JPH11274032A - Method of measuring characteristics of deflecting amplifier for charged beam drawing apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately measure settling characteristics of a deflecting amplifier used for electron beam drawing apparatus, and improve the drawing accuracy. SOLUTION: In this method of measuring the characteristics of a deflecting amplifier in an electron beam drawing apparatus, an electron beam is deflected by periodically changing the voltage of an deflecting amplifier 450, time T set from the rise of the output of the deflecting amplifier 450 is delayed, and the electron beam is moved in a direction different from a direction, where the electron beam is moved by the output change of the deflecting amplifier 450, in proportion to the delayed time T. By making the beam position a time axis, a short time electron beam is emitted for a time T by a blanking circuit 300. The beam is projected at a plurality of different points in time. Beam emission is performed a plurality of times by the blanking circuit 300 to the respective times, and a pattern for evaluation is drawn on a specimen 113.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a charged beam drawing apparatus, and more particularly to a method for measuring the characteristics of a deflection amplifier for a charged beam drawing apparatus for measuring the characteristics of a deflection amplifier for deflecting a beam.

[0002]

2. Description of the Related Art When a desired pattern is drawn on a sample using an electron beam drawing apparatus, it is necessary to turn on and off the electron beam and deflect the electron beam. In order to deflect the beam, a deflecting plate is arranged opposite to the beam axis, and an output from a deflection amplifier is applied to this deflector. The response of the deflection amplifier is relatively fast, but with the recent high-speed drawing, a time delay of the response of the deflection amplifier has become a problem.

The settling characteristics of a deflection amplifier of an electron beam writing apparatus can be obtained by measuring the output of the deflection amplifier using an oscilloscope. However, the measurement result in this case includes characteristics of the oscilloscope, and accurate measurement is difficult. It is also conceivable to draw a pattern on the resist of the sample and indirectly estimate the settling characteristics of the deflection amplifier. However, in this case, the measurement is indirect and accurate measurement is difficult.

[0004]

As described above, when the drawing speed in the electron beam drawing apparatus is increased, a time delay of the response of the deflection amplifier becomes a problem. Therefore, the settling characteristic of the deflection amplifier is measured, and this is measured. There is a need to correct. However, it has been difficult to accurately measure the settling characteristics of the deflection amplifier.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and has as its object to accurately measure the settling characteristics of a deflection amplifier used in a charged beam drawing apparatus, and to provide a drawing precision. It is an object of the present invention to provide a method for measuring the characteristics of a deflection amplifier for a charged beam writing apparatus, which can contribute to the improvement of the characteristics.

[0006]

(Structure) In order to solve the above-mentioned problem, the present invention employs the following structure.

That is, the present invention relates to a charged beam writing apparatus for writing a desired pattern on a sample by irradiating a charged beam, wherein the output of a deflection amplifier for driving a deflector is periodically changed to change the charged beam. Charged beam deflecting means for deflecting, delay means for delaying a set time T from rise or fall of the output of the deflection amplifier, and output of the deflection amplifier in proportion to the time T delayed by the delay means. The charged beam is moved in a direction different from the direction in which the charged beam is moved by the change, and a time axis generating means using the beam position as a time axis, and blanking for irradiating the electron beam for a short time T by the delay means Means for measuring the characteristics of the deflection amplifier, the time axis generating means performs the measurement at different times at a plurality of points. Irradiating a beam by King means,
In addition, beam irradiation by the blanking means is performed a plurality of times at each time, and an evaluation pattern is drawn on the sample.

When the present invention is applied to an electron beam drawing apparatus as an example, the following configurations (1) to (10) can be considered.

--- Measurement of deflection amplifier reset ring characteristics --- (1) Vertical (or horizontal) deflection amplifier output Vh (1), V
An electron beam deflecting means for periodically changing h (2) to deflect the electron beam, a delay means for delaying a set time T from the rise (or fall) of the deflection amplifier output, and a delay means The electron beam is moved in the horizontal (or vertical) direction in proportion to the delayed delay time T,
A time axis generating means having a time axis, and a blanking means for irradiating the electron beam deflected by the electron beam deflecting means with the electron beam for a short time at the time T by the delay means, and an electron to be drawn on a resist of a sample. In the (or charged) beam writing apparatus, the output of the deflection amplifier rises by moving the electron beam to X1 by the time axis generating means and irradiating the electron beam for a short time by the blanking means to the delay time T1 delayed by the delay means. (Or falling) is drawn on the resist of the sample a plurality of times, the electron beam is moved to X2 by the time axis generating means, and the delay time T is delayed by the delay means.
2. Irradiation of the electron beam for a short time by blanking means draws the rising (or falling) of the output of the deflection amplifier a plurality of times on the resist of the sample, and moves the electron beam to Xn by the time axis generating means. The rising (or falling) of the output of the deflection amplifier is drawn a plurality of times on the resist of the sample by irradiating the electron beam for a short period of time with the blanking means to the delay time Tn delayed by the above, and the vertical (or horizontal) deflection amplifier output Vh (1) plotting the transient output (rising or falling) of Vh (2) on the time axis T1 to Tn on the resist of the sample, and drawing the transient output of the deflection amplifier on the resist of the sample. A method for measuring the settling characteristics of an amplifier.

--- Measurement of Settling Characteristics of Multiple Deflection Amplifiers (Part 1) --- (2) In the above (1), the electron beam deflecting means is constituted by a plurality of vertical (or horizontal) deflection amplifier outputs. In the beam deflecting means, the outputs Vha (1) and Vha (2) of the specified vertical (or horizontal) deflection amplifier a are periodically changed (a: 1 to m), and the other specified (other than a) are not specified. The output of the vertical (or horizontal) deflection amplifier is Vh1, Vh2,
, Vhm, the electron beam is deflected, the electron beam is moved to X1 by the time axis generating means, and the deflection amplifier outputs other than a are output to Vh1 (T1), Vh2 (T1),.
(T1) and the delay time T1 delayed by the delay means
The rising (or falling) of the output of the deflection amplifier specified by irradiating the electron beam for a short time by the blanking means is drawn on the resist of the sample a plurality of times, and the electron beam is moved to X2 by the time axis generating means. , Vh1 (T2), Vh2 (T2),..., Vhm
(T2) and the delay time T2 delayed by the delay means
The rising (or falling) of the output of the deflection amplifier specified by irradiating the electron beam for a short time by the blanking means is drawn on the resist of the sample a plurality of times, and the electron beam is moved to Xn by the time axis generating means and specified. , Vh3 (Tn),..., Vh2 (Tn)
hm (Tn), and the rising (or falling) of the deflection amplifier output specified by irradiating the electron beam for a short time with the blanking means during the delay time Tn delayed by the delay means is applied to the resist of the sample a plurality of times. Draws and outputs vertical (or horizontal) deflection amplifier output Vha (1), Vha
(2) plotting the transient output (rising or falling) on the time axis T1 to Tn on the resist of the sample and drawing the transient output of the deflection amplifier on the resist of the sample; Measuring method.

--- Settling characteristic measurement of multiple deflection amplifiers (No. 2) --- (Method of setting unspecified deflection amplifier) (3) Vertical (or horizontal) deflection amplifier a specified in (2) above Electron beam deflecting means for periodically changing the outputs Vha (1) and Vha (2), setting the vertical (or horizontal) deflection amplifier outputs other than a to Vh1, Vh2,..., Vhm, and deflecting the electron beam. , The deflection amplifier output not specified at the delay time T1 is
(T1), Vh2 (T1),... The deflection amplifier output not specified at the delay time T2 is output to each Vh1
(T2), Vh2 (T2),... The deflection amplifier output not specified during the delay time Tn is output to each Vh1
(Tn), Vh2 (Tn), ..., Vh1 (T1), Vh1 (T2), ..., Vh1 (Tn) Vh2 (T1), Vh2 (T2), ..., Vh2 (Tn) ... ... Vhm (T1), Vhm (T2),..., Vhm (Tn) are given by: {Vha (1) + ΔVha (2) −Vha (1)} × exp (−t / C
R)｝ × K (t = T1, t = T2,..., T = Tn, where K is a constant determined by a deflection amplifier, CR is a time constant), and an unspecified deflection amplifier output is set, and vertical (or horizontal) deflection is performed. Amplifier output Vha (1), Vha
(2) plotting the transient output (rising or falling) on the time axis T1 to Tn on the resist of the sample and drawing the transient output of the deflection amplifier on the resist of the sample; Measuring method.

--- Reading method of settling characteristic measurement data --- (4) In (1), the resist of the sample on which the transient output of the deflection amplifier is drawn is raster-scanned with an electron beam, and the reflection generated by the raster scanning A method for measuring the settling characteristics of a deflection amplifier, comprising detecting (or absorbing, secondary) electrons, storing the detected output, and displaying the transient output of the deflection amplifier as the stored content.

--- Method for controlling minute irradiation time --- (5) In the blanking means of (1), a speed control means for controlling a speed at which an electron beam crosses a blanking plate is provided. A method for measuring the settling characteristics of a deflection amplifier, characterized in that a minute irradiation time of an electron beam is controlled by means, and a transient output of the deflection amplifier is drawn on a resist of a sample.

--- Settling characteristic adjustment method (correction pulse) --- (6) The rise of a specific deflection amplifier a periodically changed to Vha (1) and Vha (2) in (2) above. A correction pulse generating means for generating a correction pulse whose delay, amplitude, and width are controlled in synchronization with (or at the falling edge of) the correction pulse generated by the correction pulse generating means is input to the specific deflection amplifier a; The delay, amplitude and width of the correction pulse are used as parameters to adjust the settling characteristics of the deflection amplifier, and the transient output of the deflection amplifier a is superimposed and drawn on the resist of the sample. A method for measuring the settling characteristics of an amplifier.

--- Method of correcting amplitude dependency of settling characteristics --- Correction pulse generating means for generating a correction pulse whose delay, amplitude and width are controlled in synchronization with the rise (or fall) of the specific deflection amplifier a in which the amplitude of the output of the specific deflection amplifier a is changed as described above. The delay, amplitude and width of the correction pulse are controlled by the correction pulse generation means, and the delay, amplitude and width of the correction pulse are input to the specific deflection amplifier a and the settling time is minimized. The correction pulse data storage means for storing and storing the correction pulse data, the specific deflection amplifier output a is set to Vha (1) and Vha (1).
Delay Dr of correction pulse at rising of ha (2)
(2), amplitude Ar (2), width Wr (2) are stored, and Vh
The delay Df (2), the amplitude Af (2), and the width Wf (2) of the correction pulse at the time of falling of a (1) and Vha (2) are stored, and at the time of rising of Vha (1) and Vha (3). Correction pulse delay Dr (3), amplitude Ar (3), width Wr
(3) is stored, and the delay Df (3) and the amplitude Af of the correction pulse when Vha (1) and Vha (3) fall are stored.
(3), the width Wf (3) is stored,...
····· Dr (N), amplitude Ar (N) and width W of the correction pulse at the time of rising of Vha (1) and Vha (N)
r (N) is stored, and the delay Df (N) and the amplitude Af of the correction pulse at the fall of Vha (1) and Vha (N) are stored.
(N) and a width Wf (N), and a storage and reading means for reading out the delay, amplitude and width of the correction pulse stored by the storage means according to the amplitude of the deflection amplifier a, and a correction from the storage and reading means. Correction pulse generating means for generating a pulse; inputting the correction pulse generated by the correction pulse generating means to the specific deflection amplifier a; storing the amplitude dependence of the settling characteristic of the deflection amplifier a in the storage means. An electron beam writing apparatus, wherein the delay, the amplitude and the width of the corrected pulse are read by the reading means to correct the amplitude dependence of the settling characteristic of the deflection amplifier.

--- Method for Correcting Amplitude Dependence of Settling Characteristics (Sampling) --- (8) In the above (7), the deflection amplifier a has Vha (1) and Vha (1).
correction pulse data at ha (2), Dr (2), Df (2), Ar (2), Af (2), Wr (2), Wf (2) at Vha (1) and Vha (N) Correction pulse data: Dr (3), Df (3), Ar (3), Af (3), Wr (3), Wf (3) Vha (1) and Vha Measure part of the correction pulse data at the time of (N), Dr (N), Df (N), Ar (N), Af (N), Wr (N), Wf (N), and calculate the approximate value curve function. Electron beam writing characterized in that the settling characteristic of the deflection amplifier a is corrected by a correction pulse using data obtained by approximating N data of each of Dr, Df, Ar, Af, Wr and Wf of the correction pulse. apparatus.

--- Shaped deflection amplifier and beam shape at measurement of deflection amplifier --- (9) In the above (1) to (8), electron beam deflection is performed by the main and sub deflection amplifiers, and the beam forming vector shot is performed. An electron beam lithography system that draws by type, using a bar-shaped beam when measuring the settling characteristics of a shaping deflection amplifier and using a point beam when measuring the settling characteristics of a main / sub deflection amplifier. .

--- The optimization of settling characteristics is determined by CP
Operation with U --- (10) An electron beam lithography apparatus characterized in that the above (1) to (9) are controlled by a CPU.

(Operation) In the present invention, the time is set on the horizontal axis and the amplitude of the deflection amplifier is plotted on the vertical axis, and the evaluation pattern is plotted on the resist (X When measuring the amplifier, the vertical axis indicates time, and the horizontal axis indicates settling characteristics. This makes it possible to accurately measure the settling characteristics of the deflection amplifier from the obtained resist pattern.

At the time of measurement, the resist of the drawn sample is raster-scanned by a charged beam, the resulting reflected particles are detected and stored in a memory, and the stored memory is read out to display the settling characteristics of the deflection amplifier. . Further, by inputting the correction pulse to the deflection amplifier, the delay, amplitude and width of the correction pulse are adjusted so as to be optimum from the above-described display of the setting characteristics of the deflection amplifier (the delay and the width are fixed in the embodiment). Further, since the respective timings of the blanking amplifier, the shaping deflection amplifier, and the sub deflection amplifier are drawn by an actual electron beam drawing apparatus, the respective amplifiers can be adjusted with accurate timing.

[0021]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The details of the present invention will be described below with reference to the illustrated embodiments.

FIG. 1 is a schematic structural view of an electron beam drawing apparatus used in one embodiment of the present invention. Above the sample chamber 115, an electron gun 101, various lenses 102, 104,
107, 109, 111, various deflectors 103, 106,
An electron optical barrel 100 is provided which includes 110 and 112, first and second apertures 105 and 108 for forming an electron beam, aperture 116 and the like.

An electron beam extracted from the electron gun 101 passes through an electron lens 102 and a blanking electrode 103. The drawing information of the semiconductor device is stored in the pattern data generation circuit 200 from the computer 1000, and a blanking control signal is input to the blanking control circuit 300 by the pattern data generation circuit 200, and the blanking amplifier 3
On / off control of the electron beam passing through the blanking electrode 103 is performed by the output of 50.

The electron beam turned on is transmitted to the electron lens 10.
4 illuminates the first shaping aperture 105 and is shaped into a rectangular beam. The rectangular beam passes through shaped deflection electrode 106. The shaping deflection electrode 106 deflects the rectangular beam with the output of the shaping deflection amplifier 450. Pattern data generator 20
When 0, the shaping deflection control signal is input to the shaping deflection control circuit 400, the output of the shaping deflection amplifier 450 is applied to the shaping deflection electrode 106, and the rectangular beam is deflected.

The rectangular beam passing through the electron lens 107 is
Illuminate the second shaping aperture 108. The rectangular beam deflected by the shaping deflection amplifier 450 is masked by the shape of the second shaping aperture 108, and the shape is changed to a variable rectangular beam. The variable rectangular beam passes through the electron lens 109 and the aperture 116, and further passes through the main deflection electrode 110. The turned off electron beam is blocked by the aperture 116 and does not go below the aperture 116.

The sample stage 114 is moved in the X direction (left and right directions in the drawing) and the Y direction (front and back directions in the drawing) by the sample stage driving circuit 950 which receives a command from the computer 1000. The moving position of the sample stage 114 is measured by the laser length measuring circuit 900, and the length measuring information is sent to the main deflection control circuit 500 through the computer 1000.

The drawing information of the semiconductor device from the pattern data generating circuit 200 and the length measurement information of the sample stage 114 are compared by the main deflection control circuit 500, and the variable rectangular beam “irradiates the sample 115 on the sample stage 114. Is possible (drawing area)? ”Is detected by the main deflection control circuit 500. Signal that detected this drawing area (drawing area detection signal)
Are the blanking control circuit 300 and the molding deflection control circuit 4
00 is input to the sub deflection control circuit 600.

In accordance with the drawing area detection signal, the drawing information of the semiconductor device stored in the pattern data generating circuit 200 is converted into the blanking control circuit 300 and the shaping deflection control circuit 40.
The variable rectangular beam is input to the main deflection control circuit 500 and the sub deflection control circuit 600, and irradiates the sample 113 through the electron lens 111 and the sub deflection electrode 112. Then, the drawing information of the semiconductor device is irradiated on the sample 113, whereby a pattern of the semiconductor device is drawn.

On the other hand, this electron beam writing apparatus comprises a shaping deflection amplifier 450, a main deflection amplifier 550, and a sub deflection amplifier 6.
Measure and adjust the settling (settling) time to 50.
Settling measurement control circuit 700 by computer 1000
The electron beam is controlled by a short pulse whose delay is controlled by the blanking control 300 and the blanking amplifier 350. The shaping deflection control circuit 400, the main deflection control circuit 500, and the sub deflection Each amplifier is controlled by the control circuit 600 so that the electron beam
3 and the settling characteristics of the amplifier are measured using the backscattered electron detector 350 and the storage circuit 800.

The first shaping aperture 105 has a shape as shown in FIG. There is a rectangular hole 105a at the center, and this hole 105a is irradiated with the electron beam 150 uniformly. First
The electron beam 150 that has passed through the shaping aperture 105 becomes a uniform rectangular beam. As shown in FIG. 3, the shaped beam electrode 106 deflects the rectangular beam 151 to the position (A, B). The second shaping aperture 108 has an arrow shape 108a as shown in FIG. The rectangular beam 151 is masked by the arrow shape 108a and the second shaping aperture 108
, The shape of the electron beam becomes a right triangle 151a.

FIG. 4 is a diagram showing the shaping deflection electrode 106 and the shaping deflection amplifier 450 in detail. Shaped deflection electrode 106
Represents eight electrodes X, -X, Y, -Y, X + Y, -XY,
It is composed of YX and XY. Electron beam 151 shaped into a rectangular beam by first shaping aperture 105
Is deflected in the X-axis and Y-axis directions by the eight shaping deflection electrodes 106. Eight electrodes X, -X, Y, -Y, X +
Y, -XY, YX, and XY are formed by the deflection amplifier 45.
8 amplifiers X, -X, Y, -Y, X + Y, -X in 0
-Y, YX, and XY.

In order to make the rectangular beam 151 at the position (A, B) of the shaping deflection coordinates in order to form the triangular beam 151a, each amplifier input / output of the shaping deflection amplifier 450 is set as follows.

[0033] k and K are conversion coefficients from the digital input to the output voltage, and the unit is (V / BIT). K / k = 2 ^1/2 / 2,
If the DA converter (DAC) inside the amplifier is 16 bits and the maximum output of the amplifier is ± 20 V, then K = 0.610352 (mV / BIT) k = 0.431584 (mV / BIT)

To increase the size of the triangular beam 151a, the position of the rectangular beam 151 is moved only in the + X direction. Conversely, to decrease the size, the position is moved in the -X direction.

FIG. 5 shows that four types of triangular beams (151a to 151b) and one
It shows that a type of rectangular beam 151e is generated.
The rectangular beam 151 is set to (A, B), and the triangular beam 151a is formed by the second forming aperture 108,
By changing the X coordinate from A to C, the Y-axis symmetric triangular beam 151b is formed. The Y coordinate is changed from B to D and B to E, and X-axis symmetric triangular beams 151c and 151d are formed. By setting the coordinates of the rectangular beam 151 to (F, G), a rectangular beam 151e is formed.

FIG. 6A shows coordinates (F) of the rectangular beam 151.
From (1, G1) to (F1, G2), the rectangular beams 151e1 to 151
It is a principle figure which changes a dimension to e2 (broken line). Ideally, this dimensional change ends at the same time as the coordinate change of the rectangular beam 151, but a time delay occurs due to the settling characteristics of the shaping deflection amplifier 450. The following operation is performed to accurately measure the settling characteristic.

In order to periodically change the coordinates (F1, G1) and (F1, G2) of the rectangular beam 151, the settling measurement control circuit 700 gives the shaping / deflection control circuit 400 the coordinates (F1, G1) of the rectangular beam 151. G1) and (F1, G
2) are alternately transmitted, and a signal for irradiating the electron beam 150 for a short time is sent to the blanking control circuit 300 in synchronization with the change of the coordinates (F1, G1) and (F1, G2) of the rectangular beam 151. To the main deflection control circuit 500, a signal for deflecting to a drawing position first is sent (the sub deflection control circuit 600 remains unchanged).

The coordinates (F1, G1) and (F1, G2) of the rectangular beam 151, its period is 80 ns, and the sampling point 2
0, number of repetitions 10, irradiation time 2 ns
0, set value cycle 80ns and sampling point 2
From 0 to 80 nS is 0.8 mm (−400 μm to 350
μm), the time axis corresponding to 5 μS for 50 μm is calculated by the computer 1000, and the settling measurement control circuit 700
Set to. The coordinates (F1, G1) and (F1, G2) of the rectangular beam 151 are sent to the shaping / deflection control circuit 400, and the height of the rectangular beam changes at a cycle of 80 ns. It is sent to the main deflection control circuit 500.

The main deflection control circuit 500 is set to −400 μm, and the blanking control circuit 300 is set to a rectangular beam 15.
The irradiation start signal 317 is sent at the same time as the coordinates (F1, G1) of 1 change to (F1, G2), and the 2 nS electron beam is irradiated. This is repeated 10 times. Next, a signal of −350 μm is sent to the main deflection control circuit 500 to synchronize with a change in coordinates (F1, G1) to (F1, G2) of the rectangular beam 151, and after 5 ns, an irradiation start signal 317 is sent.
This is repeated 10 times. Similarly, the main deflection control circuit 5
At 00, -300 μm, 10 nS delay, and 2 nS electron beam irradiation are performed 10 times. Finally, the main deflection control circuit 500
At 50 μm, the blanking control circuit 300 is delayed by 75 nS and irradiated with 2 nS electron beams 10 times. The above is summarized as follows.

[0040] As described above, the shaping / deflecting amplifier 450 draws (set ring) characteristics for changing (F1, G2) from the coordinates (F1, G1) of the rectangular beam 151 on the resist of the sample 113.
The number of times of electron beam irradiation required for this drawing is 200 times (the number of times of irradiation depends on the type of resist and the current density of the electron beam), and the drawing time is 20 μS + the deflection time of main deflection (about 32
0 μS).

FIG. 7 is a block diagram of the blanking control circuit 300. The delay setting value DL is input from the settling measurement control circuit 700 to the register S1 (308) from the input 301.
The irradiation time set value ST is input from the input 302 to the register S2.
Set to (314). The flip-flop F1 (309) is set by the irradiation start signal 317, and the gate G1 is set.
(305) is opened, CK303 from the clock pulse generator 304 is supplied to the D counter DCR306, and
306 counts up. The comparator DCM307 compares the upper bit of the register S1 storing the delay set value DL with the content of the DCR306, and when the DCR306 has the same value as the upper bit of the register S1, the DCM307
Outputs a pulse.

This output is connected to a programmable delay device D1.
(310) delays the value of the lower bit of the register S1 (301). By the output of D1 (310), D1
CR306 and F1 (309) are reset, the output of flip-flop DFF is inverted, and flip-flop F
2 (315) is set. F2 (315) is G2 (3
11) Open and set CK303 to irradiation time counter PCR3
Count up 12. The upper bits of the register S2 are compared with the contents of the PCR 312 by the comparator PCM313, and when the contents of the PCR 312 and the upper bits of the register S2 become equal, the PCR 313 outputs a pulse.

This output is connected to a programmable delay device D2 (3
16) and is delayed by the value of the lower bit of the register S1. D2 (316), F2 (315), PCR
(312) is reset. F2 (315), PCR3
12, PCM313, S2 (314), D2 (316)
Operates on a signal from the pattern data generator 200.
The selector S3 (31) is set by the settling measurement signal 321.
8) The DFF 319 output is selected, and the blanking control circuit output 320 is sent to the blanking amplifier 350.

FIG. 8 is a configuration diagram of a blanking amplifier. Output 320 of FIG. 7 is connected to input 356 of FIG.
The voltage is supplied to the negative input of the OP amplifier 355 through the resistor 352 and is balanced. The settling measurement signal 321 of FIG.
Is connected to the input 357 of FIG.
Power supply E, and the OP amplifier 35 through the resistor 357.
5 is supplied to the-input to bias the OP amplifier output 358 in the negative voltage direction. 358 of output of OP amplifier
Are connected to the blanking electrode 103.

Irradiation control for drawing a pattern is performed by O
When the P-amplifier output 358 is zero, the electron beam
Is irradiated. Since the irradiation of the electron beam for the settling measurement is for a short time (several nS), the OP amplifier 355 does not respond, and it is difficult to perform the control with the normal electron beam irradiation. Therefore, a bias is applied to the OP amplifier 355, and the OP amplifier output 358 changes from a positive output voltage to a negative voltage output, and the electron beam is irradiated on the sample 113 for a moment when the OP amplifier output becomes zero, and the OP amplifier output 358 is There is no non-response due to the short pulse. In order to slightly change the short-time electron beam irradiation, the electron beam irradiation time can be extended by inclining the voltage of the OP amplifier output 358.

FIG. 9 is a time chart of the output of the blanking amplifier 350 and the irradiation amount of the electron beam. (1) is a blanking method at the time of normal pattern drawing. When the electron beam is irradiated, the electron beam is irradiated while the output of the blanking amplifier 350 is between the + threshold and the -threshold. The irradiation time is set in the irradiation time register S2 (314) in FIG.
Is determined by (2) and (3) are blanking methods at the time of settling measurement, and (2) is a blanking amplifier 350.
The electron beam irradiation for a short time is obtained by biasing the power supply E359 so that the output of passes between the + threshold and the −threshold instantaneously. (3) shows a method for extending the electron beam irradiation time of (2). In this case, the output time of the electron beam is extended by giving the output inclination of the blanking amplifier 350.

FIG. 10 shows the input / output of the shaping deflection amplifier 450 and the timing of electron beam irradiation (2 ns). First 10 times with no delay, 2nd 10 times with 5 ns delay, 3
The first is 10 ns for 10 times, and the last is 10 ns for 75 ns.
Each time, the resist of the sample 113 is irradiated with an electron beam.

FIG. 11 shows that 0n
A state in which a change in dimension is drawn in a time series to a rectangular beam 151e2 at a time of 75 nS from S is shown. The resist of the sample 113 can be developed and observed with a microscope, an SEM, or the like.

To shorten the observation time, the settling characteristic drawn on the resist of the sample 113 is raster-scanned by the main deflection control circuit 500 and the main deflection amplifier 550. FIG. 12 shows this state, and FIG.
3 is shown.

The settling characteristic drawn on the resist of the sample 113 is negatively charged at the position irradiated with the electron beam. Due to the raster scanning, electrons are strongly reflected in a negatively charged place, the reflected electron signal is detected by the reflected electron detector 850, and the detection signal is stored in the AD in the storage circuit 800.
AD conversion is performed by the converter 801, and the converted value is stored in the memory 802. By reading the signal stored in the memory 802 and passing it through the DA converter 803,
By changing to an analog output and modulating the brightness of a CRT display or the like with this signal, the same pattern as that shown in FIG. 11 is displayed on the display, and the resist of the sample 113 is developed. 450 settling characteristics can be observed.

A shaping deflection amplifier 450 taking a rectangular beam as an example
Explained the settling characteristics of the triangular beam 1
The same applies to 51a to 51d.

The shaping deflection amplifier 450 is composed of eight amplifiers. The above settling characteristic is obtained by combining eight amplifiers. If the settling characteristic of even one of the amplifiers deteriorates, the combined settling characteristic also deteriorates. A method for specifying a deteriorated amplifier will be described.

The change of the rectangular beams 151e1 to 151e2 is based on the coordinates (F1, G1) of the rectangular beams 151 and (F1,
With the coordinate change of G2), the Y coordinates G1 and G2 change. The shaping deflection amplifier 450 corresponding to this change is Y, -Y, X
+ Y,-(X + Y), YX,-(YX). The amplifier whose settling characteristics have been measured is denoted by Y. In the above manner, G1 and G2 are alternately input to the Y amplifier. −Y,
The X + Y, − (X + Y), Y−X, − (Y−X) amplifiers set the fixed value after 5 ns in which G1 is set in each amplifier. Next, a fixed value after 10 nS is set for each amplifier, and similarly after 15 nS,...

[0054] The inputs of the X and -X amplifiers do not change to F1 and -F1, respectively.

The -Y amplifier uses -G1-A for measurement after 5 ns.
A fixed value of -G1-A2 is set for the measurement after 1, 10 nS, and a fixed value of -G-A19 (= -G2) is set for the measurement after 75 nS.
Similarly, except for the Y amplifier, 5 nS, 10 nS,.
A fixed value is set for the measurement after S.

A1 to A19, B1 to B19, C1 to C1
The value of 9 is: AN = (G2−G1) EXP (−5N / CR1) BN = (F2−F1 + G2−G1) EXP (−5N / C
R2) CN = (G2-G1 + F1-F2) EXP (-5N / C
R3) N: 1 to 19 CR1 to 3: Fixed values determined by the time constant (nS) of the amplifier.

That is, pseudo data is input during the settling period except for the Y amplifier, and the settling characteristic of the Y amplifier alone can be measured by simulating the settling characteristic.

The shaping deflection amplifier 450 includes X, -X, -Y,
It is composed of X + Y,-(X + Y), YX,-(YX) amplifiers. Each amplifier has a function to improve settling characteristics. This function can be digitally controlled from the outside, and the settling characteristics drawn on the resist of the sample 113 can be measured to obtain the optimum settling characteristics.

FIG. 14 shows a configuration example of a Y amplifier as an example of a function for improving the settling characteristic. The DAC 452 captures the digital input 451 (G1) by the ENB pulse 453, and generates an output voltage e1. The output voltage e1 is supplied to the negative input of the OP amplifier 457 through the resistor 459, and balances the current supplied from the output e3 (k × G1) of the OP amplifier 457 to the negative input through the resistor 460. Further, the DAC output voltage e1 passes through the resistor 461,
The OP amplifier 457 is supplied to the resistors 463 and 465 and
It balances the current supplied from output e3 (k × G1) through resistor 465.

The resistor 463, the feedback capacitor 464 and the OP
The amplifier 458 forms an integrator, integrates the voltage at the point A,
It is supplied to the + input of the OP amplifier 457. The digital input 451 is alternately supplied with G1 and G2 in a cycle of 80 ns, and the output voltage e3 alternately outputs k × G1 and k × G2.

The correction pulse generator 454 has a digital input 4
51 (G1) captures (G2) by the ENB pulse 453, and calculates the difference G1-G2 (or G2)
-G1) is calculated. This digital input difference (G1-G
2) is used as an address of the internal RAM. G of RAM
The 1-G2 address stores the amplitude of the correction pulse e2 in advance. Data is read out from the address G1-G2 of the RAM, supplied to the correction DAC, and the correction pulse e
2 is generated.

FIG. 15 shows a timing chart of the Y amplifier and an output waveform of the Y amplifier. DAC452 is ENB
The digital input 451 (G2) is taken in by 453.
DAC output e1 shows a waveform changed from input G1 to G2. The broken line of the amplifier output e3 is a waveform when the correction pulse e2 is not supplied, and becomes a solid line waveform by supplying the correction pulse e2. By changing the amplitude of the correction pulse, the settling characteristics of the amplifier output e3 can be controlled.

The details of the settling correction are publicly known (Japanese Patent Application No. 7-56088, DAC amplifier). Also,
The configuration of the Y amplifier is the same for amplifiers other than the Y amplifier.

FIG. 16 shows a configuration for reading and writing the contents of the internal RAM 473 of the correction pulse generator 454. Each amplifier has a device address 478 and a digital input 45
1 has a pass function. The input of the DAC 452 and the input / output of data of the correction pulse generator 454 are also used. There is a microprogram counter (MPC) 484 for controlling the microprogram inside the amplifier. Control lines (TRG482, CP477, C
LR483).

First, MPC 484 is reset by CLR 483 and digital input 451 is applied to difference address 470
Is input to It is ready to use as a normal amplifier. M
The clock of the PC 484 is the control TRG 482, and the clear is the CLR 483. The TRG 482 is supplied to the MPC 484, the MPC 484 is set to 1, and the MPC1 decoder 481
Enables the address matching circuit 485. The device address is put on the digital input 451, and it is checked by the address matching circuit 485 whether the device address 478 set in the amplifier matches (synchronization with the CP 477). If the result of the check is a match, the address match circuit 485 sets the MPC
2 decoder 479 and MPC3 decoder 480 to CLR4
Enable until 83 comes.

The TRG 482 is supplied to the MPC 484,
Set PC484 to 2 and enable MPC2 decoder 479. Data of digital input 451 is converted to CP477.
Supplies the contents of the RAM address 471 to the address of the RAM 473.

The TRG 482 is supplied to the MPC 484,
Set PC484 to 3 and enable MPC3 decoder 480. The data put on the digital input 451 is CP
The data is taken into RAM data 475 by 477 and written into RAM 473 through gate circuit 476.

Since the difference in the amplitude of the Y amplifier is G2-G1, the address of the RAM 473 of the correction pulse generator 454 is set to G2-G1, the contents of the RAM are written, and the RAM is written.
By supplying the contents of 473 to the correction DAC 474, the amplitude e2 of the correction pulse can be controlled. The amplitude e2 of the correction pulse is adjusted, and the waveform of the Y amplifier is drawn on the resist of the sample 113 by the above operation, raster-scanned, and stored in the storage circuit 800. The waveform can be read from the recording circuit 800 to the display, the waveform can be confirmed, and the same operation can be performed by changing the drawing position of the resist on the sample 113 until the setting characteristic of the Y amplifier is optimized. If the X and Y coordinate axes and necessary ruled lines are simultaneously drawn, it is easy to observe.

Performing the above operation on all the RAM 473 addresses of the correction pulse generator 454 requires a lot of time. Since the relationship between the amplitude e2 of the correction pulse and the difference between the digital input 451 is a continuous relationship and there is no discontinuity,
By sampling several points and performing curve approximation, RAM473 data of the correction pulse generator 454 can be collected in a short time, and the data can be written to the RAM473.

The shaping deflection amplifier 450 is composed of eight pieces. It takes a lot of time to individually set these eight amplifiers to optimal settling characteristics. Therefore, by adjusting two amplifiers of X (or -X) and Y (or -Y) from the eight amplifiers, the entire settling characteristic is optimally adjusted. That is, although the settling characteristics of the individual amplifiers are not the optimum settling characteristics, the adjustment time of the amplifier is shortened by setting the optimum settling characteristics as a whole.

The measurement of the settling characteristic of the sub deflection amplifier 650 is almost the same as that of the shaping deflection amplifier 450. A square beam of about 1 μm square is formed by the shaping deflection amplifier 450. The configuration of the sub-deflection electrode 112 is the same as that of the shaped deflection electrode 106, and eight electrodes X, -X, Y, -Y, X + Y, -X-
It is composed of Y, YX and XY. The electron beam 151 shaped into a square is applied to the eight sub deflection electrodes 112.
Is deflected in the X-axis and Y-axis directions. Eight electrodes X,-
X, Y, -Y, X + Y, -XY, YX, XY
Eight amplifiers X, -X, Y,-in the sub deflection amplifier 650
Y, X + Y, -XY, YX, XY.

The square beam 15 is applied to the resist of the sample 113.
In order to set 1 to the position (A, B) of the sub deflection coordinate, each amplifier input / output of the sub deflection amplifier 650 is set as follows.

[0073] ks and Ks are conversion coefficients from the digital input to the output voltage,
The unit is (V / BIT). If there is a relationship of Ks / ks = 2 ^1/2 / 2, and the DA converter (DAC) inside the amplifier is 16 bits and the maximum output of the amplifier is plus or minus 10V, then ks = 0.215792 (mV / BIT) Ks = 0.305176 (mV / BIT).

The coordinates (A1, B1) of the square beam 151
In order to periodically change (A1, B2) and (A1, B2), the settling measurement control circuit 700 sends coordinates (A1, B1) and (A1, B2) to the sub deflection control circuit 600 alternately, and the blanking control circuit The coordinates (A1, B1) and (A
1, B2) Synchronize with the change, and for a short time, square beam 15
A signal for irradiating 1 is sent. Main deflection control circuit 5
At 00, a signal for deflecting to the first drawing position is sent. A signal for deflecting to the first drawing position is sent to the main deflection control circuit 500.

The coordinates (A1, B1) of the square beam 151
And (A1, B2), its cycle 80 ns, sampling point 20, number of repetitions 10, irradiation time 2 ns.
0, set value cycle 80ns and sampling point 2
From 0 to 80 nS is 0.8 mm (−400 μm to 350
μm), the time axis corresponding to 5 μS for 50 μm is calculated by the computer 1000, and the settling measurement control circuit 700
Set to.

The coordinates (A1, B1) of the square beam 151
And (A1, B2) are sent to the sub-deflection control circuit 600, the square beam changes in the Y-axis direction at a period of 80 ns, and the time axis sends 20 sample positions of 50 μm pitch to the main deflection control circuit 500. The main deflection control circuit 500 is set to −400 μm, and the blanking control circuit 300 sends an irradiation start signal 317 in synchronization with a change from the coordinates (A1, B1) to (A1, B2) of the square beam 151, and sends 2 ns electrons. Irradiate the beam. This is repeated 10 times.

Next, the main deflection control circuit 500 receives -350 μm.
m, and sends the coordinates (A1, B
After 5 ns after changing from 1) to (A1, B2),
An irradiation start signal 317 is sent. This is repeated 10 times. Similarly, the main deflection control circuit 500 is supplied with -300 μm, 1
0 ns delay, 2 nS electron beam irradiation 10 times. Finally, the main deflection control circuit 500 is irradiated with 350 μm, and the blanking control circuit 300 is irradiated with 75 ns delay and 2 nS electron beam irradiation 10 times. The above is summarized as follows.

[0078] As described above, the sub deflection amplifier 650 draws (settling) characteristics for changing (A1, B2) from the coordinates (A1, B1) of the square beam 151 on the resist of the sample 113.
The number of times of electron beam irradiation required for this drawing is 200 times (the number of times of irradiation depends on the type of resist and the current density of the electron beam), and the drawing time is 20 μS + the deflection time of main deflection (about 32
0 μS).

FIG. 17 shows that 0n
At 75 ns from S, the square beam 151 has coordinates (A
This shows a state in which a change in position is drawn in chronological order from (1, B1) to (A1, B2). The resist of the sample 113 can be developed and observed with a microscope, an SEM, or the like.

In order to shorten the observation time, the settling characteristic drawn on the resist of the sample 113 is raster-scanned by the main deflection control circuit 500 and the main deflection amplifier 550, and the reflected electron signal is detected by the reflected electron detector 850. Then, the detection signal is converted to A by the AD converter 801 in the storage circuit 800.
D conversion is performed, and the converted value is stored in a memory. Reading the stored signal to display the same pattern as in FIG. 17 on the display is the same as observing the settling characteristic of the shaping deflection amplifier 450. If the X and Y coordinate axes and necessary ruled lines are simultaneously drawn, it is easy to observe.

The sub deflection amplifier 650 is composed of eight amplifiers. The above settling characteristic is obtained by combining eight amplifiers. The method of specifying one of the amplifiers and measuring the settling characteristic is the same as that of the shaping deflection amplifier 450. The sub deflection amplifier 650 is
X, -X, Y, -Y, X + Y,-(X + Y), YX,
-(YX) amplifier. Like the shaping deflection amplifier 450, each amplifier has a function of improving settling characteristics. This function can be digitally controlled from the outside, and the settling characteristics drawn on the resist of the sample 113 can be measured to obtain the optimum settling characteristics.

Since the configuration of the sub deflection amplifier 650 has the same function as the configuration of the shaping deflection amplifier 450, the amplitude e2 of the correction pulse is adjusted, the waveform of the amplifier is drawn on the resist of the sample 113, and raster scanning is performed. The data is stored in the storage circuit 800. The waveform can be read out from the recording circuit 800 to the display, the waveform can be checked, and the same operation can be performed by changing the drawing position of the resist on the sample 113 until the setting characteristic of the amplifier is optimized.

The method of sampling the RAM 473 address of the correction pulse generator 454 in the shaping deflection amplifier 450, performing curve approximation, and writing data to all addresses is the same in the sub deflection amplifier 650.

The method of adjusting the entire settling characteristic optimally by adjusting two amplifiers X (or -X) and Y (or -Y) from the eight amplifiers of the shaping deflection amplifier 450 is a subordinate. The same applies to the deflection amplifier 650.

The measurement of the settling characteristic of the main deflection amplifier 550 is similar to that of the shaping deflection amplifier 450 and the sub deflection amplifier 650. A square beam of about 1 μm square is formed by the shaping deflection amplifier 450. The sub deflection amplifier 650 is fixed (for example,
All amplifiers zero). The configuration of the main deflection electrode 110 is the same as that of the shaping deflection electrode 106 and the sub deflection electrode 112, and eight electrodes X, -X, Y, -Y, X + Y, -XY, YX, X
-Y.

The electron beam 151 shaped into a square is
The eight sub-deflecting electrodes 112 deflect in the X-axis and Y-axis directions. Eight electrodes X, -X, Y, -Y, X + Y,-
XY, YX, and XY are 8 in the main deflection amplifier 550.
Amplifiers X, -X, Y, -Y, X + Y, -XY, Y
−X, XY.

The square beam 15 is applied to the resist of the sample 113.
In order to set 1 to the position (A, B) of the main deflection coordinate, each amplifier input / output of the main deflection amplifier 550 is set as follows.

Digital input Output voltage X amplifier Xa km × Xa −X amplifier −Xa −km × Xa Y amplifier Yb km × Yb −Y amplifier −Yb −km × Yb X + Y amplifier (Xa + Yb) Km × (Xa + Yb )-XY Ann-(Xa + Yb)-Kmx (Xa + Yb) YX amplifier (Yb-Xa) Kmx (Yb-Xa) XY amplifier-(Yb-Xa)-Kmx ( Yb-Xa) km and Km are conversion coefficients from digital input to output voltage,
The unit is (V / BIT). There are relationships Km / km = 2 ^1/2 2, inside the amplifier DA converter (DAC) at 20 bits, if the maximum output of the amplifier is 200V, km = 0.2697397 (mV / BIT) Km = 0 .3814697 (mV / BIT).

The coordinates of the square beam 151 (−400 μm)
In order to periodically change (m, B1) and (−400 μm, B2), the settling measurement control circuit 700 sends the coordinates (−400 μm, B1) and (−400) to the main deflection control circuit 500.
400 μm, B2) are sent alternately, and the coordinates (−400 μm, B1) and (−400 μm) are sent to the blanking control circuit 300.
m, B2) Synchronize with the change and short-time the square beam 15
A signal for irradiating 1 is sent.

The coordinates of the square beam 151 (−400 μ,
B1) and (−400 μm, B2), the cycle of which is 80 ns,
Sampling point 20, number of repetitions 1, irradiation time 20n
S computer 1000 is set, and 80 nS is set to 0.8 mm (−40) from the set value cycle 80 nS and sampling point 20.
(0 μm to 350 μm), the computer 1000 calculates a time axis corresponding to 5 nS to 50 μm, and sets the calculated time axis in the settling measurement control circuit 700.

The coordinates of the square beam 151 (−400 μm)
m, B1) and (−400 μm, B2) are sent to the main deflection control circuit 500, and the square beam changes in the Y-axis direction at a cycle of 80 ns, and the blanking control circuit 300 sends the coordinates (−400 μm) of the square beam 151 to the blanking control circuit 300. , B1) to (−4
00 m, B2), and the irradiation start signal 317
And irradiate it with a 20 nS electron beam.

Next, a pulse having a width of 20 nS is sent 5 μS after the main deflection coordinate of the square beam 151 changes from (−350 μm, B1) to (−350 μm, B2).
Main deflection coordinates (-300 m, B1), (-300 m,
B2) An electron beam irradiation of 20 nS is performed with a delay of 10 μS. Finally, the main deflection coordinates (350 μm, B1), (35
In synchronization with the change of 0 μm, B2), the blanking control circuit 300 is irradiated with a 20 nS electron beam with a delay of 75 μS.

The above is summarized as follows.

[0094] As described above, the main deflection amplifier 550 draws, on the resist of the sample 113, the characteristic of changing (setting) the main deflection Y coordinate B1 of the square beam 151 from B1 to B2. The number of times of electron beam irradiation required for this drawing is 20 times (the number of times of irradiation depends on the type of resist and the current density of the electron beam), and the drawing time is 0.4 mS + the moving time of the sample stage drive circuit 950 (about 1
Complete with S).

FIG. 18 shows that 0n is formed on the resist of the sample 113.
At 75 ns from S, the square beam 151 has the main deflection Y
A state in which a change in position from coordinates B1 to B2 is drawn in time series is shown. The resist of the sample 113 is developed and the microscope, S
It can be observed by EM or the like. If the X and Y coordinate axes and necessary ruled lines are simultaneously drawn, it is easy to observe.

In order to shorten the observation time, the settling characteristic drawn on the resist of the sample 113 is raster-scanned by the main deflection control circuit 500 and the main deflection amplifier 550, and the reflected electron signal is detected by the reflected electron detector 850. Then, the detection signal is subjected to AD conversion by the AD converter 801 in the storage circuit 800.
Convert and store the converted value in memory. Displaying the same pattern as in FIG. 18 on the display by reading the stored signal is the same as observing the settling characteristics of the shaping deflection amplifier 450 and the sub deflection amplifier 650.

The main deflection amplifier 650 is composed of eight amplifiers. The above settling characteristic is obtained by combining eight amplifiers. The method of specifying one of the amplifiers and measuring the settling characteristics is the same as that of the shaping deflection amplifier 450 and the sub deflection amplifier 650.

The main deflection amplifier 550 has X, -X, Y,-
It is composed of Y, X + Y,-(X + Y), YX,-(YX) amplifiers. As with the shaping deflection amplifier 450 and the sub deflection amplifier 650, each amplifier has a function of improving settling characteristics. This function can be digitally controlled from the outside, and the settling characteristics drawn on the resist of the sample 113 can be measured to obtain the optimum settling characteristics.

The configuration of the main deflection amplifier 550 has the same function as the configuration of the shaping deflection amplifier 450 and the sub deflection amplifier 650. Therefore, the amplitude e2 of the correction pulse is adjusted, and the waveform of the amplifier is drawn on the resist of the sample 113. Raster scanning,
The data is stored in the storage circuit 800. The waveform was read from the recording circuit 800 to the display, the waveform was confirmed, and the same operation was performed on the sample 1.
By changing the drawing position of the resist 13, the settling characteristics of the amplifier can be adjusted until it becomes optimal.

The method of sampling the RAM 473 address of the correction pulse generator 454 in the shaping deflection amplifier 450, performing curve approximation, and writing data to all addresses is the same as in the main deflection amplifier 550. X (or -X), Y from the eight amplifiers of the shaping deflection amplifier 450
The method of adjusting the overall settling characteristics optimally by adjusting the two amplifiers (or -Y) is the same for the main deflection amplifier 550.

Next, measurement of the tracking circuit will be described. The main deflection amplifier 550 is the same as the shaping deflection amplifier 450 and the sub deflection amplifier 650 as described above, but is incorporated in a deflection amplifier for tracking the position of the sample stage 114. The tracking circuit is X155
1, Y1555, X + Y1558, Y-X1561
X tracking circuit 1551 outputs X and -X
The X output is input to the X amplifier 1552, and the −X output is input to the −X amplifier 1554. The Y tracking circuit 1555 has two outputs, Y and -Y. The Y output is input to the Y amplifier 1556, and the -Y output is input to the -Y amplifier 1557.

The output of the X + Y tracking circuit 1558 is X
There are two, + Y and-(X + Y). The X + Y output is input to the X + Y amplifier 1559, and the-(X + Y) output is input to the-(X + Y) amplifier 1560. YX output is YX amplifier 1
To 562, the-(YX) output is input 1563 to the-(YX) amplifier. FIG. 19 shows the configuration of a main deflection amplifier incorporating this tracking circuit. I155 in the figure
A positive / negative inversion circuit 3 outputs -X when X is input, for example.

FIG. 20 shows an X main deflection amplifier (X in FIG. 19).
(Amplifier) 1552. Forming deflection amplifier 450
Of the tracking input e4 (56
6) is different. e4 is the resistance 565 and the resistance 56
The ratio of 8 (resistor 560 and resistor 567) becomes the amplification factor and becomes the output e3.

E3 = − (resistance 565) × e4 / (resistance 568) For the tracking input e4, the correction pulse generator 55
4 is not used. Input 551 is X main deflection input, 553 is E
NB pulse is a signal of X main deflection input.

FIG. 21 shows the configuration of the tracking circuit. A digital input 571 is provided by a DA converter (DAC) 572 and a tracking correction pulse generator 5 by the ENB 573.
74 is input. The output of DAC 572 is current i1,
It is connected to the-input of the OP amplifier 577 and connected to the output of the OP amplifier 577 through the resistor 576 to generate e3. The tracking correction pulse generator 574 includes a resistor 575
To the minus input of the OP amplifier 577.
The tracking correction pulse generator 574 generates a pulse for canceling the glitch of the DAC 572.

FIG. 22 is a timing chart of the tracking circuit. The digital input 571 inputs -1 and 0 alternately, and the ENB 573 synchronized with the digital input 571.
Generates a glitch on the DAC output i1. This glitch is inverted in phase and has a waveform like the output e3 of the OP amplifier 577. Since this glitch waveform has an adverse effect on pattern writing, the tracking correction pulse generator 574
An output e2 is generated to cancel the glitch. The output e3 of the OP amplifier 577 after the correction is a glitch-free output.

FIG. 23 shows a tracking correction pulse generator 5.
74 shows the configuration of FIG. Tracking digital input 580
Is temporarily stored in the difference address 582 by the ENB 581,
Through the selector 584, the amplitude RAM 585a and the delay RAM
585b, and the correction DAC 58
6 and the programmable delay unit 598, and the DAC 57
2 generates an output e2 for canceling the glitch of the output i1.

The amplitude RAM 585a and the delay RAM 585b
Here's how to read and write the contents of Each amplifier has a device address 588, and the digital input 580 has a pass function. The input of the DAC 572 and the input / output of data of the tracking correction pulse generator 574 are also used. There is a microprogram counter (MPC) 597 for controlling the microprogram inside the amplifier. Control lines (TRG595, CP587, CLR) for controlling the MPC597.
596).

Initially, MPC 597 is reset by CLR 596 and digital input 580 is applied to difference address 582
Is input to It is ready to use as a normal amplifier. M
The clock of the PC 597 is the control TRG 595, and the clear is the CLR 596. The TRG 595 is supplied to the MPC 597, the MPC 597 is set to 1, and the MPC1 decoder 594 is supplied.
Enable the address matching circuit 593. The device address is put on the digital input 580, and it is checked by the address matching circuit 593 that the device address matches the device address 592 set in the amplifier (synchronized by CP587). If the result of the check is that the addresses match, the address matching circuit 593 sets the MPC
2 decoder 589 and MPC3 decoder 591 to CLR5
Enable until 96 is reached.

When TRG595 is supplied to MPC597,
PC597 is set to 2 and MPC2 decoder 589 is enabled. Data of digital input 580 is transferred to CP587
Into the RAM address 583, and the selection circuit 58
4, the contents of the RAM address 584 are stored in the RAM 585
a, 585b. T to MPC597
RG595 is supplied, MPC597 is set to 3, MPC3
Enable the decoder 591. Digital input 58
The data put on 0 is stored in RAM data 59 by CP587.
0, and through the gate circuit 588, the amplitude RAM 5
85a and 585b.

The amplitude of the correction pulse is adjusted by the data of the amplitude RAM 585a, and the delay of the correction pulse is adjusted by the data of the delay RAM 585b.
A correction pulse e2 for canceling the glitch of 1 is generated. The waveform of this tracking circuit is
, And raster-scanned, and stored in the storage circuit 800. The waveform can be read from the recording circuit 800 to the display, the waveform can be checked, and the same operation can be performed by changing the drawing position of the resist on the sample 113 until the glitch of the tracking circuit is minimized. If the X and Y coordinate axes and necessary ruled lines are simultaneously drawn, it is easy to observe.

The glitch characteristic measurement of the tracking circuit is performed by
This is performed through the main deflection amplifier 550. Forming deflection amplifier 45
0 produces a square beam of about 1 μm square. The main deflection side of the sub deflection amplifier 650 main deflection amplifier is fixed (for example, all amplifiers are zero). In order to set the square beam 151 on the resist of the sample 113 at the position (A, B) of the main deflection coordinate, each amplifier input / output of the tracking circuit is as follows.

[0113] kl and Kl are conversion coefficients from the digital input to the output voltage,
The unit is (V / BIT). There are Kl / kl = 2 ^1/2 / 2 relations, the main deflection amplifier 550 equal kl = 0.2697397 (mV / BIT) Kl = 0.3814697 (mV / BIT) becomes the value.

In order to periodically change the tracking coordinates (A1, B1) and (A1, B2) of the square beam 151, the glitch measurement control circuit 700 sends the main deflection coordinates (−400 μm, 0) to the main deflection control circuit 500. ), And the coordinates (A1, B1) and (A1, B1,
B2) are alternately transmitted, and a signal for irradiating the square beam 151 for a short time is sent to the blanking control circuit 300 in synchronization with the changes in the tracking coordinates (A1, B1) and (A1, B2).

The tracking coordinates (A1, B1) and (A1, B2) of the square beam 151, the period of which is 80 ns,
Sampling point 20, repetition number 10, irradiation time 2n
S computer 1000 is set, and 80 nS is set to 0.8 mm (−40) from the set value cycle 80 nS and sampling point 20.
(0 μm to 350 μm), the computer 1000 calculates a time axis corresponding to 5 nS to 50 μm, and sets it in the settling measurement control circuit 700.

The tracking coordinates (A1, B1) and (A1, B2) of the square beam 151 and the main deflection coordinates are (-4).
00 μm, 0) is sent to the tracking circuit and main deflection amplifier of the main deflection control circuit 500, and the square beam changes at a cycle of −400 μm in the X-axis direction and 80 ns in the Y-axis direction,
An irradiation start signal 317 is sent to the blanking control circuit 300 in synchronization with the change of (A1, B2) from the tracking coordinates (A1, B1) of the square beam 151, and irradiation with the 2nS electron beam is repeated 10 times.

Next, the main deflection coordinate is set to (−350 μm, 0)
And the tracking coordinates (A1,
After 5 ns after the change from (B1) to (A1, B2),
The 2 nS electron beam irradiation is performed 10 times, the tracking coordinates are changed, and the main deflection coordinates (−300 μm, 0), 10
A delay of μS and electron beam irradiation of 2 nS are performed 10 times. Finally, the main deflection coordinates (3
50 μm, 0), 75 n in the blanking control circuit 300
S delay, 2nS electron beam irradiation are performed 10 times. The summary is as follows.

[0118] As described above, the characteristic that the main deflection amplifier 550 changes (glitch) the tracking Y coordinate B1 of the square beam 151 from B1 to B2 by the tracking circuit is drawn on the resist of the sample 113. The number of electron beam irradiations required for this drawing is 2
00 times (the number of irradiations increases / decreases depending on the type of the resist and the current density of the electron beam), and the writing time is completed by 20 μS + the deflection time of the main deflection (about 320 μS).

FIG. 24 shows that 0n is formed on the resist of sample 113.
At a time of 75 ns from S, the square beam 151 changes its position from the tracking coordinates B1 to B2 (B1-B2 is ±
1 LSB) in a time-series manner. Sample 113
Can be developed and observed with a microscope, SEM, or the like. If the X and Y axis coordinates and the necessary ruled lines are drawn simultaneously, it is easy to observe. Since the tracking circuit operates during writing, glitching of the DA converter in the tracking circuit poses a problem.

The main deflection control circuit 500 and the main deflection amplifier 550 perform raster scanning of the settling characteristics drawn on the resist of the sample 113 to shorten the observation time.
The backscattered electron signal is detected by the backscattered electron detector 850, and the detected signal is detected by the A / D converter 801 in the storage circuit 800.
D conversion is performed, and the converted value is stored in a memory. By reading out the stored signal and displaying the same pattern on the display as in FIG. 24, the forming deflection amplifier 450, the sub deflection amplifier 650, the main deflection amplifier 5
This is the same as observing 50 settling characteristics.

The tracking circuit is composed of X, Y, X + Y, Y
-X. There is a function to cancel glitches of the DA converter in the tracking circuit. This function can be digitally controlled from the outside, and the glitch characteristics drawn on the resist of the sample 113 can be measured to obtain the optimum glitch characteristics.

The amplitude e2 of the correction pulse is adjusted, a glitch waveform is drawn on the resist of the sample 113, raster-scanned, and stored in the storage circuit 800. The waveform can be read from the recording circuit 800 to the display, the waveform can be confirmed, and the same operation can be performed by changing the resist drawing position of the sample 113 until the glitch characteristics of the amplifier are optimized.

The method of sampling the RAM 473 address of the correction pulse generator 454 in the shaping deflection amplifier 450, approximating the curve, and writing data to all the addresses is the same in the tracking circuit.

[0124]

As described above in detail, according to the present invention, the settling characteristic of the deflection amplifier of the charged beam drawing apparatus, D
By writing a pattern on the resist of the sample with the AC glitch characteristic, the measurement of the settling characteristic and the glitch characteristic can be performed under the same conditions as the writing, whereby the highly accurate settling characteristic and glitch characteristic can be measured.
Based on this measurement, the settling characteristics and glitch characteristics of the deflection amplifier can be optimized using the correction pulse in the amplifier. Therefore, the timings of the blanking amplifier, the shaping deflection amplifier, the sub deflection amplifier, the main deflection amplifier and the like of the charged beam drawing apparatus can be accurately adjusted. In addition, all the above operations are performed under the control of the CPU, so that high-precision adjustment is possible, and the time can be reduced. Further, the settling characteristics and glitch characteristics of a specific amplifier can be measured, and an amplifier having deteriorated settling characteristics and glitch characteristics can be quickly found.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of an electron beam writing apparatus used in an embodiment of the present invention.

FIG. 2 is a diagram showing a shape of a first forming aperture.

FIG. 3 is a diagram showing a shape of a second shaping aperture and a triangular beam.

FIG. 4 is a diagram showing a shaping deflection electrode 106 and a shaping deflection amplifier 450 in detail.

FIG. 5 is a view showing beams of various shapes.

FIG. 6 is a view for explaining the principle of dimensional change of a rectangular beam.

FIG. 7 is a configuration diagram of a blanking control circuit.

FIG. 8 is a configuration diagram of a blanking amplifier.

FIG. 9 is a timing chart of a blanking amplifier output and an electron beam irradiation amount.

FIG. 10 is a timing chart of input / output of a shaping deflection amplifier and electron beam irradiation.

FIG. 11 is a diagram showing a state in which settling characteristics of a forming deflection amplifier are drawn on a resist.

FIG. 12 is a principle diagram of raster scanning of a pattern drawn on a resist.

FIG. 13 is a configuration diagram of a memory circuit.

FIG. 14 is a configuration diagram of a shaped deflection Y amplifier.

FIG. 15 is a timing chart of a shaping deflection Y amplifier.

FIG. 16 is a configuration diagram of a correction pulse generator in a shaping deflection amplifier.

FIG. 17 is a view showing resist drawing of settling characteristics of a sub deflection amplifier.

FIG. 18 is a view showing resist drawing of settling characteristics of a main deflection amplifier.

FIG. 19 is a configuration diagram of a tracking circuit and a main deflection amplifier.

FIG. 20 is a configuration diagram of a main deflection amplifier.

FIG. 21 is a configuration diagram of a tracking circuit.

FIG. 22 is a timing chart of a tracking circuit.

FIG. 23 is a configuration diagram of a tracking correction pulse generator.

FIG. 24 is a view showing resist writing of glitch characteristics of a tracking DAC.

[Explanation of symbols]

 105: first shaping aperture 108: second shaping aperture 113: sample 1000: computer 200: pattern data generating circuit 300: blanking control circuit 350: blanking amplifier 400: shaping deflection control circuit 450: shaping deflection amplifier 500: main deflection Control circuit 550: Main deflection amplifier 600: Sub deflection control circuit 650: Sub deflection amplifier 800: Storage circuit 850: Backscattered electron detector

Claims (1)

[Claims] 1. A charged beam writing apparatus for writing a desired pattern on a sample by irradiation of a charged beam, wherein the output of a deflection amplifier for driving a deflector is periodically changed to deflect the charged beam. Beam deflecting means;
Delay means for delaying a time T set from the rise or fall of the output of the deflection amplifier; and a direction in which the charged beam moves due to a change in the output of the deflection amplifier in proportion to the time T delayed by the delay means. A charged beam drawing apparatus comprising: a time axis generating means for moving a charged beam in a direction different from that of the charged beam and a time axis based on the beam position; In the apparatus, the beam is irradiated by the blanking means at a plurality of different times by the time axis generating means, and the beam irradiation by the blanking means is performed a plurality of times for each time, for evaluation on the sample. A method for measuring the characteristics of a deflection amplifier for a charged beam lithography apparatus, characterized by drawing a pattern.