JP3448781B2 - Charged particle beam transfer device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for forming a pattern of a semiconductor memory or the like by using a charged particle beam.

[0002]

2. Description of the Related Art Conventionally, as an apparatus of this type, a drawing apparatus for irradiating an extremely thin charged particle beam on a sample to draw a pattern one by one, and a charged particle beam for irradiating a predetermined area of a mask A transfer device that collectively reduces and transfers a pattern formed in a range onto a sample is known. In particular, as the latter transfer device, a device in which a pattern forming region on a mask is divided into a plurality of visual fields and pattern transfer is repeated for each visual field is being put into practical use. In the transfer device that performs such field division, the mask is continuously moved in a direction parallel to the sample at a constant speed, and the sample is moved in the opposite direction to the mask at a synchronous speed (at a reduction ratio of 1/4) according to the reduction ratio of the pattern. (If there is one-fourth of the moving speed of the mask)
The transfer is repeated by successively changing the field of view of the mask and the transferred region on the sample, which are continuously moved to the optical axis of the charged particle beam optical system. The sample is moved in the direction opposite to the mask because the lens system between the mask and the sample reverses the moving direction of the mask and the moving direction of the image of the pattern guided on the sample. As the mask and the sample move continuously, the beam flux of the charged particle beam that has passed through the mask is scanned at the same speed in the same direction as the moving direction of the sample.

[0003]

In the transfer device for dividing the visual field described above, unless the mask and the sample are moved accurately at the synchronous speed corresponding to the reduction ratio of the pattern, the image of the pattern guided onto the sample and the sample. During this period, relative movement occurs and the pattern transfer position deviates from the regular transfer position. In particular, since vibration is generated as the mask and the sample are driven, the moving speed of the mask and the sample constantly fluctuates, and there is a high possibility that the pattern transfer position will be displaced by this fluctuation component.
Further, due to the vibration in a direction different from the continuous movement direction, the transfer position may be displaced in a direction other than the continuous movement direction.

Here, in the drawing apparatus for drawing the above-mentioned patterns one by one, the time lag from when the current position of the sample is specified to when the pattern drawing is started is several 100 ns.
Since the density of the charged particle beam is high and the drawing time for each pattern is as short as several 100 ns. If the current position of the sample is detected immediately before drawing, the drawing is performed based on the detected position. With, it is possible to accurately form a pattern at a target position. However, in an apparatus that collectively transfers a certain range of a mask, even if the current positions of the mask and the sample are specified, the time lag until the transfer is started based on the data is as long as several tens of μs. There is a high possibility that the transfer position will shift due to fluctuations in the moving speed of the sample. In addition, the density of charged particle beams is low and the transfer time for one transfer is several tens.
Since it is as long as μs., the image of the pattern guided on the sample and the sample relatively move during the transfer, and the pattern transferred to the sample may be blurred.

An object of the present invention is to provide a transfer device capable of accurately transferring a pattern to a regular pattern transfer position on the sample even when the mask and the sample are continuously moved.

[0006]

According to a first aspect of the present invention, a charged particle beam is irradiated to a specific area of the mask while continuously moving the mask and the sample, and the pattern formed in the area is applied to the sample at a predetermined reduction ratio. It is applied to a charged particle beam transfer device for transferring. Then, a detection unit that detects the moving speed of the mask and the sample in a specific direction at a specific time before the transfer start time, a regular pattern transfer position on the sample at the transfer start time, and an image of the pattern guided on the sample. Of the deviation amount in the specific direction based on the moving speed detected by the detecting means and the acceleration obtained by differentiating the moving speed, and the mask based on the prediction result of the deviation amount predicting means. The above-described object is achieved by including a transfer position correcting unit that adjusts the deflection amount of the charged particle beam that has passed through the sample in a specific direction to correct the transfer position of the pattern actually transferred to the sample.

According to a third aspect of the present invention, a charged particle beam transfer for irradiating a specific region of the mask with a charged particle beam while continuously moving the mask and the sample to transfer the pattern formed in the region to the sample at a predetermined reduction ratio. Applies to equipment. And
The detection means for detecting the moving speed of the mask and the sample in the specific direction at the specific time before the transfer start time, and the relative moving speed in the specific direction of the sample and the image of the pattern guided on the sample during the transfer are detected. A relative moving speed predicting means for predicting the moving speed detected by the detecting means and an acceleration obtained by differentiating the moving speed; and a sample being transferred and the sample based on the prediction result of the relative moving speed predicting means. Relative moving speed correcting means for correcting the actual relative moving speed in a specific direction with respect to the image of the pattern guided above by adjusting the scanning speed of the charged particle beam passing through the mask in the specific direction. Thus, the above-mentioned object is achieved. In the above invention, the direction of movement of the mask and the sample, or the direction orthogonal to the direction of continuous movement of the mask and the sample can be selected as the direction of detection of the moving speed by the detecting means.

The deviation amount predicting means is, for example, the moving speed of the mask at a specific time is Vi, the mask position at a specific time obtained by integrating the moving speed of the mask is Si, and the moving speed of the sample is at a specific time. Vs, the time difference between the specific time and the transfer start time is Δt, the sample position at the specific time obtained by integrating the moving speed of the sample is Ss, and the reduction ratio of the pattern from the mask to the sample is k. The predicted value ΔS of the image shift amount is

[Equation 1]         ΔS = (k · Si−Ss) + (k · Vi−Vs) · Δt (1) Calculate by

When higher precision is required, when the acceleration obtained by differentiating the moving speed Vi of the mask is αi and the acceleration obtained by differentiating the moving speed Vs of the sample is αs, the pattern image is obtained. The predicted value ΔS of the deviation amount of

[Equation 2]   ΔS = (k · Si-Ss)     + [K ・ (Vi + αi ・ Δt)-(Vs + αs ・ Δt)] ・ Δt …… (2) Calculate by

Further, the relative moving speed predicting means is, for example, Vi is a moving speed of the mask at a specific time, Vs is a moving speed of the sample at a specific time, and k is a reduction rate of the pattern from the mask to the sample. At this time, the predicted value ΔV of the relative movement speed
To

[Equation 3]             ΔV = k · Vi-Vs (3) Calculate as

In another relative moving speed predicting means, the moving speed of the mask at a specific time is Vi, and the acceleration αi of the mask at a specific time obtained by differentiating the moving speed of the mask is a specific time and a transfer start time. The change amount of the mask moving speed obtained by multiplying the time difference Δt by ΔVi, the moving speed of the sample at a specific time is Vs, and the acceleration αs of the sample at a specific time obtained by differentiating the moving speed of the sample is specified. When the change amount of the sample moving speed obtained by multiplying the time difference Δt between the time and the transfer start time is ΔVs and the reduction rate of the pattern from the mask to the sample is k, the predicted value ΔV of the relative moving speed is

[Equation 4]           ΔV = (k · Vi−Vs) + (k · ΔVi−ΔVs) (4) Calculate as

[0012]

According to the invention of claim 1, the moving speed of the mask and the sample is detected at a specific time prior to the transfer of the pattern, and the detected moving speed and the acceleration obtained by differentiating the moving speed are applied to the sample at the transfer start time. The amount of deviation in the specific direction between the regular pattern transfer position of and the image of the pattern guided on the sample is predicted. At the time of transfer, processing such as adjusting the amount of deflection of the charged particle beam by the amount of displacement is performed based on the result of prediction of the amount of displacement to correct the transfer position of the pattern actually transferred to the sample in the specific direction.

According to the third aspect of the invention, the moving speeds of the mask and the sample are detected at a specific time prior to the transfer of the pattern,
From the detected moving speed and the acceleration obtained by differentiating the moving speed, the relative moving speed between the sample during transfer and the image of the pattern guided on the sample is predicted. Then, at the time of actual transfer, the relative moving speed between the sample being transferred and the image of the pattern guided on the sample is corrected based on the prediction result of the relative moving speed predicting means.

The deviation between the pattern image guided on the sample and the normal pattern transfer position and the correction of the relative moving speeds thereof will be described with reference to FIG. FIG. 4A shows continuous movement of the mask 101 and the sample 102 at the time of pattern transfer. The mask 101 moves to the right in the drawing at a velocity Vi.
The sample 102 continuously moves to the left in the drawing at a velocity Vs. A plurality of fields of view F arranged in a line in a direction orthogonal to the continuous movement direction on the mask 101 are stepwise irradiated with the charged particle beam EB, and a pattern (not shown) of each field of view F is sequentially transferred to the sample 102. To be done. This process can be realized by changing the deflection amount of the irradiation optical system of the charged particle beam EB stepwise in the column direction of the field of view F. Since the mask 101 and the sample 102 continue to move even while irradiating one field of view F, the beam of the charged particle beam EB that has passed through the mask 101 while continuously irradiating one field of view F with the charged particle beam EB. The bundle is scanned at the same speed as the sample 102 in the continuous movement direction of the sample 102 (to the left in the drawing).

In FIG. 4, mask 101 to sample 102
The reduction ratio of the pattern image by the lens system (not shown) between
Then, the moving speed of the pattern image guided to the sample 102 is k × Vi, and this speed matches the speed of the sample 102 in order to eliminate the relative movement between the pattern image and the sample 102.

## EQU00005 ## Vs = k.Vi (5) must be established. When the equation (5) is not satisfied, the difference between both sides is the pattern image on the sample 102 and the sample 102.
Corresponds to the relative movement speed with respect to, and this relative movement speed is given by the above equation (3). When ΔV calculated by the equation (3) is substantially constant, the scanning speed of the charged particle beam EB in the continuous movement direction of the sample 102 when irradiating one visual field F is increased by an amount corresponding to ΔV or If the speed is reduced, the actual relative moving speed between the pattern image on the sample 102 and the sample 102 is reduced to 0.
As a result, accurate transfer without blurring is possible.

A specific time in the transfer process (for example, the mask 10)
1 and the moving speed Vi, Vs of the mask 101 and the sample 102 with reference to the time when the sample 102 is aligned.
Are integrated, the positions Si and Ss of the mask 101 and the sample 102 in the continuous movement direction are calculated. And
By multiplying the position Si of the mask 101 by the reduction ratio k of the lens system, the position of the pattern image guided on the sample 102 can be obtained. The position of the pattern image thus obtained and the sample 10
The difference between the two positions corresponds to the amount of deviation between the pattern image on the sample 102 and the regular transfer position in the continuous movement direction.
FIG. 4B shows an example of the change over time of this shift amount. The illustrated temporal change is brought about by the vibration of the driving mechanism of the mask 101 and the sample 102.

In FIG. 4B, the mask 1 is _generated at time t _0.
01 and the sample 102 to detect the moving speed of the sample 102
A shift amount between the upper pattern image and the regular pattern transfer position of the sample 102 is calculated, and at a time t _{1, a} specific visual field F (see FIG.
It is assumed that (a) is irradiated with a charged particle beam to start transfer, and irradiation of the specific field of view F with the charged particle beam is finished at time t ₂ . Since the moving speeds of the mask 101 and the sample 102 change every moment between time t _{0 and} time t ₂ ,
The shift amount [Delta] S ₀ determined at ₀ be corrected before the time t _1, and remaining amount of deviation at the time t ₁ ΔS _1, During the time t ₁ ~t ₂ and movement pattern image only [Delta] S ₂ patterns The image is blurred.
Since the time difference Δt between the times t _{0 and} t ₁ is known in advance as the setup time for irradiating the single field of view F with the charged particle beam,
When the moving speeds of the mask 101 and the sample 102 between times t _{0 and} t ₁ can be regarded as substantially constant, in other words, in FIG.
If the slope θ of the change curve between times t _{0 and} t ₁ can be regarded as substantially constant, the moving speed of the mask 101 and the moving speed of the sample 102 detected at time t ₀ can be multiplied by the time difference Δt. Position ΔSi of mask 101 at time t ₁ , sample 1
The position ΔSs of 02 can be predicted. Therefore, the above equation (1)
Thus, the shift amount (ΔS ₀ + ΔS ₁ ) in FIG. 4B can be predicted. The blur amount ΔS ₂ can be suppressed by correcting the relative moving speed according to the above-described equation (3).

Next, correction in an area where the curvature of the displacement amount is large, such as immediately after time t ₃ in FIG. 4B, will be considered. Figure 4
4C shows the differential value of the change curve shown in FIG. 4B in the vicinity of time t ₃ , and the vertical axis of the drawing corresponds to the relative moving speed between the sample 102 and the pattern image. At time t _3, the moving speed of the mask 101 and the sample 102 is detected, and at the subsequent time t ₄ , irradiation of the charged particle beam to one visual field F is started,
Although if finished irradiation at t ₅ and the beam blur shift amount and [Delta] S ₄ of [Delta] S ₃ occurs, the range is corrected surrounded by a tangent L and velocity v ₀ for considering the acceleration data and the time t _a Therefore, the deviation amount error is ΔS _a and the blur amount error is ΔS _b.
Compressed up to. It should be noted that not only the continuous movement direction of the mask and the sample but also other directions can be similarly obtained.

[0019]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS. FIG. 1 shows an example of an electron beam transfer apparatus according to this embodiment. The electron beam emitted from the electron gun 1 is collimated into a parallel beam by the condenser lens 2 and guided by the field-of-view selection deflector 24 to one of a plurality of fields arranged in the y direction in FIG. The patterned electron beam passing through the mask 4 is imaged on the sample 10 by the lenses 8 and 9 at a predetermined reduction ratio (1/4 in this embodiment). The mask 4 is set on the mask stage 3, and is continuously moved in the x direction and stepwise moved in the y direction by the driving device 5. A laser interferometer 14 is attached to the mask stage 3. The laser interferometer 14 outputs a pulse signal corresponding to the moving direction and moving speed of the mask stage 3 to the speed calculator 16. The speed calculator 16 counts the number of pulses from the laser interferometer 14 and multiplies the counted value by a predetermined constant to calculate the moving speed of the mask stage 3. Speed calculator 1
The moving speed calculated by 6 is output to the CPU 18, the position calculator 15, and the acceleration calculator 17. The position calculator 15 calculates the position of the mask stage 3 by integrating the speed calculated by the speed calculator 16, and the acceleration calculator 17 calculates the acceleration of the mask stage 3 by differentiating the speed calculated by the speed calculator 16. The calculation results of the position calculator 15 and the acceleration calculator 17 are also output to the CPU 18.

The sample 10 is fixed to a sample table 12, and a driving device 11 makes it possible to continuously move the mask stage 3 in a direction opposite to the continuous movement in the x direction at a speed of 1/4 of the moving speed of the mask 4. The reverse direction is used because the pattern images are inverted by the lenses 8 and 9. A laser interferometer 23 is attached to the drive unit 11. Laser interferometer 23
Outputs a pulse signal corresponding to the moving speed of the sample table 12 in the above-described continuous moving direction and the direction orthogonal thereto to the speed calculator 21. The speed calculator 21 counts the number of pulses from the laser interferometer 23 and multiplies the counted value by a predetermined constant to calculate the moving speed of the sample table 12. The moving speed calculated by the speed calculator 21 is output to the CPU 18, the position calculator 20, and the acceleration calculator 22. The position calculator 20 calculates the position of the sample table 12 by integrating the speed calculated by the speed calculator 21, and the acceleration calculator 22 calculates the acceleration of the sample table 12 by differentiating the speed calculated by the speed calculator 21. The calculation results of the position calculator 20 and the acceleration calculator 22 are also output to the CPU 18.

The CPU 18 calculates the deflection amount and scanning speed of the electron beam based on the given position, speed and acceleration of the mask 4 and the position, speed and acceleration of the sample 10. Then, a signal corresponding to the calculation result of the deflection amount is output to the electromagnetic deflection coil 6 via the interface 19,
A signal corresponding to the calculation result of the scanning speed is supplied to the electrostatic deflector 7 via the interface 19. The electromagnetic deflection coil 6 sets the deflection amount of the electron beam in accordance with the calculation result of the CPU 18, and the electrostatic deflector 7 scans the electron beam in the x direction at a speed corresponding to the calculation result of the CPU 18.

The processing of the CPU 18 will be described with reference to FIG.
As shown in FIG. 2, when the transfer of the specific visual field F of the mask 4 is completed, the CPU 18 first reads the speed calculated by the speed calculators 16 and 21 in step S1 in order to transfer the next visual field F. Next, in step S2, the shift amount of the pattern image from the regular transfer position and the relative moving speed between the pattern image and the sample 10 are calculated as follows. In this embodiment, the calculation is performed in the x direction (continuous movement direction).

In order to calculate the shift amount, the CPU 18 reads the moving speed Vi of the mask 4 at a specific time (t = t ₀ ) from the speed calculator 16 and obtains it by integrating the moving speed Vi of the mask 4 from the position calculator 15. The mask position Si at t = t ₀ is read. Next, the CPU 18 makes the speed calculator 21
From the moving speed Vs of the sample 10 at t = t ₀ ,
The sample position Ss at t = t ₀ obtained by integrating the moving speed Vs of the sample 10 is read from the position calculator 20.

Further, the CPU 18 has a specific time t = t _0.
And the time difference Δt between the transfer start time t = t ₁ is given in advance. Then, the CPU 18 calculates the shift amount ΔS in the x direction between the regular pattern transfer position of the sample 10 and the pattern image guided onto the sample 10 at the transfer start time t ₁ by the following equation (6).

(6) ΔS = (k · Si−Ss) + (k · Vi−Vs) · Δt (6) where k is the reduction ratio of the lenses 8 and 9, and as described above, in the present embodiment. It is 1/4.

When higher precision transfer is required,
The CPU 18 reads the acceleration αi of the mask 4 at t = t ₀ obtained by differentiating the moving speed Vi of the mask 4 from the acceleration calculator 17, and t = obtained by differentiating the moving speed Vs of the sample 10 from the acceleration calculator 22. The acceleration αs of the sample 10 at t ₀ is read.

Then, the CPU 18 calculates the shift amount ΔS in the x direction between the regular pattern transfer position of the sample 10 and the pattern image guided onto the sample 10 at the transfer start time t ₁ by the following equation (7).

[Equation 7]   ΔS = (k · Si-Ss)       + [K ・ (Vi + αi ・ Δt)-(Vs + αs ・ Δt)] ・ Δt …… (7)

Further, the CPU 18 calculates the relative moving speed by the following equation (8).

[Equation 8]           ΔV = k · Vi = Vs (8)

If higher precision transfer is required, C
The PU 18 obtains the change amount Δi of the moving speed of the mask 4 obtained by multiplying the acceleration αi by the time difference Δt and the change amount ΔVs of the moving speed of the sample 10 obtained by multiplying the acceleration αs by the time difference Δt. Then, the relative movement speed is obtained by the following equation (9).

[Equation 9]       ΔV = (k · Vi−Vs) + (k · ΔVi−ΔVs) (9)

Then, in step S3, a signal corresponding to the shift amount is sent to the electromagnetic deflection coil 6 through the interface 19 based on the calculation result of step S2, and a signal corresponding to the relative moving speed is transmitted through the interface 19. It is sent to the electrostatic deflector 7 and the lens condition is set according to the visual field to be transferred. In the following step S4, the process stands by until the magnetic field for deflecting the electron beam beam bundle is settled. Then, in step S5, the field of view to be transferred is transferred. It should be noted that the shift amount in the y direction and the relative movement speed can also be obtained in the same manner as the above procedure.

An example of calculation of the residual error when the above-described correction is performed will be described with reference to FIG. FIG. 3A shows an example of a temporal change of the pattern image guided on the wafer 10 and the relative movement speed of the wafer 10 in the continuous movement direction, and FIG. FIG. 3C is an enlarged view of the first part, and FIG. 3C is a view showing the acceleration corresponding to FIG. The curve of relative movement speed vibrates with a cycle of 250 Hz centered on 10 mm / sec and an amplitude of ± 1 mm / sec.
It is assumed that the outputs of the speed calculators 16 and 21 are read at time t = 0, the transfer is started at t = 50 μsec, and the transfer is ended at t = 100 μsec. In addition, the relative velocity curve of FIG.

[Equation 10] V = 10 + sin (500π · t) Suppose

When the displacement amount is not corrected as in the conventional case, the pattern displacement amount Q ₁ at the start of transfer is as shown in FIG.

## EQU11 ## Q ₁ = ΔS ₁ + ΔS ₂ . here,

[Equation 12] ΔS ₁ = 10 (mm / sec) × 50 (μsec) = 0.5 (μ
m). In this embodiment, since the electromagnetic deflector 6 corrects the deviation amount of ΔS ₁ , the remaining deviation amount is ΔS ₂ .

Find ΔS ₂ . Differentiating the relative velocity curve reveals its slope α,

(13) α = 500π · cos (500π · t). a t = 0 when ^{α = 500π (mm / sec 2} ), height [Delta] S _2h triangle corresponding to [Delta] S ₂ is

[Expression 14] ΔS _2h = 50 (μsec) × 500π (mm / sec
² ) = 0.785 (mm / sec). Therefore, the deviation amount (area) of ΔS ₂ is

[Expression 15] ΔS ₂ = (1/2) × 50 (μsec) × 0.7
85 (mm / sec) = 0.01196, which is a large reduction in the deviation amount compared to the conventional case.

When it is desired to perform transfer with higher accuracy,
In consideration of the acceleration, the scanning speed of the electron beam may be corrected by the electrostatic deflector 7 as described above. A calculation example of the residual error remaining when the scanning speed is corrected will be described below.

When the acceleration is taken into consideration, FIG.
As shown in (c), since the acceleration changes greatly at t = 1 msec, the displacement amount becomes the largest at the time of transfer at t = 1 msec. As described above, the acceleration curve of FIG.

[Expression 16] α = 500π · cos (500π · t) Therefore, the slope β is

(17) β = − (500π) ² · sin (500π · t). As mentioned above, since t = 1 msec,

[Formula 18] β = − (500π) ² · sin (500π ·
1/1000) =-(500π) ² is obtained. From this, the height ΔV _h of the triangle ΔV is

[Formula 19] ΔV _h = 50 (μsec) × 500π ² (mm / sec
² ) = 123.4 (mm / sec ² ). Therefore, the area of triangle ΔV is

[Formula 20] ΔV = (1/2) × 50 (μsec) × 123.
4 (mm / sec ² ) = 3.1 (μm / sec). Therefore, since the scanning speed is deviated by 3.1 (μm / sec), the deviation amount is

[Expression 21] ΔV = 3.1 (μm / sec) × 50 (μsec) = 1.55 × 10 ⁻⁴ (μm) = 0.155 (nm), the amount of deviation is further reduced, and highly accurate transfer is possible. realizable.

In the above embodiment, the wafer 10 is the sample,
The speed calculators 17 and 21 constitute detection means, the CPU 18 constitutes deviation amount prediction means and relative movement speed prediction means, the electromagnetic deflection coil 6 constitutes transfer position correction means, and the electrostatic deflector 7 constitutes relative movement speed correction means. .

[0036]

As described above, according to the present invention, the moving speed of the mask and the sample in the specific direction is detected at the specific time before the transfer start time, and the detected moving speed and the moving speed are differentiated. The position of the pattern image with respect to the regular pattern transfer position and the relative movement speed of the sample and the pattern image are corrected based on the acceleration obtained as a result. Can be accurately transferred to the desired position.

[Brief description of drawings]

FIG. 1 is a diagram showing an outline of an apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a processing procedure by a CPU 18 of FIG.

FIG. 3 is a diagram showing a calculation example of a residual error when transfer is performed by the apparatus of FIG.

FIG. 4 is a diagram for explaining the principle of the present invention.

[Explanation of symbols]

4,101 mask 10 wafers 102 samples

─────────────────────────────────────────────────── ─── Continuation of front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) H01L 21/027 G03F 7/20

Claims (7)

(57) [Claims] 1. A charged particle beam transfer apparatus for irradiating a specific region of the mask with a charged particle beam to transfer a pattern formed in the region to the sample while continuously moving the mask and the sample from a transfer start time. A detection means for detecting the moving speed of the mask and the sample in a specific direction at a specific time before, and a regular pattern transfer position on the sample at the transfer start time and an image of the pattern guided on the sample. wherein a shift amount in a specific direction, and shift amount estimation means for estimating, based on the acceleration obtained by differentiating the moving speed and the moving speed detected by the detector, based on a prediction result of the shift amount predicting means , The mask
Adjusts the amount of deflection of the charged particle beam passing through the beam in the specified direction
The transfer position of the pattern actually transferred to the sample.
A charged particle beam transfer apparatus, comprising: a transfer position correction unit that corrects the transfer position. The charged particle beam transfer apparatus according to claim 1, wherein the shift amount predicting means, a moving speed at the specific time of the mask Vi; obtained by integrating the moving speed of the mask certain the mask position at time Si; the time difference between the particular time and the transcription start time Delta] t; a moving speed at the specific time of the sample Vs; at the specific time obtained by integrating the movement velocity of the sample The sample position is Ss; the reduction ratio of the pattern from the mask to the sample is k; the acceleration obtained by differentiating the moving speed Vi of the mask is αi; the acceleration obtained by differentiating the moving speed Vs of the sample α
s ;, the predicted value ΔS of the shift amount of the image of the pattern
ΔS = (k · Si−Ss) + [k · (Vi + αi · Δt) − (Vs + αs · Δ
t)] · Δt, which is a charged particle beam transfer apparatus. 3. A charged particle beam transfer apparatus for irradiating a specific region of the mask with a charged particle beam to transfer a pattern formed in the region to the sample while continuously moving the mask and the sample, from a transfer start time. Detection means for detecting a moving speed of the mask and the sample in a specific direction at a previous specific time, and a relative moving speed in the specific direction of the sample and an image of the pattern guided on the sample during transfer. and differentiating the moving speed and the moving speed detected by the detector
And the relative movement speed estimating means for estimating, based on the acceleration obtained Te, based on the prediction result of the relative movement speed estimating means, the specific direction of the image of the pattern to be directed onto the sample and the sample in the transfer The actual relative movement speed to the mask
The scanning speed of the charged particle beam passing through the
The charged particle beam transfer apparatus characterized by comprising: a relative moving speed correction means for correcting by section a. 4. A charged particle beam transfer apparatus according to claim 3, wherein the relative movement speed prediction means, the moving speed at the specific time of the mask Vi; obtained by differentiating the moving speed of the mask the ΔVi the variation of mask movement speed obtained by multiplying the time difference Δt between the particular time and transfer start time acceleration αi of the mask at a particular time; wherein the moving velocity at a particular time Vs of the sample; the ΔVs is the amount of change in the sample moving speed obtained by multiplying the acceleration αs of the sample at the specific time obtained by differentiating the moving speed of the sample by the time difference Δt between the specific time and the transfer start time; When the reduction ratio of the pattern to the sample is k ;, the predicted value ΔV of the relative movement speed is calculated as ΔV = (k · Vi−Vs) + (k · ΔVi−ΔVs). Charged particles beam transfer device. 5. The charged particle beam transfer apparatus according to claim 1, wherein the specific direction is a continuous movement direction of the mask and the sample. . 6. The charged particle beam transfer apparatus according to claim 1, wherein the specific direction is a direction orthogonal to a continuous moving direction of the mask and the sample. Particle beam transfer device. 7. A mask and a sample are continuously moved while being moved forward.
Irradiate a specific area of the mask with a charged particle beam to cover that area.
Charged particle beam transfer that transfers the formed pattern to the sample
In the copying method, use an interferometer attached to each of the mask and the sample.
And the mask at a specific time before the transfer start time
And the moving speed and acceleration of the sample in a specific direction
Obtained each, the regular pattern on the sample at the transfer start time
Between the transfer position and the image of the pattern guided on the sample
The amount of deviation in the specified direction is calculated at the specified time.
Prediction based on the moving speed and acceleration at the time of passing through the mask based on the predicted shift amount.
The charged particle beam is deflected in the specific direction, and
Particle beam transfer to correct the transfer position of the pattern transferred to the
Copy method.