Takahiro Kozawa1, Julius Joseph Santillan2, and Toshiro Itani2

1The Institute of Scientific and Industrial Research, Osaka University 2EUVL Infrastructure Development Center (EIDEC)

Stochastic Effects in Chemically Amplified Resists

for Extreme Ultraviolet Lithography

Evaluation of resist materials

The performance of resist materials is generally

evaluated by comparing the resolution, LER,

and sensitivity.

Sensitivity

Resolution

LWR(LER)

Trade-off

Trade-off relationship

Exposure experiments

Resolution, Sensitivity, LWR(LER)

SEM image

The evaluation of the potential capability of

resist materials is tricky because of the trade-off

relationships between these factors.

It is essential to the evaluation and development of resist materials

to extract information from SEM images as much as possible.

For example, high sensitivity does not

necessarily mean high performance.

Small Field Exposure Tool : SFET

Source (Xe DPP)

Exposure

chamber

Mask loader

Wafer loader

Wafer

track

Items Target

Specifications

NA 0.3

Illumination

mode

Annular(0.3/0.7),

x-slit

Field size 0.2 x 0.6 mm

Magnification 1/5

Wavefront error ＜0.9 nm rms

Flare ＜7% （MSFR）

Source power 0.5W @IF

Wafer size 300 mm

Objective

Establishment of scientific foundation and technology for resist evaluation

The stochastic effect in line-and-space patterns fabricated using SFET was analyzed to clarify the relationship of stochastic effect to LER and resist pattern defects.

-30

30

Dev

iati

on

fro

m h

alf-

pit

ch

(no

min

al l

ine

wid

th)

(nm

)

0

0

12

LE

R (

nm

)

6

(b) LER (a) Line width

22 23 24 25 26 28 30 32 35 40 45 50 60

Exposure dose (mJ cm-2)

Hal

f-p

itch

(nm

)

Dose-pitch matrices of EIDEC standard resist

Exp

erim

en

tal

21 22 23 24 25 26 28 30 32 35 40 45 50 60

Exposure dose (mJ cm-2)

Hal

f-p

itch

(nm

)

21

Half-pitch dependence (16 mJ cm-2 exposure dose) 60 50 45 40 35 32 30 28 26 25 24 23 22 21

Exposure dose dependence (60 nm HP) 10 11 12 13 14 15 16 17 18 19 20

nm

mJ cm-2

-30

30

Dev

iati

on

fro

m h

alf-

pit

ch

(no

min

al l

ine

wid

th)

(nm

)

0

0

12

LE

R (

nm

)

6

(b) LER (a) Line width

22 23 24 25 26 28 30 32 35 40 45 50 60

Exposure dose (mJ cm-2)

Hal

f-p

itch

(nm

)

Dose-pitch matrices of EIDEC standard resist

Exp

erim

en

tal

21 22 23 24 25 26 28 30 32 35 40 45 50 60

Exposure dose (mJ cm-2)

Hal

f-p

itch

(nm

)

21

Half-pitch dependence (16 mJ cm-2 exposure dose) 60 50 45 40 35 32 30 28 26 25 24 23 22 21

Exposure dose dependence (60 nm HP) 10 11 12 13 14 15 16 17 18 19 20

nm

mJ cm-2

Analysis procedure

Step 1: Overall fitting – Probability density model

Step 2: Refitting with Monte Calro acid generation simulation – Hybrid model

Acid generation Catalytic chain reaction

Step 3: Analysis with Monte Carlo process simulation – Stochastic model

Stochastic effect was investigated using Monte Carlo method.

However, Monte Carlo method is not suitable for accurate calculation.

EUV photon

Ionization

Ionization Next ionization or excitation

Resist

Thermalization

photon

electron

E < Eth

E > Eth

hn

hn-Ie

-Ie

+

-

+

+

- - +Acid generator

Protected unit

(polymer)

Rp

Rq

Quencher

Deprotection

Neutralization

Acid

During PEB During exposure

Diffusion

Point spread function

Monte Carlo method

Reaction diffusion equations

Monte Carlo method Monte Carlo method

Reaction diffusion equations

0.0

0.2

0.4

0.6

0.8

1.0

10 20 30 40 50 60 70

Best fit parameters

Effective quencher concentration: 0.023 nm-3

Effective reaction radius for deprotection: 0.16 nm

Diffusion constant: 10 nm2 s-1 (2~6 nm2 s-1) Proportionality constant between LER and chemical gradient fLER: 0.22

Dissolution point:

Half pitch (nm)

Dis

solu

tio

n p

oin

t

Dependence of dissolution characteristics on pattern size

Amount of chemical reaction

required for dissolution

Amount of chemical reaction

required at 16 nm half-pitch

increases by 74%, compared

with that at 60 nm half-pitch.

0

1

2

3

4

-23 -11.5 0 11.5 23

0

1

2

3

-23 -11.5 0 11.5 23

0.0

0.2

0.4

0.6

0.8

1.0

-23 -11.5 0 11.5 23

Distance (nm)

Absorbed photon image

No

rmal

ized

ab

sorb

ed

ph

oto

n d

ensi

ty

Standard deviation

Average

Acid image Latent image

12 s 24 s 36 s 48 s 60 s 0 s

Top d

ow

n

Cro

ss s

ecti

on

Acid conc. (nm-3)

0 0.6

Protected unit conc. (nm-3)

0 4

46 nm

50

nm

Stochastic effect – 23 nm LS pattern at 16 mJ cm-2 exposure Photon image

Photon density (nm-3)

Standard deviation

Average

No

rmal

ized

aci

d c

on

c.

Acid image

Distance (nm)

Latent image

Standard deviation

Average

No

rmal

ized

pro

tect

ed

un

it c

on

c.

Distance (nm)

0

5

10

15

0 5 10 15 20

Acid diffusion length in EIDEC standard resist

Initial position (nm)

Aci

d d

iffu

sio

n l

eng

th (

nm

)

10

15

0

PEB time: 60 s

Aci

d l

ifet

ime

(s)

Acid lifetime, t

ADL

5

Dependence of acid diffusion length and lifetime on the initial position of acids (23 nm HP, 16 mJ cm-2)

average

average

6

7

8

9

10

11

12

13

14

12 14 16 18 20

Exposure dose (mJ cm-2) A

cid

dif

fusi

on

len

gth

(n

m)

21 nm 22 nm 23 nm 24 nm

25 nm 26 nm 28 nm 30 nm

Half-pitch

The optimum acid diffusion length for 16 nm L&S patterns has been reported to be ~10 nm.

EIDEC standard resist has the potential for resolving 16 nm features with HQ EUV images.

T. Kozawa, JJAP 52 (2013) 016501.

Dependence of acid diffusion length on half-pitch and exposure dose

The polarity of resist polymer was sufficiently changed for the dissolution in developer with the acid diffusion length of 10 nm in the EIDEC standard resist.

0.0

0.2

0.4

0.6

0.8

1.0

-22 -11 0 11 22

Latent image of 22 nm LS pattern at 16 mJ cm-2 exposure

Distance (nm)

Norm

aliz

ed p

rote

cted

unit

conc.

Pattern boundary (Exp.)

Average

+0.31s

-0.31s

s

Dissolution point

Normalized standard deviation of protected unit connected to a polymer

LER (Exp.)

The fluctuation of protected unit concentration leads to the fluctuation

of the crossing point between latent image and dissolution threshold.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

20 22 24 26 28 30

a

Half-pitch (nm)

Ave. Monodisperse

Ave. Random

Relationship between protected unit fluctuation and LER

0.0

0.2

0.4

0.6

0.8

1.0

-23 -11.5 0 11.5 23

Distance (nm)

m ±as

±0.31-±0.37s fluctuation of protected units contributes to LER formation.

Monodisperse polymer

Randomly protected polymer

14 mJ cm-2

15 mJ cm-2

16 mJ cm-2

17 mJ cm-2

18 mJ cm-2

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

10 m

J c

m-2

Relationship between latent and SEM images of 30 nm L&S patterns

11 m

J c

m-2

67

5 n

m

1.77s

1.67s

0.36s

0.06s

Line Line Space

Line Line Space

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

12 m

J c

m-2

Relationship between latent and SEM images of 30 nm L&S patterns

13 m

J c

m-2

67

5 n

m

1.74s

1.64s

0.35s

0.72s

Line Line Space

Line Line Space

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

14 m

J c

m-2

Relationship between latent and SEM images of 30 nm L&S patterns

15 m

J c

m-2

67

5 n

m

1.56s

1.25s

0.88s

1.32s

Line Line Space

Line Line

Space

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

16 m

J c

m-2

Relationship between latent and SEM images of 30 nm L&S patterns

17 m

J c

m-2

67

5 n

m

1.09s

0.74s

2.10s

3.00s

Line Line Space

Line Line Space

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

18 m

J c

m-2

Relationship between latent and SEM images of 30 nm L&S patterns

19 m

J c

m-2

67

5 n

m 0.47s

0.30s

5.05s

3.57s

Line Line Space

Line Line Space

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Average +0.31s

-0.31s

s

Dissolution point

20 m

J c

m-2

Relationship between latent and SEM images of 30 nm L&S patterns

67

5 n

m

0.05s

6.08s

Line Line Space

The same analysis was applied to the other SEM images.

Start of space appearance

-3

-2

-1

0

1

2

3

15 20 25 30 35

Half-pitch (nm)

Dif

fere

nce

in s

unit

Random

Monodisperse

Space

Line

Half-pitch (nm)

-3

-2

-1

0

1

2

3

15 20 25 30 35

Random

Monodisperse

Dif

fere

nce

in s

unit

Half open

Relationship between stochastic effect and resist pattern defects

Half-pitch (nm)

-3

-2

-1

0

1

2

3

15 20 25 30 35

Random Monodisperse

Dif

fere

nce

in s

unit

Elimination of bridge

Half-pitch (nm)

-3

-2

-1

0

1

2

3

15 20 25 30 35

Random Monodisperse

Dif

fere

nce

in s

unit

Appearance of severe line shrinkage

Half-pitch (nm)

-3

-2

-1

0

1

2

3

15 20 25 30 35

Random

Monodisperse

Dif

fere

nce

in s

unit

Appearance of line break

Definition of “Difference in s unit”

(CP-CDP)/s

Difference between protected unit concentration and dissolution point

0.0

0.2

0.4

0.6

0.8

1.0

-30 -15 0 15 30

Distance (nm)

No

rmal

ized

pro

tect

ed u

nit

co

nc.

Summary – Advanced Resist Characterization

Line Line Space

Latent image was successfully reconstructed from SEM image using a Monte Carlo simulation.

The acid diffusion length was approximately 10 nm in the high resolution region.

To eliminate line shrinkages within 6.1 mm length, 1.2-1.6s difference is required.

±0.31-±0.37 s fluctuation of protected units contributed to LER formation.

To eliminate bridges within 6.8 mm length, 1.5-2.0s difference is required.

The fluctuation of protected unit (stochastic effect) was estimated.

The resolution dependence of dissolution point was estimated.

Chemical information was also obtained.

s

EIDEC Standard

Resist

This work was partially supported by the New Energy and

Industrial Technology Development Organization (NEDO).

Acknowledgement