10/2/08 1

RLS Trade-Off: Questions about Molecular Size and Quantum Yield

Robert Brainard and Craig Higgins

Supported by SEMATECH and Rohm and Haas

10/2/08 2

Outline

I. Introduction

II. Effect of Molecular Weight

III. Quantum Yield

IV. Ultra High PAG Resists

V. Conclusions and Questions

10/2/08 3

RLS Trade-Off1

New materials/approaches are needed to “break-through” to new

performance surfaces:

• Reduction in Polymer Molecular Weight

• Increasing Quantum Yield

(1) Brainard, DARPA Review 2002

2

4

6

8

10

12

14

0 10 20 30 40 50 60 70 80

EUV LER 100nmEUV LER 150-200nm

()

Esize (mJ/cm2)

LER

(nm

)[Base]

EUV-2D Base Study

Res

olut

ion

Sensitivity

LER

Surfaces defined byZ-Parameter or KLUP

10/2/08 4

Rg = 3-4 nmDiameter = 6-8 nm

Polymer Radius of Gyration

II. Effect of Molecular Weight

Can we reduce LER by decreasing the size of the polymer?

In this conference:61% of the papers

about new resist materialsare about Molecular Glasses

10/2/08 5

II. Effect of Polymer Mw on LER:EUV-2D using 0.088 NA in 20032

100 nm Line8 nm 3-σ LER

23 nm

8 nm Rg = 2-6 nmDiameter = 4-12 nm

0

1

2

3

4

5

6

7

0 5 10 15 20 25 30 35

Rad

ius

(nm

)

Mw K g/mol

Radius ofGyrationIn THF

Smallest SphereDensity = 1 g/mL

R = 0.9-1.8 nmDiameter = 1.8-3.6 nm

6 New Polymers Mw = 3-33 Kg/mol

Changes in Mw will alter dissolution properties:

Vary [PAG] and [base]

(2) Cutler & Brainard, SPIE 2003

10/2/08 6

[PAG] and [Base] Variations over Wide Polymer Mw Range Round 1 Round 2 Round 3

6 MwsChanging UFTLs

6 Mws, 6 [PAG]Constant UFTLChanging Eo

6 Mws6 [PAG], 6 [Base]Constant UFTL, Eo

1

2

3

4

5

6

7

8

9

5

10

15

20

25

0 5 10 15 20 25 30 35

Eo

(mJ/

cm2) U

FTL (nm)

Polymer Mw Kg/mole

DUV Eo

UFTL

1

2

3

4

5

6

7

8

9

5

10

15

20

25

0 5 10 15 20 25 30 35

Eo

(mJ/

cm2) U

FTL (nm)

Polymer Mw Kg/mole

DUV Eo

UFTL

1

2

3

4

5

6

7

8

9

5

10

15

20

25

0 5 10 15 20 25 30 35

Eo

(mJ/

cm2) U

FTL (nm)

Polymer Mw Kg/mole

DUV Eo

UFTL

UFTL = Unexposed Film Thickness Loss or Dark Loss

10/2/08 7

MW has no effect on LER

LER is UNAFFECTED by an order of magnitude change in polymer Mw

when UFTL is held constant.

Round 1 Round 2 Round 3

6 MwsChanging UFTLs

6 Mws, 6 [PAG]Constant UFTLChanging Eo

6 Mws6 [PAG], 6 [Base]Constant UFTL, Eo

2

4

6

8

10

12

14

0 5 10 15 20 25 30 35

EUV

LER

(3σ)

, nm

Polymer Mw Kg/mole

2

4

6

8

10

12

14

0 5 10 15 20 25 30 35

EUV

LER

(3σ)

, nm

Polymer Mw Kg/mole

2

4

6

8

10

12

14

0 5 10 15 20 25 30 35

EUV

LER

(3σ)

, nm

Polymer Mw Kg/mole

10/2/08 8

II. Molecular Glasses (MG)

BG

PAG

CO2H

HO3S

hνStart SmallStay Small

1) (My observation) Positive Molecular Glasses DO NOTshow improved LER, unless they are fairly slow.

2) Negative resists look good Because of polymer formation?

3) Champion resists appear to result from polymeric resists.

10/2/08 9

Questions about the role of Molecular Size:

(1) Our polymer Mw work was performed in 2003 at 0.088 NA. Should it be repeated at 0.3 NA and/or IL?

(2) Should Molecular Glass Resists be included in a systematic study against polymeric resists?

3) Why do negative MG resists appear to give good LER/Sensitivity? Must polymers be involved for top performance?

10/2/08 10

III. Can we Beat RLS by IncreasingFilm Quantum Yield?

We propose that higher quantum yield will allow us to

go from…

Here

to

Here

…with no penalty to sensitivity.

3

1REQ

LERα

∝

Increased quantum yield can break the RLS tradeoff

Gregg Gallatin:3

(3) Gallatin, EUV 2007 Symposium

Number of AcidsGenerated in the Film

Number of PhotonsAbsorbed in the Film

FilmQuantum

Yield=

Number of AcidsGenerated in the Film

Number of PhotonsAbsorbed in the Film

FilmQuantum

Yield=

10/2/08 11

Quantum Yield Increaseswith PAG Concentration

# AcidsGenerated = [PAG](1 – e(-CE))(6.02 x 1023)

2.5

3

3.5

4

4.5

5

5.5

6

0.045

0.05

0.055

0.06

0.065

0.07

0.075

0.08

0.085

4 5 6 7 8 9 10 11

Eo (m

J/cm

2) C-Param

eter[PAG] (wt%)

Eo

C-Parameter

1.5

2

2.5

3

4 5 6 7 8 9 10 11Fi

lm Q

uant

um Y

ield

[PAG] (wt%)

1.80

2.30

2.95

Used SzmandaBase-Titration Method

10/2/08 12

0

10

20

30

40

50

60

70

80

90

100

0.00 5.00 10.00 15.00 20.00 25.00 30.00

CD3·Esize (1E-9 mJ·nm)

LER2 (n

m2 )

5% PAG7.5 % PAG10% PAG

(B)

0

10

20

30

40

50

60

70

80

90

100

0.00 5.00 10.00 15.00 20.00 25.00 30.00

CD3·Esize (1E-9 mJ·nm)

LER2 (n

m2 )

5% PAG7.5 % PAG10% PAG

(B)

Higher [PAG]:• Higher FQY• Lower Z3,4

(4) Wallow, Higgins andBrainard, SPIE 2008

10%PAG

How far can we push [PAG]?

(3) Gallatin, EUV 2007 Symposium

10/2/08 13

PAG + e-

→ H+

IV. Ultra High PAG Resists

Film

Qua

ntum

Yie

ld

[PAG]

hν

Organic Polymer Film: Primarily C, H, O

e- + h+

65-87 eV e- + h+

e- + h+

e- + h+

e- + h+

+ PAG H+

+ PAG H+

+ PAG H+

+ PAG H+

hν

Organic Polymer Film: Primarily C, H, O

e- + h+

65-87 eV e- + h+

e- + h+

e- + h+

e- + h+

+ PAG H+

+ PAG H+

+ PAG H+

+ PAG H+

EUV

Question #1: How high can we make FQY?

Question #2: Do we improve RLS?

Question #3: Can we determine how many photoelectrons are made?

10/2/08 14

IV. High PAG Resist Platformsfor FQY and Imaging RLS Study

Photoacid Generator

(PAG)

BaseTBAH

Polymer

65/20/15

O O

OH

Resist Formulations

OH-

Iodonium PAG (I+)DTBI-PFBS

Sulfonium PAG (S+) TPS-PFBS

(5) Hassanein, Higgins, Thackeray, Brainard et al SPIE (2008)

10/2/08 15

Film Quantum Yields vs. [PAG]

2

4

6

8

10

12

14

0 0.2 0.4 0.6 0.8 1

I+ AMETI+ BMETS+ BMET

Film

Qua

ntum

Yie

ld

[PAG] (mol/L)

12.5

8.7

4.3

I+

S+

(6) Brainard, Higgins et al., Journal of Photopolymer Science and Tech. (2008)

I+ FQY =~10 H+ / EUV hν

S+ FQY =~4 H+ / EUV hν

10/2/08 16

Resolution of Ultra-High PAG Resists

5 wt%(0.083 mM)

120 nm 100 nm 80 nm 60 nm 50 nm 45 nm 40 nm

Iodonium PAG80 nm Film Thickness

1:1 Line/Space through PAG Loading

7.5 wt%(0.123 mM)

15 wt%(0.247 mM)

20 wt%(0.330 mM)

30 wt%(0.532 mM)

40 wt%(0.697 mM)

10/2/08 17

Resolution of Ultra-High PAG Resists

- Adhesion Failure

- Pattern Collapse

- Top Loss

Patterning Issues at Very High PAG Loadings:

I+ 125 nm FTI+ 80 nm FTS+ 125 nm FT

S+ 125 nm FT:

I+ 80 nm FT:

I+ 125 nm FT:

Resolution is consistent, then degrades at > 20% PAG

20

40

60

80

100

120

0 10 20 30 40 50

(p

)

[PAG] (wt%)

Res

olut

ion

(nm

)

10/2/08 18

Sensitivity of Ultra-High PAG Resists

I+ 125 nm FTI+ 80 nm FTS+ 125 nm FT

Esize (100 nm L/S) Data:

- Saturated at 15-20% PAG

Eo Data:

- Saturated at 15-20% PAG

All data obtained from BMET

0

2

4

6

8

10

12

0 10 20 30 40 50

Dos

e (m

J/cm

2 )

[PAG] (wt%)

10/2/08 19

I+ 125 nm

I+ 80 nm

S+ 125 nm

LER is consistent, then degrades > 30-40 wt% PAG

0

2

4

6

8

10

12

0 10 20 30 40 50

LER

(nm

)

[PAG] (wt%)

100 nm Half-Pitch

LER of Ultra-High PAG Resists

10/2/08 20

Exposure Latitude, Acid Diffusion and KLUP

Curves: half pitch

All Data at 125 nm Film Thickness

Iodonium PAG

Exposure Latitude decreases with [PAG]

0

0.1

0.2

0.3

0.4

0.5

0 10 20 30 40 50

120 nm100 nm80 nm 60 nm50 nm45 nm

Expo

sure

Lat

itude

[PAG] (wt%)

Curves: half pitch

Sulfonium PAG

0

0.1

0.2

0.3

0.4

0.5

0 10 20 30 40 50

120 nm100 nm80 nm 60 nm50 nm45 nm

Expo

sure

Lat

itude

[PAG] (wt%)

10/2/08 21

Acid Diffusion Increases with [PAG]

I+ 125 nm FT

I+ 80 nm FT

S+ 125 nm FT

Why does LD increase for increasing PAG? (7) Van Steenwinckel, Lammers,

Koehler, Brainard, and Trefonas JVST (2005)

NILSELMTFDIFF ∝

Acid diffusion was determined from exposure latitude using

the following method:7

20

40

60

80

100

120

140

0 10 20 30 40 50

Aci

d D

iffus

ion

Leng

th (n

m)

[PAG] (wt%)

10/2/08 22

Ultra-High PAG Resist Performance: KLUP

Best performance is at ~20% PAG:

Sensitivity Gains are Cancelled by Acid Diffusion Increases

100 nm Line/Space

0

0.2

0.4

0.6

0.8

1

0 10 20 30 40 50

Esize ResolutionLEREL

Nor

mal

ized

Val

ue[PAG] (wt%)

Resolution

Sensitivity

LER Exposure Latitude

I+ (125 nm) FT

I+ 80 nm FT

S+ 125 nm FT

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 10 20 30 40 50

K LUP

[PAG] (wt%)

10/2/08 23

V. Conclusions and Additional Questions

LER and Resolution Appear to be Flat with [PAG], but then degrades at > 30% PAG.

Sensitivity improves, but then flattens out.

EL decreases/ diffusion increases with [PAG]

The KLUP analysis shows that the sensitivity gains are cancelled by the increased diffusion.

10/2/08 24

Why do the improvements in sensitivity stop?

Style borrowed from Chris Anderson

10/2/08 25

Why do the improvements in sensitivity stop?

I+ 125 nm FT

I+ 80 nm FT

S+ 125 nm FT

0

2

4

6

8

10

12

0 10 20 30 40 50

Dos

e (m

J/cm

2 )

[PAG] (wt%)

a) Base is being overwhelmed

b) Used all available electrons (but why the high FQY for I+?)

c) Not enough deblockinggroups

Esize

Eo

10/2/08 26

Why does diffusion increase with [PAG]?

10/2/08 27

Why does diffusion increase with [PAG]?a) Film Tg may change

b) Acid solubility parameter may change

c) More free volume for acid to diffuse

d) Base is being overwhelmed

0

0.1

0.2

0.3

0.4

0.5

0 10 20 30 40 50

120 nm100 nm80 nm 60 nm50 nm45 nm

Expo

sure

Lat

itude

[PAG] (wt%)

20

40

60

80

100

120

140

0 10 20 30 40 50

Aci

d D

iffus

ion

Leng

th (n

m)

[PAG] (wt%)

Acid DiffusionExposure Latitude

10/2/08 28

2) Test Low Diffusion Material and Processesa) Lower PEB temperature

Increased generated acids no longer need to diffuse as far

b) Increase TBA deblocking group in polymer

[TBA] may be limiting deblocking rate in high [PAG]

Optimize surface for better adhesion

V. Planned and Possible Future Work1) Verify Film Quantum Yield Results

(SEMATECH Funded – In Collaboration with G. Denbeaux)

a) Use direct method for determining optical density

b) Use acid sensitive dye to directly measure acid generation

10/2/08 29

Will Smaller Molecules give better LER?

10/2/08 30

V. Possible Future Work

3) Side-by-side comparison of polymeric and MG resists using:

• 10 X range in Polymer Mw

• Best MG available

• High Resolution tools (BMET 0.3 NA, PSI IL)

• Apples-to-Apples comparison (KLUP or Z Parameter)

10/2/08 31

Break-Through Strategies:An Editorial

Res

olut

ion

Sensitivity

LER

Surfaces defined byZ-Parameter or KLUP

Higher Absorption

Molecular Glass (Pos)

Molecular Glass (Neg)

Higher Quantum Yield

High ERED PAGs

Ultra-High PAG

Anisotropic Diffusion

10/2/08 32

Acknowledgements

CNSE:

Craig Higgins

Srividya Revuru

Alin Antohe

Greg Denbeaux

Richard Matyi

SEMATECH:

Jacque Georger

Kim Dean

Andrea Wüest

Berkeley:

Patrick Naulleau

Chris Anderson

Rohm and Haas:

Jay Machevich

Charlotte Cutler

Jim Thackeray

Peter Trefonas

Kathleen Spear

Applied Math Solutions:Gregg Gallatin

10/2/08 33

Future WorkWe propose that higher quantum

yield will allow us toimprove resolution, LER…

…with no penalty to sensitivity.

Future Work:

Increased generated acids no longer need to diffuse as far

This Work- Increased the # of Acids- Saturated Sensitivity Improvement- Used Constant Process

10/2/08 34

Traditional Quantum Yield Film Quantum Yield

Experiments are done in transparent solvents

Light is primarily absorbed by single molecules

QY < 1

hν

Organic Polymer Film: Primarily C, H, O

e- + h+

65-87 eV e- + h+

e- + h+

e- + h+

e- + h+

+ PAG H+

+ PAG H+

+ PAG H+

+ PAG H+

hν

Organic Polymer Film: Primarily C, H, O

e- + h+

65-87 eV e- + h+

e- + h+

e- + h+

e- + h+

+ PAG H+

+ PAG H+

+ PAG H+

+ PAG H+

EUV

vs.

Moles of AcidsGenerated in the Film

Moles of PhotonsAbsorbed by the Film

FilmQuantum

Yield≡Quantum

YieldMoles of Product

Moles of Photons Absorbed

≡

Light is absorbed by everything in the film.

Multiple electrons are made.

FQY of Acid > 1

What is Film Quantum Yield?

92 eV