JP2907527B2 - Charged particle beam exposure apparatus and exposure method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Summary of the Invention] A charged particle beam exposure apparatus and an exposure method for performing a block exposure are described, which enable accurate clock correction for each pattern of an actual block mask and enable calibration. In a charged particle beam exposure apparatus having a block mask having an aperture group for shaping a charged particle beam into one of a plurality of basic patterns or a variable rectangular pattern, an actual measurement is performed with respect to a basic irradiation time for each pattern on the block mask. Having a block mask clock memory for storing the clock correction amount for each pattern obtained by
Further, in an exposure method using a charged particle beam exposure apparatus including a block mask having an aperture group for shaping a charged particle beam into one of a plurality of basic patterns or a variable rectangular pattern, the reference irradiation time for each pattern on the block mask is The reference irradiation time is corrected with the clock correction amount for each pattern on the block mask obtained by actual measurement, and the exposure time for the pattern is determined.

[Industrial applications]

The present invention relates to a charged particle beam exposure apparatus for performing block exposure and an exposure method.

In recent years, with the increase in the density of integrated circuits, a new exposure technique using an electron beam has been studied instead of photolithography, which has been the mainstream of a fine pattern forming method for many years, and it has come into practical use.

[Conventional technology]

A conventional electron beam lithography apparatus having a practical range is an apparatus that generates an arbitrary shot using a variable rectangular beam and draws a pattern on an irradiation target. Such an apparatus has a pattern generator function of decomposing pattern data, which is the minimum unit on the design side, into shots with an arbitrary variable rectangular beam. However, as the integration advances, the pattern size decreases and the number of patterns increases, so that the number of beam shots resolved by the pattern generator increases and the throughput decreases. Therefore, a block exposure method has been proposed in order to obtain a realistic throughput even when an ultrafine pattern is generated.

Semiconductor devices that require ultra-fine patterns include, for example,
Like a 64 mega DRAM, most areas that are fine or drawn are repetitions of any basic pattern. If the repetition of the basic pattern can be processed in one shot, the throughput is improved. Therefore, a method of generating the basic pattern in one shot by holding the above basic pattern on a transmission mask and irradiating it with an electron beam is block exposure.

FIG. 6 shows an example of the block exposure method. The main part is a block mask 15 in which a plurality of basic patterns (15b and the like) are formed, and a rectangular pattern 15a for a variable rectangular beam is also formed in the center. Electrons emitted from the electron gun 11 pass through the shaping aperture 12, and select and irradiate an arbitrary pattern portion of the block mask 15 by the mask deflector 14 placed at the image point of the crossover formed by the electromagnetic lens 13. . A single pattern having no repetitive pattern is a rectangular pattern in the block mask 15.
15a is selected, and it becomes the same as the conventional variable rectangular beam exposure method. The patterned image is returned to the optical axis by the convergence action of the lens, the pattern image is reduced by the reduction lens 16, and is radiated to an arbitrary portion of the irradiation target 20 by the projection lens 17, the deflection coil, and the like.

Various element technologies are required to irradiate the pattern image of the electron beam onto the object and obtain a high-precision pattern.However, whether or not the desired pattern can be obtained depends on the electron beam on the object. It is determined by the setting of the pattern irradiation time determined by the sensitivity of the photosensitive resist and the current density of the electron beam pattern image. In the variable rectangular beam exposure method, this is set as a basic irradiation time, and clock correction is performed on each pattern data in order to take into account the proximity effect and the like resulting from pattern arrangement and high density. In the clock correction, the exposure amount (irradiation time) is corrected, and the shot time of each pattern is determined by the ratio to the basic irradiation time, thereby enabling all patterns to be exposed with high precision.

[Problems to be solved by the invention]

In the block exposure method, since the repetitive pattern is exposed in one shot, the absolute number of clock corrections for each pattern used in the variable beam exposure method decreases, but
It is necessary. However, clock correction in conventional block exposure is based on design data. That is, the clock correction amount for each pattern on the block mask is
The size is determined based on the size and shape of the pattern (both are design values). The same applies to the correction for the proximity effect, which does not take into account the manufacturing error of the block mask, the position of each pattern on the block mask, the characteristics of each exposure apparatus, and the like.

SUMMARY OF THE INVENTION It is an object of the present invention to improve such a point, perform accurate clock correction for each pattern of an actual block mask, and enable calibration.

[Means for solving the problem]

As shown in FIG. 1, in the present invention, a block mask clock memory 28 is provided in the control unit 25 of the charged particle beam exposure apparatus 10 for performing block exposure. As shown in FIG. 1 (c), this memory 28 stores the clock correction amount for each mask NO (each pattern on the block mask 15).
The clock correction amount is a correction amount for the basic irradiation time of the pattern determined by the beam current density of the pattern image and the sensitivity of the beam-sensitive resist on the irradiation target, and is obtained by actual measurement.

In the exposure, the clock correction amount for the pattern is obtained by reading the block mask clock memory 28, and the basic irradiation time of the pattern generating circuit unit 27 is corrected using the clock correction amount, and the result is used as the exposure time.

[Action]

According to this exposure apparatus and method, each pattern can be exposed accurately and with an appropriate exposure time. That is, the processor 21 reads data from the magnetic drum 22 or the magnetic tape 23 and
The pattern information developed in FIG. 26 is as shown in FIG. 1 (b), and the shape codes 1, 2, 0,..., The pattern starting point, the pattern size, and the clock correction amount are as follows. Exposure is performed on the basis of the clock correction amount, which is determined based on the design data, and does not correspond to an actual exposure apparatus or block mask. On the other hand, according to the present invention, a clock correction amount is calculated by performing a sample current measurement based on an actual exposure apparatus and a block mask, and this is used as a block mask clock memory.
This is stored in the memory 28, read out at the time of exposure, and the basic irradiation time is corrected to obtain the exposure time, so that correct exposure can be performed.

〔Example〕

Referring to FIG. 1 in detail, exposure data is stored on a magnetic disk 22 or a magnetic tape 23, and is transferred by a processor 21 to a data memory 26 via an interface 24. As shown in FIG. 2B, the data memory 26 stores the shape codes 1 and 2 (these are variable beam patterns), their pattern start points (coordinates X1 and Y1), and the pattern size (the length X2 of the two sides of the rectangle). , Y2), clock correction (CLK; numerical value such as 1.112), and matrix information (repeated shape 1
Pitch PX, PY, number NX, NY, etc.) are stored. For the shape code 0 (this is a block pattern), the pattern start point (X1, Y1), mask NO, clock correction, and matrix information are stored.

The clock correction data stored in the data memory 26 is at the design stage. The clock correction data of the shape code 0, that is, the block pattern, has its general value CLK in the memory 26, and its correction value for each pattern in the block mask clock memory 28. The contents of the memory 28 are as shown in FIG. That is, the memory 28 stores the clock correction amounts CLK1, CLK2,... Of the masks (block patterns) 15b,... For each mask NO, and the mask NO is ffff, for the aperture 15a for the variable rectangular beam. The clock correction amount is set to 0. When the memory 26 is read and the shape code 0, the mask NO ××××, and the clock correction amount CLK are obtained, the memory 28 is read and the clock correction amount of the mask NO is obtained.
CLK _X is obtained, and this is added to or multiplied by the clock correction amount of the memory 26 and used.

The clock correction data and the pattern information obtained in this way are sent to the block exposure apparatus 10 via the pattern generation circuit 27.
To form a pattern. The basic irradiation time is set in advance by the processor 21 in the pattern generation circuit block 27, and the actual exposure time is obtained by correcting the basic irradiation time with the clock correction data.

FIG. 2 shows the relationship between the data memory, the block mask clock memory, and the pattern generation circuit. Data memory 26
To the pattern data X1, etc. Y1 pattern generating circuit 27 read from a clock correction CLK ₁ is added / multiplier 31
To also mask NO is sent to the memory 28, the clock correction CLK ₂ read out from the memory 28 in this mask NO is sent to the adder / multiplier 31. The pattern generation circuit 27 sends pattern information to a D / A converter such as a deflector according to the pattern data from the memory 26, and the basic irradiation time CLK
₀ is sent to the second adder / multiplier 32. Addition / multiplication results CLK ₃ of the adder / multiplier 31 is also fed to the adder / multiplier 32, adder / multiplier 32 outputs the addition / multiplication results CLK of CLK ₀ and CLK ₃ to the exposure apparatus.

The data stored in the block mask clock memory 28 is obtained by actual measurement. Specifically, this is performed by measuring a sample current when each pattern of the block mask is irradiated with a beam. As shown in FIG. 3, the block mask 15
The pattern 15a provided on the optical axis portion is a rectangular pattern used for variable rectangular beam exposure. The current density obtained by using the rectangular pattern 15a provided on the optical axis (the sample current ÷
The area of the rectangular pattern 15a) is used for setting the basic irradiation time.

The correction data for the basic patterns 15b, 15c,... Is obtained as follows. That is, a measurement mask 15A having a rectangular pattern 15a provided on the optical axis of the operation mask is prepared on the entire surface of the operation mask (block mask) shown in FIG. The sample current when each pattern 0, 1, 2,... Is irradiated is measured, and the ratio of the optical axis to the current density of the master pattern 15a is calculated. As a result, data as shown in FIG. 4 (b) is obtained. The correction rate for pattern No. 0 is 1.000
, But this depends on the criteria. The correction factor for the patterns 1, 2, 3,... Of the measurement mask is larger than 1.000, which is because the beam current density is lower at the periphery of the mask than at the center of the mask. Fig. 5 (a)
Shown in As shown in the figure, a pattern is selected and a sample current is measured to obtain a current density correction value for pattern 0.
This is performed for all patterns on the mask (N is the number of patterns). Thus, correction data for the position of the pattern on the mask is obtained. Although there may be a difference between the measurement mask and the operation mask, a change in the state of the exposure apparatus accompanying replacement when the mask cannot coexist, and the like, these are ignored as small.

Next, as shown in FIG. 5B, the sample current of each pattern of the operation mask is measured. Optical axis rectangular pattern 15a
Is calculated, and each sample current is estimated based on the aperture (beam irradiation area) of each pattern determined at the time of mask production. The current density used in the calculation of the estimated sample current is corrected by a current density map based on the pattern selection obtained by the measurement mask. Then, final correction is performed by comparing the measured sample current value of each pattern with the estimated sample current value. In this final correction, corrections based on mask manufacturing accuracy and pattern aperture are mixed, but can be determined by inspection processing. For example, when sample currents of patterns having the same or close aperture are compared, the mask manufacturing accuracy is related, and thus the mask itself can be evaluated. The final clock correction amount is stored in the block mask clock memory.

This carburetion is performed at appropriate intervals to update the block mask clock memory. In this way, the clock correction amount can always be kept at the optimum value.

〔The invention's effect〕

As described above, according to the present invention, it is possible to eliminate the non-uniformity of the beam irradiation due to the selected pattern on the block mask, thereby performing high-precision exposure. In addition, it is possible to cope with the manufacturing accuracy of each pattern on the block mask, the characteristics of the exposure apparatus, and the like, and the former leads to an improvement in the throughput of manufacturing the block mask.

[Brief description of the drawings]

FIG. 1 is a principle diagram of the present invention, FIG. 2 is a block diagram showing a configuration of a control unit, FIG. 3 is an explanatory diagram of an operation mask, FIG. 4 is an explanatory diagram of a measurement mask and correction data, and FIG. FIG. 6 is a flowchart showing a correction procedure, and FIG. 6 is an explanatory view of an exposure apparatus having a block mask. In FIG. 1, reference numeral 28 denotes a block mask clock memory, and FIG.
14A is a measurement mask, and 15 in FIG. 6 is a block mask.

Claims (3)

(57) [Claims] 1. A charged particle beam exposure apparatus comprising a block mask (15) having an aperture group for shaping a charged particle beam into one of a plurality of basic patterns or a variable rectangular pattern, wherein a basic irradiation for each pattern on the block mask is performed. A charged particle beam exposure apparatus comprising a block mask clock memory (28) for storing a clock correction amount for each pattern obtained by actually measuring time with respect to time. 2. A measurement mask (15A) in which the clock correction amount stored in the block mask clock memory has the same opening arrangement as the block mask, and each opening is the same as the rectangular opening for the variable rectangular pattern at the center of the block mask. Is used to measure the sample current, and from this result the current density correction of each opening with respect to the rectangular opening is obtained.Then, the sample current is measured for each pattern using a block mask, and each current is corrected using the current density correction value. 2. The charged particle beam exposure apparatus according to claim 1, wherein an estimated sample current of the pattern is calculated, and the estimated sample current is calculated from the sample current value of each pattern measured. 3. An exposure method using a charged particle beam exposure apparatus having a block mask having an aperture group for shaping a charged particle beam into one of a plurality of basic patterns or a variable rectangular pattern, wherein each pattern on the block mask is A reference irradiation time is given, and the exposure time for the pattern is determined by correcting the reference irradiation time with a clock correction amount for each pattern on the block mask obtained by actual measurement. Method.