Granted Patent B2

US 12,730,382 · App. 18/806,848 · Granted Sep 8, 2026

Inventors: Yeongchan Cho (Suwon-si, KR); Minjae Kim (Suwon-si, KR); Sungmin Nam (Suwon-si, KR); Seunghune Yang (Suwon-si, KR)

Assignee: Samsung Electronics Co., Ltd.

G03F7/70516G03F7/70504G03F7/70033

Abstract

Claims (202)

1 . A method of configuring an extreme ultraviolet (EUV) illumination system, the method comprising:

generating a library of independent aerial images, each of the aerial images being generated by performing an optical simulation with respect to each of a plurality of EUV point sources, each EUV point source corresponding to a respective optical path defined by an assignment between a field facet mirror (FFM) and a pupil facet mirror (PFM);

performing an EUV mapping that defines combinations of the EUV point sources, wherein the EUV mapping corresponds to a predetermined condition of selecting optical paths from M FFMs to N PFMs, where M represents a positive integer and N represents an integer greater than M;

for each of the combinations defined by the EUV mapping, summing up the aerial images corresponding to EUV point sources included in the combination, without performing an additional optical simulation for the combination;

searching, from among the combinations defined by the EUV mapping, for a combination of the EUV point sources using a fitness value with respect to the summed aerial images, wherein the fitness value is calculated based on quantifying at least one of (i) optical indexes or (ii) performance indexes of patterning in an EUV light exposure process; and

physically configuring the EUV illumination system by directing light from a plurality of FFMs along optical paths to a plurality of PFMs according to the combination of the EUV point sources having a maximum fitness value.

2 . The method of claim 1 , wherein

each of the plurality of EUV point sources is a minimum unit that is configured to be individually turned on/off, and

EUV light from the plurality of EUV point sources at different locations has incoherence with each other.

3 . The method of claim 1 , wherein the aerial images are represented as graphs of an intensity of a cutline with respect to a mask pattern.

4 . The method of claim 1 , wherein the aerial images, with respect to EUV point sources, are generated through the optical simulation.

5 . The method of claim 1 , wherein, in the searching of the combination of the EUV point sources, a genetic algorithm is used.

6 . The method of claim 5 , wherein,

in the summing up of the aerial images, P combinations of the EUV point sources are generated, with P representing a positive integer, and the aerial images are summed with respect to each of the combinations of the EUV point sources;

wherein the searching of the combination of the EUV point sources comprises

determining the fitness value with respect to each of the P combinations of the EUV point sources,

selecting Q combinations of the EUV point sources, which are set in an order of large fitness values, with Q representing a positive integer less than P, and

modifying a subset of the EUV mappings in each of the Q combinations of the EUV point sources while maintaining remaining EUV mappings,

wherein, after the modifying the subset of the EUV mappings, the summing up of the aerial images is performed;

wherein the determining of the fitness value, the selecting of the Q combinations, and the modifying of the subset of the EUV mappings are repeatedly performed a predetermined number of times; and

wherein, in the configuring of the EUV illumination system, after the performance of the determining of the fitness value, the selecting of the Q combinations, and the modifying the subset of the EUV mappings the predetermined number of times, the EUV illumination system is configured using the EUV point sources included in the combination having the maximum fitness value.

7 . The method of claim 6 , wherein the fitness value is a weighted total of performance indexes comprising at least one of a critical dimension (CD) longer axis, a CD shorter axis, a CD aspect ratio that is a ratio between the CD longer axis and the CD shorter axis, a longer axis normalized image log slope (NILS), a shorter axis NILS, or dose.

8 . The method of claim 1 , wherein, in the searching of the combination of the EUV point sources, a Lagrange multiplier method is used.

9 . The method of claim 8 , wherein, in the Lagrange multiplier method,

a function L, representing the predetermined condition, is defined by Equation 1 below in order to find an optimal combination of the EUV point sources based on normalized image log slopes (NILS) including NILSx and NILSy,

L ( I i ,λ)=NILS−λ( I tot −I 0 )  (1),

the NILSx of the NILS is expressed as

∑

i

=

1

M

CD

x

I

tot

⁢

dI

i

dx

,

the NILSy of the NILS is expressed as

∑

i

=

1

M

CD

y

I

tot

⁢

dI

i

dy

,

and

I tot is

∑

i

=

1

M

I

i

and represents a total intensity of the aerial images in a target critical dimension (CD) and I 0 denotes a value used in implementing the target CD,

under a condition of I tot =I 0 , a problem of finding the optimal combination of the EUV point sources is changed into a problem of maximizing the NILS, and

the optimal combination determined by maximizing the NILS under the condition of I tot =I 0 is used to physically configure the EUV illumination system.

10 . The method of claim 9 , wherein

the NILS corresponds to a slope of an intensity at a point where the target CD is defined, and

the NILS has a characteristic in which the CD is less changed with respect to a change in processes as a value of the NILS increases.

11 . The method of claim 1 , wherein the method is applied to a mask pattern to which a bias is applied or to a plurality of mask patterns.

12 . A method of configuring an extreme ultraviolet (EUV) illumination system, the method comprising:

generating aerial images, indicated as an intensity of a cutline with respect to a mask pattern, each of the aerial images being generated by performing an optical simulation on each of a plurality of EUV point sources that are individually configured to turn on/off incoherent light;

performing an EUV mapping for selecting optical paths from M field facet mirrors (FFMs) to N pupil facet mirrors (PFMs) in order to configure an illumination system of the EUV point sources, wherein the EUV mapping corresponds to a predetermined condition of selecting the optical paths from M FFMs to N PFMs, where M represents a positive integer and N represents an integer greater than M;

summing up the aerial images based on the EUV mapping;

searching for a combination of the EUV point sources using a fitness value with respect to the summed aerial images, wherein the fitness value is calculated based on quantifying at least one of (i) optical indexes or (ii) performance indexes of patterning in an EUV light exposure process; and

physically configuring the EUV illumination system based on the combination of the EUV point sources having a maximum fitness value.

13 . The method of claim 12 , wherein the aerial images with respect to N EUV point sources are generated through the optical simulation.

14 . The method of claim 12 , wherein,

in the summing up of the aerial images, P combinations of the EUV point sources are generated, with P representing a positive integer, and the aerial images are summed with respect to each of the combinations of the EUV point sources;

wherein in the searching of the combination of the EUV point sources, a genetic algorithm is used;

wherein the searching of the combination of the EUV point sources comprises

determining the fitness value with respect to each of the P combinations of the EUV point sources,

selecting Q combinations of the EUV point sources, which are set in an order of large fitness values, with Q representing a positive integer less than P, and

modifying a subset of the EUV mappings in each of the Q combinations of the EUV point sources,

wherein, after the modifying the subset of the EUV mappings, the summing of the aerial images is performed,

wherein the determining of the fitness value, the selecting of the Q combinations, and the modifying of the subset of the EUV mappings are repeatedly performed a predetermined number of times, and

wherein, in the configuring of the EUV illumination system, after the performance of the determining the fitness value, the selecting the Q combinations, and the modifying the subset of the EUV mappings the predetermined number of times, the EUV illumination system is configured using the EUV point sources included in the combination having the maximum fitness value.

15 . The method of claim 12 , wherein

in the searching of the combination of the EUV point sources, a Lagrange multiplier method is used, and

wherein in the Lagrange multiplier method a function L, representing a predetermined condition, is defined by Equation 1 below in order to find an optimal combination of the EUV point sources based on normalized image log slopes (NILS) including NILSx and NILSy,

L ( I i ,λ)=NILS−λ( I tot −I 0 )  (1),

the NILSx of the NILS is expressed as

∑

i

=

1

M

CD

x

I

tot

⁢

dI

i

dx

,

the NILSy of the NILS is expressed as

∑

i

=

1

M

CD

y

I

tot

⁢

dI

i

dy

,

and

I tot is

∑

i

=

1

M

I

i

and represents a total intensity of the aerial images in a target critical dimension (CD) and I 0 denotes a value used in implementing the target CD,

under a condition of I tot =I 0 , a problem of finding the optimal combination of the EUV point sources is changed into a problem of maximizing the NILS, and

the optimal combination determined by maximizing the NILS under the condition of I tot =I 0 is used to physically configure the EUV illumination system.

16 . An extreme ultraviolet (EUV) exposure method comprising:

preparing an EUV mask;

configuring an EUV illumination system corresponding to the EUV mask; and

performing an EUV exposure on a wafer using the EUV illumination system,

wherein the configuring of the EUV illumination system comprises

generating aerial images, each of the aerial images being generated by performing an optical simulation on each of a plurality of EUV point sources that are individually configured to turn on/off incoherent light,

summing up the aerial images based on an EUV mapping, wherein the EUV mapping corresponds to a predetermined condition of selecting optical paths from M field facet mirrors (FFMs) to N pupil facet mirrors (PFMs), where M represents a positive integer and N represents an integer greater than M,

searching for a combination of the EUV point sources using a fitness value with respect to the summed aerial images, wherein the fitness value is calculated based on quantifying at least one of (i) optical indexes or (ii) performance indexes of patterning in an EUV light exposure process, and

physically configuring the EUV illumination system based on a combination of the EUV point sources having a maximum fitness value.

17 . The EUV exposure method of claim 16 , wherein

the aerial images, with respect to N EUV point sources, are generated through the optical simulation.

18 . The EUV exposure method of claim 17 , wherein,

in the summing up of the aerial images, P combinations of the EUV point sources are generated, with P representing a positive integer, and the aerial images are summed with respect to each of the combinations of the EUV point sources;

wherein the searching of the combination of the EUV point sources, a genetic algorithm is used;

wherein the searching of the combination of the EUV point sources comprises

determining the fitness value with respect to each of the P combinations of the EUV point sources,

selecting Q combinations of the EUV point sources, which are set in an order of large fitness values, with Q representing a positive integer less than P, and

modifying a subset of the EUV mappings in each of the Q combinations of the EUV point sources,

wherein, after the modifying the subset of the EUV mappings, the summing up of the aerial images is performed,

wherein the determining of the fitness value, the selecting of the Q combinations, and the modifying of the subset of the EUV mappings are repeatedly performed a predetermined number of times, and

wherein, in the configuring of the EUV illumination system, after the determining the fitness value, the selecting the Q combinations, and the modifying the subset of the EUV mappings the predetermined number of times, the EUV illumination system is configured using the EUV point sources included in the combination having the maximum fitness value.

19 . The EUV exposure method of claim 17 , wherein,

in the searching of the combination of the EUV point sources, a Lagrange multiplier method is used, and

wherein in the Lagrange multiplier method a function L, representing the predetermined condition, is defined by Equation 1 below in order to find an optimal combination of the EUV point sources based on normalized image log slopes (NILS) including NILSx and NILSy,

L ( I i ,λ)=NILS−λ( I tot −I 0 )  (1),

the NILSx of the NILS is expressed as

∑

i

=

1

M

CD

x

I

tot

⁢

dI

i

dx

,

and

the NILSy of the NILS is expressed as

∑

i

=

1

N

CD

y

I

tot

⁢

dI

i

dy

,

I tot is

∑

i

=

1

M

I

i

and represents a total intensity of the aerial images in a target critical dimension (CD) and I 0 denotes a value used in implementing the target CD, and

under a condition of I tot =I 0 , a problem of finding the optimal combination of the EUV point sources is changed into a problem of maximizing the NILS, and

the optimal combination determined by maximizing the NILS under the condition of I tot =I 0 is used to physically configure the EUV illumination system.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (19)