JP2004072123A - Electron beam exposure system and its operating method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve the throughput of an electron beam exposure system and its operating method by shortening the time required for deflection and exposure without degrading the exposing accuracy when the deflection by a main deflector is transferred to a region in which exposure is possible by a sub-deflector at a high speed by using a feedback deflector. <P>SOLUTION: The election beam exposure system, comprising a main deflection settling feedback circuit for extracting a difference between an output of a D-A converter for converting the amount of deflection of the main deflector from digital to analog and a voltage applied to the main deflector and deflecting the position of the electron beam to the position at which exposure is possible by the sub-deflector based on the extracted output, is provided with a main deflector settling wait time control section. <P>COPYRIGHT: (C)2004,JPO

Description

The present invention relates to an electron beam exposure apparatus and an operation method thereof, and particularly to an electron beam exposure apparatus having high exposure accuracy and high throughput and an operation method thereof.

2. Description of the Related Art In recent years, the degree of integration and functions of semiconductor integrated circuit devices have been more and more improved, and the role as a core technology of technological progress over the entire industry such as computer, communication, or machine control is expected, and semiconductors that play this role are expected. The degree of integration of integrated circuit devices has increased four times over two to three years, and in the case of DRAM (dynamic random access memory), up to 1 Gbit has been prototyped. The improvement in degree is due to the progress of microfabrication technology.

When electron beam exposure is used as a fine processing technology, fine processing of 0.05 μm or less can be realized with an alignment accuracy of 0.02 μm or less, but the throughput of such a fine pattern drawing process is extremely low. It was thought that it could not be used for mass production of VLSI (ultra large scale integrated circuit device).

However, such a prediction is merely a discussion of a one-stroke electron beam using an electron beam exposure apparatus that has been conventionally marketed.
A conventional electron beam exposure apparatus forms an arbitrary pattern by scanning an electron beam on a sample using a variable rectangular beam and connecting rectangular patterns.

Such an apparatus is a pattern generator that creates an actual pattern from soft pattern data.However, since a pattern of an arbitrary shape is generated by connecting rectangular patterns, the finer the pattern rule, the more the unit area per unit area There is a disadvantage that the number of exposure shots increases and the throughput decreases.

To overcome such limitations, the present inventors have proposed a block exposure method or a multi-beam blanking aperture array (BAA) exposure method (for example, see Patent Documents 1 to 3). With this proposal, a throughput of about 1 cm □ / sec can be expected, and the fineness, alignment accuracy, quick turnaround, and reliability are all superior to other methods. is there.

Among them, the blanking aperture array (BAA) exposure method uses a light source, that is, an electron beam from an electron gun through a blanking aperture array (BAA) controlled by a blanking control unit to have an arbitrary shape. Then, it is deflected by a main deflector and a sub deflector, and irradiated onto an exposure sample mounted on a moving stage controlled by a moving stage controller.

This blanking aperture array (BAA) has a plurality of rectangular windows, each of which can be opened and closed one by one, arranged in a grid pattern. An arbitrary electron beam pattern composed of a rectangular electron beam block is formed, and the electron beam pattern is irradiated on a sample to perform exposure.

In such an exposure method, the deflection time by the main deflector is reduced, and a pre-read calculation process for correcting the main deflector, a feedback deflector (FBC: feedback coil) of the main deflector, and With the proposal of the main differential settling feedback circuit (MSF), which is a control circuit, the settling waiting time, which took about 50 μsec for the conventional 100 μm deflection by the main deflector, becomes about 3 μsec. Exposure can be started by the sub deflector in 3 μsec.

Here, a method of operating the conventional electron beam exposure apparatus will be described with reference to FIGS.
Referring to FIG. 8, at the time of electron beam exposure, an electron beam from a light source, that is, an electron gun, is formed into an arbitrary shape via a blanking aperture array (BAA) controlled by a blanking control unit, and then the main beam is formed. The light is deflected by the deflector and the sub deflector, and is irradiated onto the exposure sample placed on the moving stage controlled by the moving stage controller.

In the deflection of the main deflector in this case, the main deflection data from the main deflector controller is converted into an analog quantity by a main deflector DAC (digital-analog converter), and then the input resistance R _in 12 and the feedback resistance R _f. 13, and, together applied to the coil of the main deflector through the comparison amplifier 11 having a output resistor R _o 14, voltage and main deflector that is actually applied to the main deflector by the main differential settling feedback circuit By detecting the difference from the output of the DAC and applying this output to a feedback deflector having a small deflection amount but capable of following at high speed, the deflection by the main deflector is brought to the exposure range by the sub deflector. The apparent settling time is reduced to about 3 μsec.

The gain / rotation calculation unit 16 indicated by an arrow in the main differential settling feedback circuit is provided to correct a deviation from an electron beam due to a coil winding, a coil installation position, a coil characteristic, or the like.

Next, when the deflection by the main deflector becomes the exposure range by the sub deflector, the sub deflection data from the sub deflector control unit is converted into an analog quantity by the sub deflector DAC, and then the sub deflection is transmitted via the amplifier (AMP). Exposure is started by applying to the deflector, and when the exposure is completed, the exposure end signal is output from the sub deflector controller, and the same operation is repeated by blowing the electron beam to the next exposure area by the main deflector. .

In this case, the timing of each signal is controlled by the main deflector settling waiting time control unit. The main deflector setting waiting time control unit calculates the jump amount of the main deflector from the main deflection data. The output of the waiting time flag generating circuit, that is, the waiting time flag is sent to the sub deflector. Input to start exposure.

When exposure is performed by an electron beam exposure apparatus having such a main differential settling feedback circuit, when the exposure pattern data by the sub deflector is extremely sparse, the settling wait time of the main deflector is, for example, 3 μsec. If the exposure cycle by the deflector is, for example, 10 MHz (10 ⁷ Hz, 10 ⁻⁷ = 100 n) and the exposure pattern is, for example, 10 shots, the time required from the start of deflection by the main deflector to the end of exposure is
3 μsec + (100 nsec × 10 shots) = 4 μsec
Thus, an improvement in throughput can be expected.

In addition, in such a multi-beam blanking aperture array exposure method, it has been proposed to adopt a continuous movement exposure method and to shorten a settling time of a main deflector, thereby improving a throughput. Therefore, this proposal will be described with reference to FIG.

Referring to FIG. 10, first, a voltage based on main deflection data for largely deflecting the electron beam is applied to a position at which exposure can be started by the sub deflector to the main deflector, and deflection by the main deflector is started. In this case as well, a large settling wait time of about 50 μsec for stabilizing the main deflector is required as in the conventional case.

Since the sample is continuously moving during this large settling time, the position of the sample is largely deflected after the large settling time by just deflecting the first main deflection data. Means that the electron beam is not deflected to a position where the exposure by the electron beam becomes possible.

Therefore, during the large settling wait time, the movement of the sample is read periodically, and correction deflection data corresponding to the movement amount is applied to the main deflector to correct the deflection position of the electron beam. In the case of a proper correction, a minimum settling wait time, that is, a corrected settling wait time C is required, and is, for example, about 3 μsec.

In such a case, the overall large settling wait time WAIT1 set as a target is the absolute value ΔX of the difference between the value before deflection in the X direction and the value to be deflected and the difference between the value before deflection and the value to be deflected in the Y direction. By comparing the absolute value ΔY with the absolute value, the larger absolute value is set as the flying amount ΔG, and further, an offset B of about 3 μsec is set, and the setting waiting time calculation is performed.

In this calculation, the overall large settling wait time WAIT1 set as the target is:
WAIT1 = ΔG × A + B
It is represented by the following equation.
Note that A and B are preset coefficients, which do not change at least during exposure of the sample. Here, when the movement of the sample is read periodically and correction deflection data corresponding to the movement amount is applied to the main deflector to correct the deflection position of the electron beam, the correction settling wait time C is equal to the WAITC shown in FIG. Will appear periodically in the form.

As described above, since the deflection of the main deflector is started during the movement of the sample and the deviation due to the movement is corrected by the periodic correction, the time required for the entire exposure process is shortened, and the throughput is improved.
JP-A-06-084772 JP-A-06-224112 JP 05-062890 A

Here, problems in the case of the conventional blanking aperture array (BAA) exposure method will be described with reference to FIGS.
Looking at the trajectory of the electron beam on the exposure sample with reference to FIG. 11, the electron beam has arrived at a predetermined position from (1) to (5) with the lapse of time by the feedback deflector, and 4 μsec from the start of deflection by the main deflector. Later (settling waiting time: 3 μsec, exposure: 100 nsec × 10 shots = 1 μsec), it appears that the deflection by the next main deflector is possible, but as is apparent from the trajectory of the main deflector, the deflection by the main deflector actually occurs. The deflection has not been swung to a predetermined position.

When the exposure data is extremely sparse, for example, in the case of exposure of 10 shots, at the time point (4) 1 μsec after the start of the exposure, the actual position of the beam by the main deflector is the position of the exposure start. When the next deflection is performed by the main deflector at the time of (4), the output of the comparison amplifier of the main differential settling feedback circuit does not reach the predetermined value applied to the main deflector. Is input, and the difference from the output of the main deflector DAC does not become an accurate difference, so that there is a disadvantage that the exposure accuracy is extremely deteriorated.

Referring to FIG. 12, for example, when the exposure data is dense to some extent, as shown in the upper part of FIG. 12, after deflecting to a predetermined position by a feedback deflector in about 3 μsec, exposure is started by a sub deflector, and about 50 μsec. After the exposure, the same operation is repeated to expose the entire area.

However, when the exposure data is sparse, as shown in the lower part of FIG. 12, the exposure ends early, and immediately after the end of the exposure, if the main deflector is to be deflected to the next exposure area, The actual deflection position is different from the exposure start position by d, and when the value of d is large, that is, when the exposure data is extremely sparse and the exposure ends very early, the required deflection is required. Since the amount exceeds the deflection range of the feedback deflector, exposure at an accurate position becomes impossible.

In order to prevent the exposure accuracy from deteriorating when the exposure data is extremely sparse, it is necessary to wait for the main deflector to settle for a sufficient time. In this case, even in the case of a field where the exposure data amount is dense, the exposure by the sub deflector after the end of the long settling wait time is started.

In this case, when the settling wait time of the main deflector is, for example, 50 μsec, and the exposure cycle by the sub deflector is, for example, 4 μm / 1 shot at 10 MHz for the 100 μm field, the entire field is exposed from the start of deflection by the main deflector. The time needed to do
50 μsec + (100 μm □ ÷ 4 μm □ × 100 nsec) = 112.5 μsec
As shown in the upper part of FIG. 13, the time when exposure by the sub deflector is performed 3 μsec from the start of deflection of the main deflector,
50 μsec + (100 μm □ ÷ 4 μm □ × 100 nsec) − (50 μsec−3 μsec) = 65.5 μsec
There is a disadvantage that it is greatly increased as compared with.

In the second equation, the first term is the minimum settling time, that is, the minimum stay time, the second term is the actual exposure time, and the third term is the settling time in the settling time. It is the time to cover the exposure, but in the following equations, to simplify the explanation,
3 μsec + (100 μm □ ÷ 4 μm □ × 100 nsec) = 65.5 μsec
Notation like.

Again, see FIG. 10 Also in the continuous moving exposure method, the periodic reading cycle is an arbitrary cycle with respect to the whole large settling wait time WAIT1 set as a target, and therefore is always an integer multiple of the integer. Waiting time WAIT and WAIT1 are not the same,
WAIT = ΔG × A + B + α
Is represented by

Therefore, when the actual settling time WAIT of the main deflector WAIT, WAIT1, and the reading cycle happen to coincide,
WAIT = ΔG × A + B (α = 0)
, And there is no problem in the efficiency of the settling wait time, but when the deviation between WAIT1 and the reading cycle is the largest, assuming that the corrected settling wait time is C,
WAIT = ΔG × A + B + C
And has the disadvantage of being the least efficient actual settling wait time.

Accordingly, the present invention requires an electron beam exposure apparatus that uses a feedback deflector to quickly bring deflection by a main deflector to an exposure range of a sub deflector without deteriorating exposure accuracy. It is an object of the present invention to shorten the time required to improve the throughput.

Here, means for solving the problem in the present invention will be described with reference to FIG.
See FIG. 1 (1) The present invention extracts the difference between the output of a DA converter that converts the amount of deflection of the main deflector from a digital amount to an analog amount and the voltage applied to the main deflection, and extracts the extracted output. In an electron beam exposure apparatus having a main de-settling feedback circuit for deflecting the position of an electron beam to an exposure-possible position of a sub-deflector on the basis of A minimum stay time including a part of the exposure time by the deflector is set, and after the minimum stay time and after the exposure of the sub deflector is completed, control is performed so that the next main deflector can be deflected. A main deflector settling waiting time control unit is provided.

Thus, in the multi-beam blanking aperture array electron beam exposure apparatus, when deflecting by the main deflector, apart from the settling wait time of the main deflector, a part of the exposure time by the sub deflector is also included. By setting the minimum stay time, even if the exposure data is extremely sparse, after the exposure is completed, the deflection by the next main deflector is performed at least after waiting for the minimum stay time, thereby preventing the deterioration of the exposure accuracy. Can be.

(2) Further, the present invention is characterized in that, in the above (1), at the time of deflection by the main deflector, counting of the minimum stay time and exposure by the sub deflector are started immediately after the settling wait time of the main deflector ends. And

As described above, at the time of deflection by the main deflector, by counting the minimum stay time and starting exposure by the sub deflector immediately after the end of the settling wait time of the main deflector, deterioration of exposure accuracy can be prevented. In addition, it is possible to reduce the average time required for exposing the entire area.

(3) According to the present invention, in the above (1), at the time of deflection by the main deflector, counting of the settling time of the main deflector and counting of the minimum stay time are simultaneously started, and the settling time of the main deflector is also started. The exposure by the sub deflector is started after the end.

In this way, at the time of deflection by the main deflector, the counting of the settling wait time of the main deflector and the counting of the minimum stay time are started simultaneously, and the exposure by the sub deflector is started after the end of the settling wait time of the main deflector. As a result, the required time when exposure is sparse can be shortened without sacrificing exposure accuracy, so that the average required time required for exposure of the entire region can be further reduced.

(4) In the present invention according to the above (1), at the time of deflection by the main deflector, counting of the settling time of the main deflector and counting of the minimum stay time are simultaneously started, and the setting of the main deflector is waited. Exposure by the sub deflector is started after the end of the time, and immediately after the end of the exposure, deflection by the next main deflector is started. After the settling time and after the end of the previous minimum stay time, the next deflector is used. Is started.

In this way, at the time of deflection by the main deflector, the counting of the settling waiting time of the main deflector and the counting of the minimum stay time are simultaneously started, and the exposure by the sub deflector is started after the end of the settling waiting time of the main deflector, Immediately after the end of the exposure, the next main deflector starts deflecting, and after the settling wait time and after the end of the previous minimum stay time, the next exposure by the sub deflector is started, so that the exposure data is sparse. In such a case, the waiting time for the next main deflector to settle the deflection can be further reduced.

(5) In the present invention, in the above (4), means for comparing the next settling wait time by the main deflector with the remaining time of the minimum staying time of deflection by the previous main deflector and the initial setting wait time. , And a longer time between the two.

In this way, by providing the means for comparing the next settling wait time of deflection by the main deflector with the remaining time of the minimum stay time of deflection by the previous main deflector and the initial settling wait time, Time control becomes easier.

(6) Further, according to the present invention, a difference between an output of a DA converter for converting a deflection amount of a main deflector from a digital amount to an analog amount and a voltage applied to a main deflection is extracted, and the extracted output is output. A main de-settling feedback circuit for deflecting the position of the electron beam to an exposing position of the sub-deflector based on the main deflector; The main deflector settling time for controlling to enable the next main deflector to deflect after the minimum stay time and after the exposure of the sub deflector is completed after setting the minimum stay time including the part. A method of operating an electron beam exposure apparatus that performs exposure while continuously moving the sample, wherein the end of the correction stabilization waiting time accompanying the correction deflection by reading the amount of movement of the sample determines the entire deflection by the main deflector. Settling As the end previous Chi time, and wherein the determining the read timing of the sample on the basis of the remaining amount of the overall settling time.

As described above, in the case of performing exposure while continuously moving the sample using the above-described electron beam exposure apparatus, the end of the correction settling time accompanying the correction deflection by reading the movement amount of the sample is settled by the main deflector. By determining the readout timing of the sample based on the remaining amount of the entire settling time so that it is before the end of the waiting time, the waste due to the last correction settling time immediately before the end of the entire settling time is eliminated. be able to.

(7) In the present invention, in the above (6), the entire settling time is calculated as a digital number from the flying amount of the main deflector, and this digital number is counted by a down counter to wait for the settling time. The value of this down counter is used as the address of the read timing determination memory, and data for starting reading is stored at an address a fixed time before the end of the entire settling wait time, and the read correction immediately before the end of the entire settling wait time is performed. The end of the correction settling time accompanying the deflection is set to be before the end of the entire correction wait time.

In this way, the entire settling time is calculated as a digital number from the amount of flight of the main deflector, and this digital number is counted by the down counter to wait for the settling time, and the value of the down counter is read to determine the timing for reading timing. By storing data for starting reading at an address a fixed time before the end of the entire settling wait time, the last reading timing can be automatically determined without waste.

(8) Further, according to the present invention, in the above (7), the read timing determination memory is provided with a time function for changing the read cycle immediately before the end of the entire settling wait time and the other read cycle. It is characterized by.

As described above, the read timing determination memory is provided with a time function for changing the read cycle immediately before the end of the entire settling wait time and the other read cycles, thereby reducing the number of times of reading.

(9) Further, according to the present invention, in the above (7), a memory having, as an address, an absolute value of the deflection jump amount obtained by the main deflector with respect to the X axis and the Y axis is provided. It is characterized by storing a settling waiting time map depending on the flight direction.

As described above, a memory having, as an address, an absolute value of the amount of deflection of the deflection by the main deflector with respect to the X axis and the Y axis is provided, and a stabilization waiting time map depending on the amount of flight and the direction of flight is stored in this memory. By doing so, the end of the settling wait time can be automatically determined by counting the data of the settling wait time map.

(10) Further, according to the present invention, in the above (9), when the absolute value of the deflection jump amount by the main deflector is small, the settling wait time is a constant value, and otherwise, it is proportional to the absolute value of the jump amount. It is characterized by storing a settling time map, which is settling time.

(4) As described above, when the absolute value of the amount of deflection by the main deflector is small, by setting the fixed waiting time to a constant value, it is possible to ensure good deflection accuracy when the amount of flight is small.

(11) The present invention is characterized in that, in the above (9), the settling waiting time depending on the direction of the deflection by the main deflector is set differently in the X-axis direction and the Y-axis direction.

As described above, by setting the settling time depending on the direction of the deflection by the main deflector so as to be different between the X-axis direction and the Y-axis direction, the stage movement characteristics and the characteristics of the amplifier in the X-axis direction and the Y-axis direction are different. If different, the correction can be made freely.

According to the present invention, since the minimum stay time control mechanism is provided in addition to the settling waiting time control mechanism, high-speed deflection using the main de-settling feedback circuit and the feedback deflector can be performed without deteriorating the exposure accuracy, and the throughput can be reduced. Can be improved.

In the case of performing exposure while continuously moving the sample using this electron beam exposure apparatus, the end of a large settling wait time for deflection by the main deflector and the correction settling time associated with the last movement reading just before the end of the large settling wait time. Since the end of the time can be made substantially coincident, waste in the actual settling waiting time can be eliminated, and the throughput can be improved.

In the present invention, a difference between an output of a DA converter for converting a deflection amount of a main deflector from a digital amount to an analog amount and a voltage applied to a main deflection is extracted, and an electron beam is extracted based on the extracted output. In the electron beam exposure apparatus having the main de-settling feedback circuit for deflecting the position of the sub-deflector to the exposing position of the sub-deflector, when deflecting by the main deflector, the exposure time The main deflector setting which controls so that deflection by the next main deflector becomes possible after the minimum stay time and after the end of the exposure of the sub deflector is completed. A waiting time control unit is provided to prevent deterioration of exposure accuracy even when exposure data is extremely sparse by performing deflection by the next main deflector after waiting for at least the minimum stay time after exposure is completed. That.

Further, in the case of performing exposure while continuously moving the sample using such an electron beam exposure apparatus, the end of the correction settling time accompanying the correction deflection by reading the movement amount of the sample is the time for the entire deflection settling by the main deflector. By determining the sample readout timing based on the remaining amount of the entire settling wait time so that it is before the end of the time, waste due to the last correction settling time immediately before the end of the entire settle wait time is eliminated.

Here, a first embodiment of the present invention will be described with reference to FIGS.
FIG. 1 shows a control mechanism of the electron beam exposure apparatus common to the first to third embodiments. In particular, FIG. 1 is a diagram for explaining in detail a main deflector settling waiting time control unit. The mechanism and operation mechanism are the same as those of the conventional electron beam exposure apparatus having the feedback deflector (FBC) described with reference to FIG.

The feature of the electron beam exposure apparatus of the present invention is that the main deflector settling time control unit has a minimum staying time other than a waiting time arithmetic circuit, a waiting time counter circuit, and a waiting time flag generation circuit for the settling time control. The minimum stay time control circuit for setting is provided.

In this minimum stay time control circuit, the count of the minimum stay time is controlled by a set trigger signal from the main deflector control unit and an output (wait time flag) from the wait time flag generation circuit. For example, 50 μsec, which is a settling waiting time necessary for deflection of only the main deflector, is adopted.

In this case, if the setting is made to start counting the minimum stay time simultaneously with the set trigger signal from the main deflector control unit, at the time of deflection by the main deflector, the minimum stay time is counted simultaneously with the counting of the settling wait time of the main deflector. Counting is started, and even when the exposure data is extremely sparse, the minimum stay time flag is not set unless the exposure end flag from the sub deflector is input and the minimum stay time is not ended. Since no output is made, the exposure accuracy does not deteriorate.

When the count of the minimum stay time is controlled by the output (wait time flag) from the wait time flag generation circuit, the count of the minimum stay time is started after the end of the settling wait time (for example, 3 μsec). However, even when the exposure data is extremely sparse, the minimum stay time flag is not output after the exposure end flag is input from the sub deflector and after the minimum time, so that the exposure accuracy is deteriorated. I can't.

See FIG.
FIG. 2 is an explanatory diagram of the counting of the settling wait time, the counting of the minimum stay time, and the timing of exposure in the first embodiment of the present invention. In the first embodiment, the control of the settling wait time of the main deflector and the minimum stay time are performed. A counter is provided for time control, and starts counting the settling time of the main deflector simultaneously with the start of deflection of the main deflector, and starts counting the minimum stay time and exposure by the sub deflector after the settling time is over. . Then, after the end of the minimum stay time and the end of the exposure, the deflection operation by the next main deflector is repeated to expose the entire area.

Here, when the exposure data is dense, that is, for example, when the exposure cycle is 10 MHz and the 100 μm square field is filled with one shot at 10 MHz by the sub deflector, the start of the deflection by the main deflector and the start of the exposure by the sub deflector are performed. The time required for the termination is set to a minimum stay time of 50 μsec (a waiting time for the main deflector to stabilize when the correction by the main deflector feedback circuit and the feedback deflector is not included) and a waiting time for the main deflector to settle to 3 μsec. If
3 μsec + (100 μm □ ÷ 4 μm □ × 100 nμsec) = 65.5 μsec
It becomes.

Further, when the exposure data is extremely sparse, for example, when 10 shots are taken at 4 μm square, the required time is the same as above,
3 μsec + 50 μsec = 53 μsec
It becomes.

Here, the total time required for exposure when the exposure data of 1 mm square per cell is dense / dense / dense repeated data, that is, the average required time is:
1 mm □ ÷ 100 μm □ = 100 fields. For example, assuming that 50 fields have sparse exposure data and 50 fields have dense exposure data.
50 fields × (53 μsec + 65.5 μsec) = 5.925 msec
It becomes.

On the other hand, when the exposure is controlled only by the settling wait time (50 μsec) of the main deflector without using the main de-settling feedback circuit and the feedback deflector, under the same conditions,
If the exposure data is dense,
50 μsec + (100 μm □ ÷ 4 μm □ × 100 nsec) = 112.5 μsec
If the exposure data is sparse,
50 μsec + (10 shots × 100 nsec) = 51 μsec
Becomes
50 fields × (51 μsec + 112.5 μsec) = 8.175 msec
It becomes.

Therefore, for example, when the exposure field is 100 fields and the exposure data is only sparse,
2 μsec (= 53 μsec−51 μsec) × 100 (field) = 0.2 msec Although the conventional method has better throughput, for example, when there is a density of 50 fields each, the average time required by the exposure method of the present invention is Is 8.175 msec-5.925 msec = 2.25 msec
And the throughput is improved.

Next, a second embodiment will be described with reference to FIG.
FIG. 3 is an explanatory diagram of the counting of the settling waiting time, the counting of the minimum stay time, and the timing of the exposure in the second embodiment of the present invention. In the second embodiment, the setting of the main deflector is controlled. And a counter for controlling the minimum stay time, start counting the settling wait time and minimum stay time of the main deflector at the same time as the deflection of the main deflector starts, and start the exposure by the sub deflector after the settling wait time ends. I do.
Then, after the end of the minimum stay time and the end of the exposure, the deflection operation by the next main deflector is repeated to expose the entire area.

Here, under the same conditions as in Example 1 described above, the time required for exposure is
If the exposure data is dense,
3 μsec + (100 μm □ ÷ 4 μm □ × 100 nsec) = 65.5 μsec
If the exposure data is extremely sparse,
50 μsec
It becomes.

Therefore, the average travel time is
50 fields × (50 μsec + 65.5 μsec) = 5.775 msec
And if the exposure data is only sparse,
1 μsec (= 51 μsec−50 μsec) × 100 (field) = 0.1 msec.
8.175 msec-5.775 msec = 2.4 msec
However, the average required time is shortened and the throughput is improved by the exposure method of the present invention.

Next, a third embodiment will be described with reference to FIG.
FIG. 4 is an explanatory diagram of the counting of the settling waiting time, the counting of the minimum stay time, and the timing of the exposure in the third embodiment of the present invention. In the third embodiment, the setting of the settling waiting time of the main deflector is controlled. And a counter for controlling the minimum stay time, start counting the settling wait time and minimum stay time of the main deflector at the same time as the deflection of the main deflector starts, and start exposure by the sub deflector after the settling wait time ends. I do.

Then, even if the minimum stay time has not ended, the deflection by the next main deflector is started after the end of the exposure.
Here, the remaining time of the previous minimum stay time and the initial setting value (3 μsec) of the settling wait time are compared by a comparator, and the larger value is set as the settling wait time of the next main deflector.

Here, under the same conditions as in the above-described first and second embodiments, the required time for exposure is 65.5 μsec as in the first and second embodiments when the exposure data is dense, but is extremely high. If the sparse exposure data continues, the settling waiting time of the next main deflector is included in the previous minimum stay time, so it is 50 μsec or less, for example, 47 μsec when the initial setting value is 3 μsec.

Therefore, the average required time is improved by 0.4 msec or more even when the exposure data is only sparse, and the average required time is improved by 2.4 msec or more even when the exposure data is sparse.

Next, a fourth embodiment in which the minimum stay time is controlled using an analog comparator will be described with reference to FIG.
FIG. 5 shows a configuration in which analog comparators 17 and 18 are provided at the output section of the difference extracting section of the main differential settling feedback circuit. The counting of the settling time of the main deflector is started, and after the settling time, the sub-deflector is set. Exposure by the deflector is started, and two different thresholds A and B (A> B) are input to the two analog comparators 17 and 18 as reference values, and are input from the difference extracting unit of the main differential settling feedback circuit. The outputs of the two analog comparators 17 and 18 compared with the output of the AND circuit 19 are input to an AND circuit 19, and the output via the AND circuit 19 is used as a minimum stay time flag.

In this case, when the output of the difference extraction unit of the main differential settling feedback circuit is equal to or smaller than A and equal to or larger than B, a minimum stay time flag is output to enable deflection of the next main deflector. Only the deflection settling time by the main deflector is managed, and the minimum stay time is controlled by an analog waveform, so that the time management is simplified.

Next, a fifth embodiment in which the settling wait time and the minimum stay time are controlled using an analog comparator (not shown) will be described.
In the fifth embodiment, an analog comparator for settling wait time control and an analog comparator for minimum stay time control are provided at the output of the difference extraction unit of the main differential settling feedback circuit similar to FIG. When the output of the difference extraction unit of the main de-settling feedback circuit is input from the start of deflection by the main deflector to the settling waiting time control analog comparator using the value that can be exposed as the reference value, and this input reaches the reference value Exposure starts.

Next, after the end of the exposure, and using two analog comparators as in FIG. 5, if the output of the difference extraction unit of the main differential settling feedback circuit becomes A or less and B or more, the minimum stay time flag is set. Is output to enable deflection by the next main deflector.
Accordingly, since the settling wait time and the minimum stay time are controlled only by the analog waveform without performing the time management, the entire system is simplified.

Next, a sixth embodiment of the present invention relating to the continuous movement exposure method will be described with reference to FIGS.
First, the upper few bits of the absolute value of the difference between the value before deflection and the value after deflection of the main deflector in binary were extracted from the X side and the Y side, respectively, and these were previously obtained as addresses. The data of the ideal value of the settling time is called from the memory storing the ideal value of the settling time.

Referring to FIG. 6, for example, as shown in the upper part of FIG. 6, assuming that the deflection range of the main deflector is about ± 1 mm, if the upper 8 bits (2 ⁸ = 256) are taken out, about 8 μm (strictly, 8. A jump amount map by deflection can be formed by a mesh of 192 μm) LSB (least significant bit), and arbitrary data can be stored in each address of the jump amount map meshed by 8 μmL SB.

When the jump amount ΔG is 100 μm or less, the minimum settling wait time of 3 μsec is sufficient. When the jump amount ΔG is 100 μm or more, the settling wait time longer than 3 μsec is required. Settling wait time WAIT1,
WAIT1 = 3 μsec: ΔG ≦ 100 μm
WAIT1 = (ΔG−100) × A + 3 μsec: ΔG> 100 μm
Set as follows.

In this case, the settling wait time obtained by the jump amount map is preloaded into the time counter (down counter), and the end of the settling wait time is indicated by the countdown value becoming 0.
However, if the clock supplied to the time counter is 10 MHz (10 ⁷ Hz), the LSB of the settling wait time of the map is 10 ⁻⁷ sec, that is, 100 nsec. For example, if the settling wait time is 3 μsec, 100 nsec Therefore, 30 is stored in an address where the flying distance ΔG is 100 μm or less.

As shown in the example of the map storing such data in the lower part of FIG. 6, the jump amount ΔG is 30 near the center of the XY coordinate, that is, near the (0,0) point, that is, 3 μsec. Is input, and the larger the distance from the (0,0) point, the larger the numerical value.

Next, a reading timing determining memory for determining the timing of reading the movement of the sample is provided, and the value of the down counter is used as an address of the reading timing determining memory, and the movement of the sample is moved to an address a fixed time before the end of the entire settling wait time. A bit for outputting a flag to start reading is stored.

When the output of the down counter reaches a predetermined time before the end of the whole settling wait time, a flag for starting reading of the movement of the sample is output, and the last reading immediately before the end of the whole settling wait time WAIT is performed. Therefore, the end of the correction settling time accompanying the correction deflection due to the last reading can be made before the end of the entire settling wait time, and the time required for the end of the entire large settling wait time WAIT is reduced. The read flag can be generated at a good timing, and the minimum waiting time C is not wasted.

Also, by providing the read timing determination memory with a time control function, the read flag generation interval can be arbitrarily changed. Therefore, the read interval is increased near the start of the large settling wait time, and the settling wait time is increased. In the vicinity of the end point of the above, the interval can be set to be narrow, the waste of the settling wait time WAIT can be eliminated, and the number of times of reading can be reduced.

Further, the data stored in the settling time map may be set differently in the X-axis direction and the Y-axis direction. If the directions differ, the difference in the characteristics can be freely corrected.

FIG. 8 is an explanatory diagram of a settling waiting time control unit of the electron beam exposure apparatus common to the first to third embodiments of the present invention. FIG. 4 is an explanatory diagram of a count of a settling waiting time, a count of a minimum stay time, and an exposure timing in the first embodiment of the present invention. FIG. 9 is an explanatory diagram of a count of a settling waiting time, a count of a minimum stay time, and an exposure timing in a second embodiment of the present invention. FIG. 10 is an explanatory diagram of a count of a settling waiting time, a count of a minimum stay time, and an exposure timing in a third embodiment of the present invention. FIG. 14 is an explanatory diagram of a control mechanism when controlling a minimum stay time using an analog comparator according to a fourth embodiment of the present invention. FIG. 14 is an explanatory diagram of a jump amount map according to the sixth embodiment of the present invention. FIG. 14 is an explanatory diagram of a sample reading timing in Embodiment 6 of the present invention. FIG. 3 is an explanatory diagram of a control mechanism of an electron beam exposure apparatus in a conventional blanking aperture array exposure system. FIG. 4 is an explanatory diagram of a settling waiting time control unit of an electron beam exposure apparatus in a conventional blanking aperture array exposure system. FIG. 9 is an explanatory diagram of a sample reading timing in a conventional continuous movement exposure method. FIG. 7 is an explanatory diagram of a locus of an electron beam in a conventional blanking aperture array exposure method. FIG. 7 is an explanatory diagram of a problem in a conventional blanking aperture array exposure method.

Explanation of reference numerals

REFERENCE SIGNS LIST 11 comparison amplifier 12 input resistance 13 feedback resistance 14 output resistance 15 analog comparator 16 gain / rotation calculator 17 analog comparator 18 analog comparator 19 and circuit

Claims (11)

The difference between the output of the DA converter that converts the amount of deflection of the main deflector from a digital amount to an analog amount and the voltage applied to the main deflector is extracted, and the position of the electron beam is determined based on the extracted output. In an electron beam exposure apparatus having a main de-settling feedback circuit that deflects to an exposure-possible position of a sub-deflector, when deflecting by the main deflector, separately from a settling wait time of the main deflector, A main deflector setting for setting a minimum stay time including a part thereof and performing control so that deflection by the next main deflector becomes possible after the end of the minimum stay time and after the end of exposure of the sub deflector. An electron beam exposure apparatus comprising a waiting time control unit. The main deflector settling time control unit is set so as to immediately start counting of the minimum stay time and exposure by the sub deflector immediately after the end of the settling time of the main deflector. 2. The electron beam exposure apparatus according to claim 1. The counting of the settling time of the main deflector and the counting of the minimum stay time are started simultaneously, and the main deflector is set so that the exposure by the sub deflector is started after the settling time of the main deflector is completed. 2. The electron beam exposure apparatus according to claim 1, wherein a waiting time control unit is set. The counting of the settling wait time of the main deflector and the counting of the minimum stay time are started simultaneously, and the exposure by the sub deflector is started after the settling wait time of the main deflector is completed. Starting the deflection by the main deflector and starting the next exposure by the sub deflector after the settling wait time in the next deflection and after the end of the previous minimum stay time. 2. The electron beam exposure apparatus according to claim 1, wherein a unit is set. The next settling waiting time of the main deflector is compared with the remaining time of the minimum staying time of the deflection by the previous main deflector and the initial settling waiting time, and is set to a larger time between the two. Item 5. An electron beam exposure apparatus according to Item 4. The difference between the output of the DA converter that converts the amount of deflection of the main deflector from a digital amount to an analog amount and the voltage applied to the main deflector is extracted, and the position of the electron beam is determined based on the extracted output. A main de-settling feedback circuit for deflecting the sub-deflector to the exposure-possible position; A main deflector settling waiting time control unit for setting a stay time, after the minimum stay time, and after the exposure of the sub deflector, to control the deflection by the next main deflector. In the operating method of the electron beam exposure apparatus for exposing while continuously moving the sample, the end of the correction settling time accompanying the correction deflection of the main deflector by reading the movement amount of the sample is provided to the main deflector. Yo So that at the end the previous overall settling time of the deflection, the operation method of the electron beam exposure apparatus characterized by determining the reading timing of the sample based on the remaining amount of the overall settling time. The whole settling wait time is calculated as a digital number from the amount of flight of the main deflector, the digital number is counted by a down counter, the settle wait time is waited, and the numerical value of the down counter is read and the address of the memory for timing determination is determined. The data for starting reading is stored at an address a fixed time before the end of the entire settling wait time, and the end of the correction settling time accompanying the read correction deflection immediately before the end of the entire settling wait time is the entire time. 7. The method of operating an electron beam exposure apparatus according to claim 6, wherein the correction waiting time is before the end of the correction waiting time. 8. The electron beam exposure apparatus according to claim 7, wherein the read timing determination memory has a time function of changing a read cycle immediately before the end of the entire settling wait time and another read cycle. Operation method. A memory having, as an address, an absolute value of the amount of deflection of the deflection by the main deflector with respect to the X-axis and the Y-axis is provided, and a stabilization waiting time map depending on the amount of flight and the direction of flight is stored in the memory. The method for operating an electron beam exposure apparatus according to claim 7, wherein: The settling wait time map stores a fixed settling time when the absolute value of the amount of deflection by the main deflector is small, and stores a settling time that is proportional to the absolute value of the amount of flight otherwise. The method for operating an electron beam exposure apparatus according to claim 9, wherein: 10. The operating method of the electron beam exposure apparatus according to claim 9, wherein the settling time depending on the direction of the deflection by the main deflector is set to be different between the X-axis direction and the Y-axis direction.

Images