deviation amount from the normal beam passing position on the round aperture, and are the deviation amounts including error factors such as an assembly error of the apparatus.

Therefore, similarly to the adjustment of the first and second deflecting means described above, the third and fourth deflecting means are adjusted so that the charged particle beam comes to the normal beam passing position on the round aperture. Therefore, it becomes possible to collectively correct the error factors.

According to the present invention, in the sixth step, the deflection phase and the deflection amount of the third deflecting means obtained in the fourth step, and the deflection phase and the deflection amount of the fourth deflecting means obtained in the fifth step. But the third and fourth are offset by each other
By adjusting the deflecting means of (i.e., the deflection efficiency of the crossover image on the round aperture is set to zero), the charged particle beam is adjusted to come to the normal beam passing position on the round aperture. There is.

As described above, by carrying out the fourth to sixth steps, no matter how the third and fourth deflecting means are driven, the charged particle beam will move to the normal beam passing position on the round aperture. It will always be illuminated.

In the state where the first to sixth steps are completed, the charged particle beam is always irradiated to the regular beam passing position on the round aperture regardless of how the first to fourth deflecting means are driven. Will be configured. However,
It is not on the round aperture that the charged particle beam exposure apparatus actually wants to expose the charged particle beam, but on the sample disposed below the round aperture.

Therefore, even if the charged particle beam is simply applied to the normal beam passing position on the round aperture, the charged particle beam is not always applied to the predetermined position on the sample. For example, the charged particle beam is applied to the normal position on the round aperture. The charged particle beam obliquely advancing to the beam passage position is irradiated onto a position on the sample that is largely deviated from a predetermined position.

In the seventh and eighth steps, the charged particle beam is made to enter the round aperture most vertically, and the deflection phase difference and the deflection amount ratio are not changed, and the third and the third deflection means. The charged particle beam is irradiated to a predetermined position on the sample by obtaining the relative angle and the relative intensity with respect to the fourth deflecting means.

When the first to eighth steps are completed, the charged particle beam emitted from the charged particle beam generating means is on the sample regardless of how the first to fourth deflecting means are driven. It is configured so that it is always irradiated at a predetermined position.

However, when the first to eighth steps are completed, the adjustment of the first to fourth deflecting means, the round aperture and the sample (the adjusted configurations are collectively referred to as an apparatus system) is completed. However, the adjustment of the device system for the transmission hole mask plate has not been performed. Therefore, in the ninth step, the deflection amount and the deflection direction of the deflection coordinates on the transmission hole mask plate are adjusted to match the coordinate system of the apparatus system.
As a result, it is possible to prevent an error from occurring when selecting a transmission hole pattern on the transmission hole mask plate, and it is possible to perform a highly accurate irradiation process of the charged particle beam.

[0115]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of the present invention will now be described with reference to the drawings.

FIG. 1 is a block diagram showing a charged particle beam exposure apparatus 30 (electron beam exposure apparatus) which is adjusted by the method of the present invention. In the figure, the same components as those of the charged particle beam exposure apparatus described with reference to FIG. 34 are designated by the same reference numerals and the description thereof will be omitted.

In the figure, 31 is an electromagnetic lens,
A round aperture 20 is provided for a charged particle beam (hereinafter, referred to as an electron beam) which is rectangularly shaped by the slit deflector 6.
It has a function of forming an image on the upper side. Therefore, the electron beam incident on the electromagnetic lens 31 is focused on a predetermined position according to the focal length of the electromagnetic lens 31. Further, the block mask 8 has a so-called lens-in-mask structure in which the block mask 8 is integrally incorporated in the electromagnetic lens 31. With this configuration, the charged particle beam exposure apparatus 30 can be downsized.

Next, a specific adjusting method of the charged particle beam exposure apparatus 30 having the above configuration will be described.

FIG. 2 is an enlarged view showing the vicinity of the block mask 8 in FIG. In the drawings and the drawings used for the following description, only the components related to the adjusting method of the present invention are shown, and the other components are not shown.

To adjust the charged particle beam exposure apparatus 30, first, as the first step, only the first mask deflector 13 is driven and the electron beam 1 is emitted from the electron gun 1. At this time, the voltage applied to the first mask deflector 13 is V1.

The electron beam 3 is deflected by driving the first mask deflector 13, and the electron beam 3 passes through a predetermined block pattern 12 formed on the block mask 8 and is swung back by the electromagnetic lens 31. And the round aperture 20 is illuminated. FIG. 3 is a view of the round aperture 20 as viewed from the side where the electron gun 1 is disposed (that is, the upper position).

For example, by driving only the first mask deflector 13, the electron beam 3 has a white circle in FIG.
It is assumed that the position indicated by the mark V1 is irradiated and a crossover image is formed. In the present embodiment, vector display is used to represent this irradiation position, and the magnitude of the vector V1 (hereinafter referred to as the deflection amount) is defined as the counterclockwise angle with respect to the r1 and X _R axes (hereinafter referred to as the deflection phase). Let θ1.

This vector V1 contains an error factor such as an assembly error of the charged particle beam exposure apparatus 30. If there is no error factor, the electron beam 3 is returned to the lens 31 and the round aperture 2
It should pass through the aperture hole 20a formed at the center position of 0 (corresponding to the origin in FIG. 3). Therefore, in order to perform highly accurate exposure processing by the charged particle beam exposure apparatus 30, it is necessary to adjust the apparatus 30 so that the deflection amount r1 becomes zero.

Then, as the second step, only the second mask deflector 14 is driven to emit the electron beam 3 from the electron gun 1. At this time, the voltage applied to the second mask deflector 14 is V2. The electron beam 3 is deflected by driving the second mask deflector 14, and as in the first step, the electron beam 3 passes through the predetermined block pattern 12 formed on the block mask 8 and the electromagnetic lens 31. Then, it is swung back and is irradiated onto the round aperture 20.

For example, by driving only the second mask deflector 14, the electron beam 3 is filled with black dots in FIG.
It is assumed that irradiation is performed on the position indicated by mark V2. When this irradiation position is displayed as a vector, the deflection amount of the vector V2 is r
2, the deflection phase can be expressed as θ2. Note that this vector V2 also includes error factors such as an assembly error of the charged particle beam exposure apparatus 30 like the vector V1 described above, so that the charged particle beam exposure apparatus 30 can perform high-precision exposure processing. Must adjust the device 30 so that the deflection amount r2 becomes zero.

In the following third step, the deflection amounts r1 and r
Perform processing to set 2 to zero. In the present invention, in order to set the deflection amounts r1 and r2 to zero, the vectors V1 and V2 are set.
The method of offsetting 2 is used. Below, vector V1,
A method of canceling V2 will be described with reference to FIG.

According to the first and second steps, the first or second step
Driving the mask deflectors 13 and 14 separately for the vector V
1 and V2 are obtained, in the third step, the first and second mask deflectors 13 and 14 are simultaneously driven, and the vector V
2 is moved to a position symmetrical about the origin with respect to the vector V1, and the magnitude of the vector V2 is changed to the vector V1.
The adjustment processing is performed so that the size becomes the same.

Specifically, the vector V2 is replaced by the vector V1.
To move to a symmetrical position around the origin
Angle difference (hereinafter, this angle difference is called the deflection phase difference)
The size of vector V1 and the size of vector V2
A deflection amount correction coefficient G1 for making the sizes the same is obtained.
The deflection amount correction coefficient G1 is obtained by G1 = r1 / r2 (21) and the vector is obtained by multiplying r2 by G1.
Make the magnitude of V1 and the magnitude of vector V2 the same
You can The deflection phase difference is shown by the arrow θ in FIG. _d1Indicated by
Is the value to be set.

When the deflection amount correction coefficient G1 and the deflection phase difference θ _d1 are obtained as described above, the second mask deflector 1 is made to correspond to the deflection amount correction coefficient G1 and the deflection phase difference θ _d1.
The voltage applied to 4 and the phase of the power supply are controlled. Second
Of the vector V by controlling the phase of the power source applied to the mask deflector 14 in accordance with the deflection phase difference θ _d1.
2 moves as indicated by the double-dot chain line arrow in FIG. 3B, and the voltage applied to the second mask deflector 14 is controlled so as to correspond to the deflection amount correction coefficient G1. And a vector V2 '(a vector shown by a chain line in the figure) symmetrical with respect to the origin is generated.

Now, let us say that the voltage applied to the second mask deflector 14 to generate the above-mentioned vector V2 'is V2', and in this case, the applied voltage V1 to the first mask deflector 13 and the second mask Applied voltage V2 'to the deflector 14
Is defined as the deflection amount ratio (V1: V2 ′).

Then, control is performed so as to correspond to the phase deflection phase difference θ _d1 of the power source applied to the second mask deflector 14,
Moreover, by controlling the deflection amount ratio so that V1: V2 'is always established, the vector V1 generated by the first mask deflector 13 and the vector V2' generated by the second mask deflector 14 cancel each other out. is, the origin of always combining vector of the vectors V1 and vector V2 'X _R -Y _R coordinate.

This is because the second mask deflector 14 is controlled based on the deflection phase difference θ _d1 and the deflection amount ratio (V1: V2 ′), so that the first and second mask deflectors 13,
No matter how the electron beam 14 is driven, the electron beam 3 always passes through the aperture hole 20a formed in the round aperture 20, that is, the first and second mask deflectors 13 and 14.
It is possible to set the deflection efficiency of the crossover image on the round aperture 20 to zero when both are simultaneously driven.

By the above series of adjustment processing, the first and second
The adjustment of the mask deflectors 13 and 14 has been completed. Subsequently, a method of adjusting the third and fourth mask deflectors 15 and 16 will be described.

In order to adjust the third and fourth mask deflectors 15 and 16, as the fourth step, only the third mask deflector 13 is driven and the electron beam 3 is emitted from the electron gun 1 ( FIG. 4 shows this state). At this time, the voltage applied to the third mask deflector 15 is V3.

Incidentally, the third and fourth mask deflectors 15 and 1
Although the first and second mask deflectors 13 and 14 are not driven in the present embodiment when the adjustment of No. 6 is performed, the charged particle beam exposure apparatus 30 to be adjusted in the present embodiment is an electromagnetic lens as described above. 31 is a device having a structure for providing a lens swing-back system in which 31 is arranged. Therefore, if the adjustments of the first to third steps are completed, the electron beam 3 is incident on the aperture hole 20a of the round aperture 20 no matter how the first and second mask deflectors 13 and 14 are driven. Therefore, it is possible to adjust the third and fourth mask deflectors 15 and 16 while driving the first and second mask deflectors 13 and 14. However, for convenience of description, in the following description, it is assumed that the first and second mask deflectors 13 and 14 are not driven.

Since the first and second mask deflectors 13 and 14 are not driven, the electron beam 3 emitted from the electron gun 1 is irradiated on the block mask 8 along the optical axis 2 to form a predetermined block pattern. It passes through and reaches the third mask deflector 15.

Since the third mask deflector 15 is driven as described above, the electron beam 3 is deflected by the third mask deflector 15 as indicated by a solid arrow A in FIG. 4, for example. . In the fourth step, since the fourth mask deflector 16 is not driven, the electron beam 3 travels as shown by the broken line arrow B and irradiates the round aperture 20. FIG. 5 shows a round aperture 20 with an electron gun 1.
It is the figure seen from the installation position side (namely, upper position).

Now, it is assumed that by driving only the third mask deflector 15, the electron beam 3 is irradiated to the position shown by the white circle V3 in the figure and a crossover image is formed. Then, as described above, vector display is used to represent this irradiation position, and the vector V3
Let r3 be the deflection amount and θ3 be the deflection phase (counterclockwise angle with respect to the X _R axis). This vector V3 also includes an error factor such as an assembly error of the charged particle beam exposure apparatus 30.

Subsequently, in the fifth step, only the fourth mask deflector 16 is driven to emit the electron beam 3 from the electron gun 1. At this time, the voltage applied to the fourth mask deflector 16 is set to V4, and the first to third mask deflectors 13 are used.
-15 shall not drive.

By driving the fourth mask deflector 16, the electron beam 3 is deflected as shown by the chain line arrow C in FIG. 4, and is irradiated onto the round aperture 20. Now, it is assumed that by driving only the fourth mask deflector 16, the electron beam 3 is irradiated onto the position shown by the black-filled mark V4 in FIG. When this irradiation position is displayed as a vector, the deflection amount of the vector V4 can be represented as r4 and the deflection phase can be represented as θ4. The vector V2 also includes an error factor such as an assembly error of the charged particle beam exposure apparatus 30.

As described above, in the fourth step, the vector V
The deflection amount r3 of 3 and the deflection phase θ3 are obtained, and in the fifth step, the deflection amount r4 and the deflection phase θ4 of the vector V4.
Is required. At this time, the deflection amounts r1 and r2 and the deflection phases θ1 and θ2 obtained in the first and second steps described above.
Is used as an initial value for obtaining the deflection amounts r3, r4 and the deflection phases θ3, θ4, the processing time can be shortened.

This is because the vectors V1 and V2 also include error factors such as an assembly error of the device 30, and the deflection amount r
1, r2 and deflection phases θ1 and θ2 are values for correcting this. Therefore, the deflection amounts r3, r4 and the deflection phases θ3, θ
4 is also likely to take values close to the deflection amounts r1 and r2 and the deflection phases θ1 and θ2. Therefore, the deflection amounts r1 and r2 and the deflection phase θ obtained in the first and second steps as initial values.
By using 1 and θ2, the processing time can be shortened.

In the subsequent sixth step, adjustment processing is performed to make the deflection amounts r3, r4 of the vectors V3, V4 obtained in the fourth and fifth steps zero. By performing this adjustment process, the electron beam 3 can be guided to the aperture hole 20a of the round aperture 20 no matter how the third and fourth mask deflectors 15 and 16 are driven.

In the present invention, similar to the above-mentioned third step,
In order to set the deflection amounts r3 and r4 to zero, the vector V
The method of canceling 3, V4 is used. Hereinafter, a method of canceling the vectors V3 and V4 will be described with reference to FIG.

According to the fourth and fifth steps, the third or fourth step is performed.
Driving the mask deflectors 15 and 16 of
3 and V4 are obtained, in the sixth step, the third and fourth mask deflectors 15 and 16 are simultaneously driven, and the vector V
4 is moved to a position symmetrical about the origin with respect to the vector V3, and the magnitude of the vector V4 is changed to the vector V3.
The adjustment processing is performed so that the size becomes the same.

Specifically, the vector V4 is replaced by the vector V3.
On the other hand, the deflection phase difference θ _d2 for moving to the symmetrical position about the origin is obtained, and the deflection amount correction coefficient G2 for making the magnitude of the vector V3 and the magnitude of the vector V4 the same is obtained. The deflection amount correction coefficient G2 is obtained by G1 = r3 / r4 (22), and the size of the vector V3 and the size of the vector V4 can be made equal by multiplying r4 by G2.

When the deflection amount correction coefficient G2 and the deflection phase difference θ _d2 are obtained as described above, the fourth mask deflector 1 is made to correspond to the deflection amount correction coefficient G2 and the deflection phase difference θ _d2.
The voltage applied to 6 and the phase of the power supply are controlled. Fourth
By controlling so as to correspond to the power of the phase to be applied to the mask deflector 16 to deflect the phase difference theta _d2, vector V
4 is moved as indicated by the double-dashed line arrow in FIG. 5B, and the voltage applied to the fourth mask deflector 16 is controlled so as to correspond to the deflection amount correction coefficient G2. And a vector V4 '(a vector shown by a chain line in the figure) symmetrical with respect to the origin is generated.

Now, the voltage applied to the fourth mask deflector 16 to generate the vector V4 'is set to V4',
In this case, the ratio between the voltage V3 applied to the third mask deflector 15 and the voltage V4 'applied to the fourth mask deflector 16 is defined as the deflection amount ratio (V3: V4').

Then, control is performed so as to correspond to the phase deflection phase difference θ _d2 of the power source applied to the fourth mask deflector 16,
Moreover, by controlling the deflection amount ratio so that V3: V4 'is always established, the vector V3 generated by the third mask deflector 15 and the vector V4' generated by the fourth mask deflector 16 cancel each other. It is synthetic vector V3 and the vector V4 'is always the origin of X _R -Y _R coordinate.

This is because the fourth mask deflector 16 is controlled on the basis of the deflection phase difference θ _d2 and the deflection amount ratio (V3, V4 ′), so that the third and fourth mask deflectors 15,
No matter how the 16 is driven, the electron beam 3 always passes through the aperture 20a formed in the round aperture 20, that is, the third and fourth mask deflectors 15, 16
It is possible to set the deflection efficiency of the crossover image on the round aperture 20 to zero when both are simultaneously driven.

The first to sixth steps described above are completed,
By adjusting the mask deflectors 13 to 16, the electron beam 3 emitted from the electron gun 1 always passes through the aperture 20a of the round aperture 20.

However, even if the electron beam 3 passes through the aperture 20a of the round aperture 20, the electron beam 3 does not always form a shaped beam image at a predetermined position on the sample 25. This will be described with reference to FIG. As shown in the figure, there are various modes in which the electron beam 3 passes through the aperture hole 20a, and the electron beam 3 is obliquely incident on the aperture hole 20a as indicated by the two-dot chain line arrows D and E. In some cases, as shown by the solid arrow F, the throttle hole 20a
May be incident vertically on.

The exposure position of the electron beam 3 on the sample 25 (indicated by an arrow H in the figure) is configured to face the aperture hole 20a of the round aperture 20, so that the electron beam 3 is directed to the aperture hole 20a. When the light is obliquely incident, a shaped beam image is formed at a position deviated from the predetermined exposure position H, and proper exposure processing cannot be performed. Therefore, in order to perform an appropriate exposure process, it is necessary to vertically inject the electron beam 3 into the aperture hole 20a of the round aperture 20.

The seventh and eighth steps are adjustment processes for making the electron beam 3 vertically incident on the aperture 20a of the round aperture 20. In the seventh step and the eighth step, the current value passing through the round aperture 20 is maximized, and the beam position shift on the sample 25 is minimized when different block patterns formed on the block mask 8 are selected. Is a process for adjusting the first to fourth deflecting means 13 to 16. First, the adjustment principle will be described.

As described above, in order to make the electron beam 3 incident on the aperture hole 20a of the round aperture 20, the coordinate system X _R -Y _R is assumed on the round aperture 20, and the mask deflectors 13 to 16 are driven. The vectors V1 to V4 generated by the above are offset from each other and converged to the origin of the coordinate system X _R -Y _R (corresponding to the arrangement position of the aperture 20a).

The electron beam 3 is irradiated onto the sample 25 at a predetermined exposure position H.
Also in the adjustment of irradiating the same, adjustment similar to the above adjustment method is performed. That is, as shown by the two-dot chain line arrows E and F in FIG. 6, when the electron beam 3 is not vertically incident on the aperture hole 20a, the electron beam 3 is displaced from the predetermined exposure position H. 3 is illuminated. Therefore, assuming the coordinate system X _S -Y _S having the origin at the predetermined exposure position H on the sample 25, the electron beam 3 to be irradiated may be adjusted so as to converge on the origin of the coordinate system X _S -Y _S. .

In addition, the deflection phase difference θ _{d1, the} deflection amount ratio (V1: V2 ′) obtained in the first to third steps, and the fourth
Or deflection phase difference theta _d2 obtained by the sixth _step, the deflection amount ratio (V3: V4 '), since the electron beam 3 when not maintain the value does not enter the aperture hole 20a, the deflection phase difference theta _{d1 ,} θ _d2 , the deflection amount ratio (V1: V2 ′) _{, and the} deflection amount ratio (V3: V4 ′) are not changed, and the electron beam 3 is converged on the origin of the coordinate system X _S -Y _S. Less than,
A specific adjustment method will be described.

In order to adjust the electron beam 3 to be vertically incident on the aperture hole 20a, first, the first mask deflector 13 and the second mask deflector 14 are simultaneously driven to emit the electron beam 3 from the electron gun 1. Let At this time, since the deflection phase difference θ _{d1 and the} deflection amount ratio (V1: V2 ′) set between the first and second mask deflectors 13 and 14 are maintained as described above, the electron beam 3 passes through the aperture 20a and is irradiated onto the sample 25. At this time, the third and fourth
Both mask deflectors 15 and 16 are not driven.

FIG. 7 is a view of the sample surface 25 viewed from the side where the electron gun 1 is disposed (that is, the upper position). Now, by simultaneously driving the first and second mask deflectors 13 and 14, the electron beam 3 is set on the sample 25 at X _S
It is assumed that a shaped beam image is formed by irradiating the position of (V1: V2 ′) indicated by the white square □ on the −Y _S coordinate (in the following description, the lower left corner of the shaped beam image indicated by □ is As a reference point).

Then, similarly to the above, the vector display is used to represent this irradiation position, and the vector (V
The deflection amount of 1: V2 ′) is r5, and the deflection phase (counterclockwise angle with respect to the X _S axis) is θ5.

Subsequently, the first and second mask deflectors 1
3 and 14 are stopped, the third and fourth mask deflectors 15 and 16 are simultaneously driven, and the electron beam 3 is emitted from the electron gun 1. At this time, since the deflection phase difference θ _{d2 and the} deflection amount ratio (V3: V4 ′) set between the third and fourth mask deflectors 15 and 16 are maintained as described above, the electron beam 3 passes through the aperture 20a of the round aperture 20 and is irradiated onto the sample 25.

Now, the third and fourth mask deflectors 1
By driving 5 and 16, the electron beam 3 is moved to the position shown in FIG.
It is assumed that the irradiation is performed on the position (V3: V4 ′) indicated by □ in satin. When this irradiation position is displayed as a vector, the deflection amount of the vector (V3: V4 ′) is r6 and the deflection phase is θ6.
It can be expressed as.

As described above, the vector (V1: V2 '),
The deflection amounts r5 and r6 of (V3: V4 ′) and the deflection phase are θ
5 and θ6 are obtained, the deflection amounts r5 and r6 are subsequently calculated.
The adjustment process is performed so that

By performing this adjustment process, the electron beam 3 is made to vertically enter the aperture hole 20a of the round aperture 20 no matter how the first to fourth mask deflectors 13 to 16 are driven. Thus, the crossover image can be formed at the predetermined exposure position H (the position of the hatched square in FIG. 7).

In the present invention, in order to set the deflection amounts r3 and r4 to zero, the vectors (V1: V2 ') and (V3: V4') are used for the sample 25 in the same manner as in the above steps.
The method of offsetting is used above. Hereinafter, a method of canceling the vectors (V1: V2 ′) and (V3: V4 ′) will be described with reference to FIG.

Vectors (V1: V2 '), (V3: V
4 ′) is obtained, the third and fourth mask deflectors 15 and 16 are simultaneously driven as a seventh step, and the vector (V3: V4 ′) is centered on the origin with respect to the vector (V1: V2 ′). A relative angle θ _d3 for moving to a symmetrical position is obtained.

Then, as an eighth step, the vector (V1:
The relative intensity G3 for making the magnitude of V2 ′) and the magnitude of the vector (V3: V4 ′) the same is obtained. The relative intensity G3 is obtained by G1 = (V1: V2 ') / (V3: V4') (23), and the size of the vector (V1: V2 ') and the vector (V3 are obtained by multiplying r6 by G3. : V4 ')
Can have the same size. Further, when the magnitudes of the vector (V1: V2 ′) and the vector (V3: V4 ′) are the same, the third and fourth mask deflectors 15,
The value of the voltage applied to 16 is (V3: V4 ')'.

As described above, the relative intensity G3 and the relative angle θ
_{When d3} is obtained, this relative intensity G3 and relative angle θ _d3
To correspond to the third and fourth mask deflectors 15, 1
The voltage applied to 6 and the phase of the power supply are controlled. Third
By controlling the phases of the power supplies applied to the first and fourth mask deflectors 15 and 16 so as to correspond to the relative angle θ _d3 , the vector (V3: V4 ′) is indicated by a chain double-dashed line in FIG. By moving as indicated by the arrow and controlling the voltage applied to the third and fourth mask deflectors 15 and 16 so as to correspond to the relative intensity G3, the vector (V1: V
2 ') and a vector (V3: V4') 'that is symmetric with respect to the origin
(A vector shown by a chain line in the figure) is generated.

Therefore, the third and fourth mask deflectors 1
The phases of the power supplies applied to the terminals 5 and 16 are controlled so as to correspond to the relative angle θ _d3 , and the first and second mask deflectors 1
The ratio of the voltage applied to the third and fourth mask deflectors 15 and 16 to the voltage applied to the third and fourth mask deflectors is always (V1: V
2 ') :( V3: V4') 'is controlled so that (V1: V2') and third and fourth generated by the first and second mask deflectors 13 and 14 are controlled. Of the vector (V
3: V4 ')' cancel each other out and the vector (V1: V
2 ') and vector (V3: V4' resultant vector) 'is always the origin of X _S -Y _S coordinates.

This is achieved by controlling the first to fourth mask deflectors 13 to 16 as described above.
No matter how the mask deflectors 13 to 16 are driven,
This means that the electron beam 3 is vertically incident on the aperture hole 20a formed in the round aperture 20 and is applied to a predetermined exposure position H on the sample 25.

The adjustment work of the charged particle beam exposure apparatus 30 itself is completed through the series of adjustment steps described above.
As described above, in the adjustment according to the method of the present invention, the first to fourth mask deflectors 13 to 16 are selectively driven, and the round aperture 20 or the sample 25 is irradiated with the error factors of the apparatus included. The position of the electron beam 3 is displayed as a vector, and the vector indicating the irradiation position is canceled for each set of the mask deflectors 13 to 16 to adjust the device 30. Therefore, the adjustment processing can be performed by measuring the irradiation position of the electron beam 3 once in each step, and therefore, compared to the conventional method of collecting the data for the adjustment many times, it can be performed in an extremely short time. Adjustment work can be performed.

On the other hand, the adjustment work of the charged particle beam exposure apparatus 30 itself is completed by carrying out the first to eighth steps, but the position adjustment of the charged particle beam exposure apparatus 30 and the block mask 8 is still carried out. Not not. Therefore, in the ninth step, as shown in FIG. 8, deflection coordinates (coordinates indicated by XY in the drawing) set in advance in the block mask 8 and mask coordinates (set in the device 30) ( figure, to match the coordinates) and indicated by X _m -Y _m.

As a result, the block mask 8 and the device 30 are
Since the coordinate systems of the two coincide with each other, the electron beam 3 is surely exposed to the predetermined exposure position H regardless of which of the plurality of block patterns 12 formed on the block mask 8 is selected.
Therefore, it is possible to prevent the deviation of the irradiation position on the sample 25 from occurring.

In the above-described embodiment, the voltage and phase of the first mask deflector 13 are fixed and the characteristics of the second mask deflector 14 are changed in the third step. The mask deflector 14 may be fixed and the characteristic of the first mask deflector 13 may be changed. Also,
In this way, when adjusting a pair of mask deflectors, one of them may be fixed and the other may be changed, and the mask deflector to be fixed is not limited. The sixth step and the seventh and eighth steps The same can be said for.

In the above, when an arbitrary point on the block mask 8 is selected, the electron beam passes through that point and vertically passes through the aperture hole 20a of the round aperture 20 to reach the exposure position H. The method of adjusting the deflection signal to be applied to the first to fourth deflectors 13 to 16 has been described.

Next, a mask deflector adjusting method performed by developing the above adjusting method will be described.

In the block mask exposure apparatus, the mask deflector is adjusted before the first operation after the installation. Once activated, the block deflector is readjusted each time the block mask is moved to select a new mask area.

The adjustment method of the charged particle beam exposure apparatus according to the present invention is summarized in that the items to be obtained are divided into two items, that is, the relative relationship between the deflecting means and the output value-beam coordinate position relationship.

As shown in FIG. 9, the adjusting method is roughly composed of a deflector relative relationship determining step 70 and a first deflector deflection output value-beam coordinate positional relationship determining step 71.

In the deflector relative relation obtaining step 70, in order to irradiate the charged particle beam through the round aperture 20 to the reference position on the sample surface, the above-mentioned first, second, third and fourth steps are performed.
The deflection value output to be applied to the deflectors 13, 14, 15, 16 of the above-mentioned second, third, and fourth deflectors 14, 15, 16 with respect to the deflection value output to the first deflector 13. This is a step of obtaining the relative relationship of the deflection value outputs by excluding the constraint condition for the block mask 8.

In the first deflector deflection output value-beam coordinate positional relationship obtaining step 71, the deflection output value to the first deflector 13 and the deflection output value to the first deflector 13 are added. This is a step of obtaining the relationship between the charged particle beam and the coordinate position on the block mask 8 at that time.

The relative positional relationship of the deflector is determined irrespective of the position of the block 8. Therefore, even when the block mask 8 is moved to select another mask area, the relationship obtained in step 70 is still valid. Therefore, the readjustment only needs to perform the step 71.

To perform block exposure by the electron beam exposure apparatus shown in FIGS. 1 and 38, prior to exposure, the electron beam emitted from the electron gun 1 at that time correctly passes through the designated block pattern and the exposure position. Therefore, the deflection signals to be added to the first to fourth deflectors 13 to 16 are adjusted in accordance with the state of the electron gun 1 at that time.

The mask deflector adjusting method is performed in this adjusting step.

Here, for convenience of explanation, an electron beam exposure apparatus to which the above-mentioned mask deflector adjusting method is applied will be described.

The electron beam exposure apparatus is different from FIG. 37 only in the pattern generating means.

As shown in FIG. 10, the pattern generating means 55A is supplied with the exposure data from the data memory 54 and generates the pattern data code.
0, a mask memory 81A that stores the position coordinates of the block pattern of the block mask 8, a mask area correction circuit 85, and a mask deflector relative correction circuit 86.

The mask area correction circuit 85 has a coefficient part, receives the block pattern arrangement position data X _m and Y _m from the mask memory 81, performs the operation of the equation 25 described later, and executes the first mask deflection. Deflection output value to the device 13 X _ma ,
Output Y _ma .

This output is applied to the first mask deflector 13 via the AMP section 57.

The mask deflector relative correction circuit 86 receives the output from the mask area correction circuit 85, performs the operation of Expression 24 described later, and outputs it to the second, third and fourth mask deflectors 14, 15 and 16. The deflection output value of is output.

This output is applied to the second, third and fourth mask deflectors 14, 15, 16 via the AMP unit 57.

Next, a method of adjusting the mask deflector will be described.

FIG. 11 shows the adjusting means.

The deflection output value applied to the mask deflector is represented by the equations (24) and (25) in FIGS. 15 and 16.

In FIG. 14, the mask deflector relative correction coefficient is a coefficient that gives a relative relationship between the two mask deflectors among the first to fourth mask deflectors 13 to 16.

In FIG. 16, the mask area correction coefficient is a coefficient that gives the relationship between the coordinates on one mask area of the block mask 8 having a plurality of mask areas and the output data to the first mask deflector 13. is there. In other words, the mask area correction coefficient is a correction coefficient determined for each mask area.

The mask area means an area where the electron beam can reach by the deflection by the first and second mask deflectors 13 and 14.

Adjustment of the mask deflector means that no matter which block pattern is selected, the mask deflector relative correction is performed so that the mask deflector is not shaded by the aperture and there is no displacement on the sample surface. Determining the coefficient,
When changing the block mask or moving the block mask to use a different mask area,
To determine the mask area correction coefficient for the mask area. This is shown in FIG.

The adjustment of the mask deflector is performed by obtaining the relative relationship between the mask deflectors without setting constraint conditions, and the relationship between the output data to the first mask deflector and the position on the block mask 8. It is divided into two items:

The mask deflector relative correction coefficient is measured and determined by the procedure shown in FIG. 11 (step 90).

The relative relationship of the second mask deflector 14 is obtained with the first mask deflector 13 as a reference (step 90 _{−1 in} FIG. 11).

At this time, the block mask 8 has been removed.

Here, as shown in FIG. 18, when the first mass deflector 13 performs an arbitrary deflection, the second mask deflector 14 deflects the electron beam so as to pass through the round aperture 20. Seeking a relationship to bring back.

Specifically, this is performed as described below.
(Steps 100 to 104 in FIG. 12).

-1 Third Mask Deflector 15 and Fourth
The mask deflector 16 is turned off.

An appropriate deflection output value is added to the first mask deflector 13 so that the electron beam passes through the measurement point in FIG.

The deflection output value is applied to the second mask deflector 14 while measuring the sample current, and the electron beam is swung back. The deflection output value applied to the second mask deflector 14 is measured when the electron beam is swept back and passed through the hole 20a of the round aperture 20, that is, when the sample current value reached a maximum.

Subsequently, the deflection output value applied to the first mask deflector 13 is changed so that the electron beam passes through the measurement point, and the second mask deflector when the sample current value becomes maximum. The deflection output value applied to 14 is measured.

This is performed for the measurement points → ... →. This is one unit.

Furthermore, as shown by the arrow in FIG. 19, the measurement points (1) to (7) are gradually expanded to the outside, and the operation of one unit is repeated.

This is performed up to the maximum portion of the maximum area (mask area) where the measurement point can select the pattern.

As a result, a map which is a set of deflection output values to the second mask deflector 14 is obtained.

-2: The obtained map is substituted into the equation 26 of FIG. 20, and the first coefficient that gives the deflection output value of the second mask deflector 14 corresponding to the deflection output value of the first mask deflector 13 is obtained. A (see FIG. 17) is obtained. This coefficient is referred to as a relative correction coefficient of the second mask deflector 14 with respect to the first mask deflector 13.

Here, since the coefficient is obtained based on the map, the coefficient can be obtained with high accuracy.

The relative relationship of the third mask deflector 15 is obtained with the fourth mask deflector 16 as a reference (step 90 _{-2 in} FIG. 11).

At this time, the block mask 8 has been removed.

Here, as shown in FIG. 21, when the fourth mask deflector 16 performs arbitrary deflection, the third mask deflector 15 causes the electron beam to pass through the round aperture 20. Find the function to swing in the opposite direction beforehand. Specifically, it is performed as described below (steps 105 to 109 in FIG. 13).

-1 First mask deflector 13 and second
The mask deflector 14 is turned off.

An appropriate deflection output value is added to the fourth mask deflector 16 so that the electron beam passes through the measurement point in FIG.

The deflection output value is applied to the third mask deflector 15 while measuring the sample current, and the electron beam is swung in the opposite direction in advance. The deflection output value applied to the third mask deflector 15 is measured when the electron beam is oscillated and passes through the hole 20a of the round aperture 20, that is, when the sample current unit reaches its maximum.

Subsequently, the deflection output value applied to the fourth mask deflector 16 is changed so that the electron beam passes through the measurement point, and the third mask deflector when the sample current value becomes maximum. The deflection output value applied to 15 is measured.

This is performed for the measurement points → ... →. This is one unit.

Further, as shown by the arrow in FIG. 22, the measurement points (1) to (7) are gradually expanded to the outside, and the operation of one unit is repeated.

This is performed up to the maximum part of the maximum area (mask area) where the measurement point can select the pattern.

As a result, a map which is a set of deflection output values to the third mask deflector 15 is obtained.

-2. Substituting the obtained map into Expression 27 of FIG. 23, the second coefficient giving the deflection output value of the third mask deflector 15 corresponding to the deflection output value of the fourth mask deflector 16. Find B ₁ (see FIG. 21). This coefficient is referred to as a relative correction coefficient of the third mask deflector 15 with respect to the fourth mask deflector 16.

Here, since the coefficient is obtained based on the map, the coefficient can be obtained with high accuracy.

The relationship between the first mask deflector 13 and the position on the block mask 8 is obtained (step 90 in FIG. 11).
_-3 ).

Here, the beam position on the block mask and the first mask deflector 13 when the electron beam is deflected by the first mask deflector 13 and the second mask deflector 14 according to the above relationship. The relationship with the deflection output value is calculated.

Specifically, the operation is performed as follows (steps 110 to 112 in FIG. 14).

-1 As shown in FIG. 24, the block mask 8 is set, and the calibration mask pattern mask area 120 is set at the center.

Calibration mask area 120
As shown in FIG. 25, the rectangular opening patterns are arranged in a grid pattern at a predetermined pitch.

-2, as shown in FIG. 24, the first mask deflector 13 and the second mask deflector 14 are interlocked with each other in accordance with the relationship obtained above (in accordance with the equation (24) in FIG. 15). Select a pattern for use.

-3 From the position of each calibration mask pattern on the mask area 120 and the deflection output value to the first mask deflector 1 required to select this position, from Expression 28 shown in FIG. , The mask area correction coefficient D (see FIG. 17) is obtained.

First and second mask deflectors 13, 1
4 and the relative relationship between the third and fourth mask deflectors 15 and 16 are obtained (step _{90-4 in} FIG. 11).

Here, when the electron beam is arbitrarily deflected by the first mask deflector 13 and the second mask deflector 14 in accordance with the relation (equation (24) in FIG. 15) obtained above, the sample surface In the above, in order to eliminate the positional deviation of the electron beam, the third mask deflector 15 and the fourth mask deflector 16 perform deflection while maintaining the relationship (Equation 27 in FIG. 23) obtained above, The first and second mask deflectors 13 and 14 and the third and fourth mask deflectors 15 and 1
Find the relative relationship with 6.

Specifically, it is performed as follows (steps 113 to 117 in FIG. 14).

-1 As shown in FIG. 27, the calibration pattern 122 at the mask coordinate origin 121 (see FIG. 25) is selected according to the relationship (equation (25) in FIG. 16) obtained above, and As shown enlarged in FIG. 28, a measuring beam 124 that is considerably thinner than the electron beam 123 in the mask deflector is formed. This is to improve the measurement accuracy.

By using this beam 124 to detect a mark on the surface of the sample 25, a beam position 125 serving as a reference on the surface of the sample is measured. Since the measuring beam 124 is thin, the beam position 125 can be accurately measured.

-2 Then, as shown in FIG. 29, another calibration pattern other than the pattern 122 in FIG. 25 is selected in accordance with the relationship (equation (25) in FIG. 16) obtained above. Similarly to the above, the measurement beam 124 is formed, the position 126 of the mark on the sample surface irradiated by this beam 124 is detected, and the deviation δ of the position 126 from the reference beam position 125 is measured.

In order that the beam position deviation δ becomes zero,
According to the relation (Equation 27 of FIG. 23) obtained above, the third
The mask deflector 15 and the fourth mask deflector 16 are interlocked.

As a result, the beam 124 changes from the state shown by the solid line in FIG. 29 to the state shown by the chain double-dashed line.

When the beam position deviation becomes zero, the third
The deflection output value to the mask deflector 15 and the deflection output value to the fourth mask deflector 16 are measured.

The above is performed for each of the plurality of calibration patterns other than the pattern 122.

As a result, a map of the deflection output value to the third mask deflector 15 and a map of the deflection output value to the fourth mask deflector 16 are acquired.

-3 The map of the deflection output value to the first mask deflector 13 when each calibration pattern is selected, the map of the deflection output value to the third mask deflector 15 and the fourth map From the map of the deflection output value to the mask deflector 16 and the equation (29) shown in FIG.
Mask deflector relative correction coefficient B (see FIG. 17) that gives the deflection output value of the third mask deflector 15 to the deflection output value of the mask deflector 13 of FIG. Relative to the mask deflector relative correction coefficient C (FIG. 1).
7)) is derived.

From the above, the relative correction coefficients A, B and C in FIG. 17 were obtained.

Next, step 91 in FIG. 11 is performed.

Here, the block mask 8 is moved to position the mask area used for block exposure at the center of the exposure apparatus, and the mask area correction coefficient D ₁ of this mask area is obtained.

Specifically, this is performed as shown in FIG.

First, all the mask deflectors 13 to 16 are interlocked with each other in accordance with the mask deflector correction coefficients A, B and C to select one rectangular aperture pattern existing in the new mask area (step 120).

Next, the position of the selected rectangular opening pattern on the block mask 8 and the first position required for the selection.
A new mask area correction coefficient is calculated by Equation 30 shown in FIG. 32 based on the output value to the mask deflector 13 in step 121.

The time required to calculate a new mask area correction coefficient is as short as about 1 minute.

Next, in FIG. 11, steps 92 and 93 are performed. In step 92, the mask area correction coefficient D ₁ is set as shown in FIG.
It is set in the coefficient part of the mask area correction circuit 85 which is 0.

In step 93, the mask deflector relative correction coefficients A, B and C are set in the coefficient portion of the mask deflector relative correction circuit 86 in FIG.

As a result, the preparation before exposure is completed.

After that, the block exposure 93 is started.

In the case where the block mask 8 has a structure in which block mask patterns are mixed in the calibration mask pattern mask area, FIG.
In the middle, step 91 is unnecessary, and the block area exposure can be started by setting the mask area correction coefficient D obtained in step 90 _-3 in the coefficient part of the mask area correction circuit 85.

If the desired mask pattern is not present in the currently selected mask area of the mounted block mask 8, the block mask 8 is moved to position it around another mask area.

In this case, it is necessary to readjust the mask deflector.

The work at this time will be described below.

The mask deflector relative correction coefficient A,
Since B and C are determined independently of the block mask 8, it is not necessary to recalculate the mask deflector relative correction coefficients A, B and C. It suffices to recalculate the mask area correction coefficient for the other mask area.

As shown in FIG. 33, first of all, the mask deflectors 13 to 13 are corrected in accordance with the mask deflector correction coefficients A, B, and C.
16 is interlocked to select one rectangular opening pattern existing in another mask area (step 130).

Next, the position of the selected rectangular opening pattern on the block mask 8 and the first necessary for selecting
Of the deflection output value to the mask deflector 13 of FIG.
A new mask area correction coefficient is calculated by Equation 30 shown in 2 (step 131).

Next, another calculated mask area correction coefficient D ₂ is set in the coefficient portion of the mask area correction circuit 85 in FIG. 10 (step 132).

Thereafter, the block exposure is restarted (step 1
33).

The time required for readjustment after moving the block mask is as short as about 1 minute.

Conventionally, in addition to the coefficient similar to the above-mentioned mask area correction coefficient, it is necessary to re-determine the coefficient similar to the mask deflector relative correction coefficient, which takes about half a day.

Therefore, according to the present embodiment, when the electron beam exposure apparatus is continuously moved, the time during which the block exposure is interrupted due to the movement of the block mask is greatly reduced as compared with the conventional case. As a result, the operating efficiency of the electron beam exposure apparatus can be improved, and the throughput of the semiconductor device can be improved as compared with the conventional case.

[0270]

As described above, according to the first aspect of the invention, it is possible to surely perform the adjustment processing of the apparatus in a short adjustment time.

According to the fourth aspect of the present invention, the deflecting means relative relation obtaining step and the first deflecting means deflection output value-beam coordinate positional relation obtaining step are divided, and the deflecting means relative relation obtaining step is performed. The configuration is performed without taking into consideration the arrangement of the transparent hole mask plate.

Therefore, in the readjustment when the transmission hole mask plate is moved to select another mask area or when the transmission hole mask plate is exchanged with another mask area, the relative relation of the deflecting means is changed. It is not necessary to recalculate, and it is sufficient to recalculate only the deflection output value of the first deflecting means-beam coordinate positional relationship.

Therefore, the time required for readjustment is about 15
It is about a minute, and it can be greatly shortened compared to the case where it took about half a day to date. The charged particle beam exposure apparatus is stopped during the readjustment.

Considering that the readjustment is performed at a relatively short interval, the operating efficiency of the charged particle beam exposure apparatus can be significantly improved as compared with the conventional one, and the throughput of the semiconductor device as a product can be improved. .

The invention of claim 5 has the same effect as the invention of claim 1.

According to the invention of claim 6, readjustment during operation can be performed in a short time, and an apparatus having high operation efficiency can be realized.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of an electron beam exposure apparatus which is adjusted by the method of the present invention.

FIG. 2 is an enlarged view showing the vicinity of a block mask for explaining the first to third steps.

FIG. 3 is a diagram for explaining an adjusting method in first to third steps.

FIG. 4 is an enlarged view showing the vicinity of a block mask for explaining the fourth to sixth steps.

FIG. 5 is a diagram for explaining an adjustment method in fourth to sixth steps.

FIG. 6 is an enlarged view showing the vicinity of a block mask for explaining the seventh and eighth steps.

FIG. 7 is a diagram for explaining an adjusting method in seventh to eighth steps.

FIG. 8 is a diagram for explaining a ninth step.

FIG. 9 is a diagram showing an outline of the present invention.

FIG. 10 is a diagram showing a configuration of a pattern generating means 55A in a control section of an embodiment of the electron beam exposure apparatus of the present invention.

FIG. 11 is a diagram showing an adjustment procedure of the mask deflector.

FIG. 12 is a flowchart of step 91 _-1 in FIG.

[13] In FIG. 11, a flowchart of the process 90 _-2.

14 is a flowchart of steps 90 _-3 and 90 _-4 in FIG.

FIG. 15 is a diagram showing an arithmetic expression for obtaining a deflection output value to a mask deflector.

FIG. 16 is a diagram showing an expression representing a deflection output value of the first mask deflection to 13.

FIG. 17 is a diagram illustrating adjustment of a mask deflector.

FIG. 18 is a diagram illustrating a step 90 ₋₁ in FIG. 11.

FIG. 19 is a diagram illustrating a step 90 ₋₁ in FIG. 11.

20 is a diagram showing an expression for calculating a relative correction coefficient A of the second mask deflector 14 with respect to the first mask deflector 13. FIG.

FIG. 21 is a diagram illustrating step _90-2 in FIG. 11.

FIG. 22 is a diagram for explaining step _90-2 in FIG. 11.

23 is a diagram showing an equation for calculating a relative coefficient B ₁ of the fourth mask deflector 16 with respect to the third mask deflector 15. FIG.

FIG. 24 is a diagram illustrating a step 90 _-3 in FIG. 11.

FIG. 25 is a diagram showing a mask area of a calibration mask pattern in FIG. 35.

FIG. 26 is a diagram showing an equation for calculating a mask area correction coefficient.

FIG. 27 is a diagram showing reference position measurement of a sample surface.

FIG. 28 is a diagram showing shaping of a measurement beam.

FIG. 29 is a diagram illustrating beam position deviation measurement.

FIG. 30 shows the first of the third and fourth mask deflectors 15 and 16.
7 is a diagram showing an equation for calculating mask deflector relative correction coefficients C and B for the mask deflector 13 of FIG.

FIG. 31 is a diagram illustrating step 91 in FIG. 11.

FIG. 32 is a diagram showing an equation for calculating a new mask area correction coefficient.

FIG. 33 is a diagram showing readjustment when a block mask is moved to set another mask area.

FIG. 34 is a conceptual diagram showing an example of an electron beam exposure apparatus.

FIG. 35 is a plan view showing a block mask.

FIG. 36 is a plan view showing one area portion of the block mask.

FIG. 37 is a plan view showing an example of a block pattern.

FIG. 38 is a diagram showing the control unit of the electron beam exposure apparatus together with the exposure unit.

FIG. 39 is a view showing the trajectory of the electron beam from the mask deflector at the most upstream position to the aperture.

40 is a diagram of the trajectory of the electron beam shown in FIG. 38 as seen from above the block mask.

FIG. 41 is a diagram showing changes in the trajectory of the electron beam when the deflection direction of the mask deflector arranged on the most upstream side is changed.

FIG. 42 is a diagram showing changes in the orbit of the electron beam when the deflection direction of the mask deflector arranged second from the most upstream side is changed.

FIG. 43 is a diagram showing a change in the orbit of the electron beam when the deflection directions of the third and fourth mask deflectors arranged from the most upstream side are changed.

FIG. 44 is a diagram showing a change of a plane on which an electron beam trajectory exists when the deflection intensity of a mask deflector arranged on the most upstream side is changed.

FIG. 45 is a diagram showing a change in a plane in which the trajectory of the electron beam exists when the deflection intensity of the mask deflector arranged second from the most upstream side is changed.

FIG. 46 is a diagram showing a change in the plane in which the orbit of the electron beam exists and the mask deflector arranged fourth from the most upstream side when the deflection intensity of the mask deflector arranged third from the most upstream is changed. FIG. 6 is a diagram showing a plane on which the trajectory of the electron beam exists when the deflection intensity of is changed.

FIG. 47 is a flow chart showing a main routine in the case of deflecting the electron beam and returning to the optical axis in the non-deflected state by a simulation to obtain a deflection condition required.

FIG. 48 is a flowchart showing a main routine in the case of deflecting the electron beam and returning to the optical axis in the non-deflected state by a simulation to obtain a deflection condition required.

FIG. 49 is a flowchart showing a subroutine for simulating deflection conditions necessary for deflecting the electron beam and returning it to the optical axis in the non-deflected state.

FIG. 50 is a flowchart showing a subroutine for simulating the deflection conditions necessary for deflecting the electron beam and returning it to the optical axis in the non-deflected state.

FIG. 51 is a plan view showing a conventional block mask having a calibration pattern.

52 is a plan view for explaining a calibration method using the block mask shown in FIG. 25. FIG.

FIG. 53 is a diagram showing a conventional pattern generating means.

FIG. 54 is a diagram showing a relationship between moving a block mask and readjustment.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 electron gun 2 optical axis 3 electron beam 6 slit deflector 8 block mask 13 first mask deflector 14 second mask deflector 15 third mask deflector 16 fourth mask deflector 25 sample 30 charged particle beam Exposure apparatus (electron beam exposure apparatus) 31 Electromagnetic lens 50 Control section 51 Storage medium 52 CPU 53 Interface 54 Data memory section 55 Pattern generating means 57 DAC / AMP section 58 Mask moving mechanism 70 Deflector relative relationship obtaining step 71 First Deflection output value to deflector-Beam coordinate position 120 Calibration mask pattern Mask area 121 Mask coordinate origin 122 Calibration mask pattern 123 Electron beam 124 Measurement beam 125 Reference beam position

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Junko Hatta 1015 Kamiodanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture, Fujitsu Limited (72) Inventor, Kiichi Sakamoto 1015, Kamiodanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture, Fujitsu Limited

Claims (6)

[Claims] 1. A charged particle beam generating means (1) for generating a charged particle beam (3) and a polygon arranged at a position through which the charged particle beam (3) passes and shaping the charged particle beam (3). A shaping plate (6), a transmission hole mask plate (8) in which one or more transmission hole patterns (12) having an arbitrary shape are formed, and an arrangement position of the charged particle beam generating means (1) is an upper part. In this case, the first to fourth deflecting means (13 to 13) are arranged so as to be substantially mirror-symmetrical one by one on the upper and lower sides of the transmission hole mask plate (8) so as to deflect the charged particle beam (3). 1
6) and a round aperture (20) arranged below the first to fourth deflecting means (13 to 16), which is a method for adjusting a charged particle beam exposure apparatus, wherein the polygon shaping plate The first installed at the position closest to (6)
Of the crossover image on the round aperture (20) by the first deflecting means (13) and the deflection amount (r1).
And a second deflecting means (14) arranged below the first deflecting means (13) and above the transmitting hole mask plate (8) to drive the second deflecting means (14). Second step of obtaining the deflection phase (θ2) and the deflection amount (r2) of the crossover image on the round aperture (20) by the deflecting means (14), and the first and second deflecting means (13, 14), the first and second such that the deflection efficiency of the crossover image on the round aperture 20 is made zero when driven simultaneously.
Third step of obtaining the deflection difference (θd ₁ ) and the deflection amount ratio (V1: V2 ′) of the deflecting means (13, 14) and the third step disposed at the lower position of the transmission hole mask plate (8). 4th step of driving only the deflecting means (15) of (3) and obtaining the deflection phase (θ3) and the deflection amount (r4) of the crossover image on the round aperture (20) by the third deflecting means (15). And a fourth portion arranged at a lower position of the third deflecting means (15)
Of the crossover image on the round aperture (20) by the fourth deflecting means (16) and the deflection amount (r4).
And a third step for making the deflection efficiency of the crossover image on the round aperture (20) to be zero when the third and fourth deflecting means (15, 16) are simultaneously driven. 3 and 4
The sixth step of obtaining the deflection difference (θd ₂ ) and the deflection amount ratio (V3: V4 ′) of the deflecting means (15, 16), and driving the first to fourth deflecting means (13-16), The charged particle beam (3) is most vertically incident on the round aperture (20), and the first and second deflecting means (13, 14) and the first and second deflecting means (13, 14) are used without changing the deflection phase difference and the deflection amount ratio. Seventh step of obtaining the relative angle (θd ₃ ) with the third and fourth deflecting means (15, 16), and after obtaining the relative angle, the first to fourth deflecting means (13-16) The charged particle beam (3) is most vertically incident on the round aperture (20) and the first and second deflecting means (13, 13) without changing the deflection phase difference and the deflection amount ratio. 14) and the relative intensity (G3) between the third and fourth deflecting means (15, 16) is obtained. And the relative deflection amount and relative phase difference of the first to fourth deflection means (13 to 16), the deflection amount and deflection of the deflection coordinate on the transmission hole mask plate (8). A ninth step of matching the directions and selecting the transmission hole pattern (12) on the transmission hole mask plate (8)
A method of adjusting a charged particle beam exposure apparatus, comprising: 2. In the fourth step and the fifth step, the deflection phase and the deflection amount already obtained in the first and second steps are used as the deflection phase of the third and fourth deflecting means (15, 16). The method of adjusting a charged particle beam exposure apparatus according to claim 1, wherein the adjustment method is used as an initial value when obtaining the deflection amount. 3. In the seventh step and the eighth step, the current value passing through the round aperture (20) is maximized, and different transmission hole patterns (12) are formed on the transmission hole mask plate (8). If selected, the sample (2
5) The charged particle beam exposure according to claim 1 or 2, wherein the deflection phase and the deflection amount of the first to fourth deflecting means (13 to 16) are obtained so that the above beam position deviation is minimized. How to adjust the device. 4. A charged particle beam generating means (1) for generating a charged particle beam (3) and a polygon arranged at a position through which the charged particle beam (3) passes and shaping the charged particle beam (3). A shaping plate (6), a transmission hole mask plate (8) in which one or more transmission hole patterns (12) having an arbitrary shape are formed, and an arrangement position of the charged particle beam generating means (1) is an upper part. In this case, the first and second deflecting means (13, 14) arranged above the transparent hole mask plate (8) for deflecting the charged particle beam (3) and the transparent hole mask Third and fourth deflecting means (15, 1) disposed on the lower side for deflecting the charged particle beam (3)
A method for adjusting a charged particle beam exposure apparatus, comprising: 6) and a round aperture (20) arranged below the fourth deflecting means (16), wherein the charged particle beam is rotated by the round aperture (20). )
To irradiate a reference position on the sample surface through the first, second, third and fourth deflecting means (13, 14, 15,
With respect to the deflection value output to be added to 16), the deflection value output to the second, third and fourth deflection means (14, 15, 16) with respect to the deflection value output to the first deflection means (13). Deflection means relative relationship obtaining step (84 _-1 ) for obtaining the relative relationship by excluding the constraint conditions for the transmission hole mask plate (8), the deflection output value to the first deflection means, and the first deflection means. First deflection means deflection output value-beam coordinate positional relation calculation for obtaining the relation between the charged particle beam coordinate position on the transmission hole mask plate (8) when the deflection output value is applied to the deflection means A method of adjusting a charged particle beam exposure apparatus, characterized in that the configuration is divided into a step (84 _-2 ). 5. A charged particle beam generating means (1) for generating a charged particle beam (3) and a polygon arranged at a position through which the charged particle beam (3) passes and shaping the charged particle beam (3). A shaping plate (6), a transmission hole mask plate (8) in which one or more transmission hole patterns (12) having an arbitrary shape are formed, and an arrangement position of the charged particle beam generating means (1) is an upper part. In this case, the first and second deflecting means (13, 14) arranged above the transparent hole mask plate (8) for deflecting the charged particle beam (3) and the transparent hole mask Third and fourth deflecting means (15, 1) disposed on the lower side for deflecting the charged particle beam (3)
A method for adjusting a charged particle beam exposure apparatus, comprising: 6) and a round aperture (20) arranged below the fourth deflecting means (16), the first mask deflector (13) For each of the plurality of deflection output values added to the second mask, the deflection output value applied to the second mask deflector (14) when the sample current value becomes maximum is measured, and the second mask A map of the deflection output value to the deflector (14) is obtained, and from the map, the deflection output value to the second mask deflector (14) with respect to the deflection output value to the first mask deflector (13). Step (90 _-1 , 100 to ₁ ) for obtaining the first mask deflector relative correction coefficient (A) giving a value
04) and for each of the plurality of deflection output values added to the fourth mask deflector (16), in addition to the third mask deflector (15) when the sample current value becomes maximum. The deflection output value is measured to obtain a map of the deflection output value to the third mask deflector (15), and the map is used for the deflection output value to the fourth mask deflector (16). Step (90 _-2 , 105-1) for obtaining a second mask deflector relative correction coefficient (B ₁ ) giving the deflection output value to the third mask deflector (15)
09) and the second mask deflector (14) according to the first coefficient (A), and the charged particle beam is transmitted through the first mask deflector (13). Deflecting to a plurality of beam positions on the hole mask plate (8),
A map of the deflection output value to the first mask deflector (13) corresponding to the plurality of beams is obtained, and from the map, the first mask deflector (13) for the beam position on the transmission hole mask is obtained. step (90 _-3, 110 to 112) for obtaining the mask Elijah correction coefficient (D) giving the deflection output value and, performs deflected by the first mask deflector (13), said first coefficient (a When the charged particle beam is deflected to a plurality of positions on the transmission hole mask plate (8) by performing the deflection by the second mask deflector (14) according to (4), the positional deviation of the beam on the sample surface is caused. A third mask deflector (15) determined according to the map of the deflection output value applied to the fourth mask deflector (16) and the deflection output value and the second coefficient (B ₁ ) in order to eliminate it. The map of the deflection output value to , A third mask deflector relative correction coefficient (B) which gives the deflection output value to the third mask deflector (15) with respect to the deflection output value to the first mask deflector (13) from the two maps. And the first mask deflector (1
4) the fourth mask deflector (1
Step (90 ₋₄ , 113 to 11) for obtaining a fourth mask deflector relative correction coefficient (C) giving a deflection output value to 6).
7) and the first mask deflector relative correction coefficient (A), the mask area correction coefficient (D), the third mask deflector relative correction coefficient (B), and the fourth mask deflector relative According to the correction coefficient (C), the deflection output values to the first, second, third and fourth mask deflectors for the desired transmission hole pattern of the transmission hole mask plate (8) are determined. A method for adjusting a charged particle beam exposure apparatus, which is characterized. 6. A charged particle beam generating means (1) for generating a charged particle beam (3) and a polygon arranged at a position through which the charged particle beam (3) passes and shaping the charged particle beam (3). A shaping plate (6), a transmission hole mask plate (8) in which one or more transmission hole patterns (12) having an arbitrary shape are formed, and an arrangement position of the charged particle beam generating means (1) is an upper part. In this case, the first and second deflecting means (13, 14) arranged above the transparent hole mask plate (8) for deflecting the charged particle beam (3) and the transparent hole mask Third and fourth deflecting means (15, 1) disposed on the lower side for deflecting the charged particle beam (3)
6) and a charged particle beam exposure apparatus having a round aperture (20) arranged below the fourth deflecting means (16), storing arrangement coordinates of each transmission hole pattern of the transmission hole mask plate. Therefore, when the charged particle beam is to be shaped into a desired transmission hole pattern, a storage unit (81) for outputting the arrangement coordinates of the transmission hole pattern, and the mask area correction coefficient ( D) is set in the coefficient portion, the transmission hole pattern arrangement coordinates from the storage means are input, and the deflection output value for deflecting the charged particle beam to the desired transmission hole pattern is output to the first deflection means. Mask area correction means (85) for performing the above, and the first, third, and fourth mask deflector relative correction coefficients (A, B, C) obtained in claim 5 are set in the coefficient part, and the mask area The deflection output value from the correcting means to the first deflecting means is input, and the beam shaped by the desired transmission hole pattern is set at the reference position of the sample to the second, third and fourth deflecting means. A charged particle beam exposure apparatus comprising a mask deflector correction means (86) for outputting a deflection output value for guiding.