IEUVI-TWG 2020 1

EUV Photoresists:

What Needs to be Done?

SUNY Polytechnic Institute;

College of Nanoscale Science and Engineering

Robert L. Brainarda

IEUVI-TWG 2020 2

There is a Lot that Needs to

be Done!

But There are Many

Opportunities!

IEUVI-TWG 2020 3

I. The Mechanisms of the

Exposure of EUV Resists

We Don’t Know

Much About:

We Have Not Reached

the RLS Potential of EUV

Resists:

II. Metal-Containing Resists

Are the Best Solution to RLS.

III. What is the Best Way for EUV Resists to

Reach their Full Potential?

IEUVI-TWG 2020 4

Mechanisms of EUV Exposure are Very Complicated

How

Many?

EUV

hn92 eV

e-/p+ +C,H,O

e- (80 eV) 2 eV

• 80, 60, 40, 30, 20, 10, 5, 2 eV electrons probably behave completely

different from each other.

• The electrons have very short life-times.

• The electrons cannot be detected or measured without leaving the film!

IEUVI-TWG 2020 5

There is a Lot We Don’t Know

How many

e-/p+ pairs do

we make?

EUV

hn92 eV

e-/p+ +C,H,O

Can two

e-/p+ pairs exist

in the same time

and space?

What is the

probability that

e-/p+ pairs will

recombine?

How far do

e- go?

How is H+

generated

from PAG?

IEUVI-TWG 2020 6

2008: Is it possible to “Titrate” the number of e- using PAG?

Two PAG-Loading studies agree:

Quantum Yields of 5-6 H+/Photon are possible.

2. Higgins & Brainard SPIE 2008 and JJAP 2011 V. 50

3. Kozawa & Tagawa JJAP 2010 V. 49

0

1

2

3

4

5

6

0 100 200 300 400 500 600 700 800

DPI TfTPS TfNDI Tf

Fil

m Q

ua

ntu

m Y

ield

[PAG] (mM)

Higgins & Brainard2

Max H+ = 5.6

Kozawa & Tagawa3

Max H+ = 4.9

0

1

2

3

4

5

6

0 200 400 600 800 1000

Film

Qu

an

tum

Yie

ld

[PAG] (mM)

I.A. How Many Electrons?

IEUVI-TWG 2020 7

Electron Trappinge- + PAG H+

(DE = 2-3 eV)

e- + Ph3S+ X- Ph3S

.X- H+

Hole-Initiated Chemistry(Kozawa Mechanism)

p+ + Polymer H+

Our Current Thinking: Both of these

mechanisms can occur independently.

I.A. How Many Electrons? By What Mechanisms?

IEUVI-TWG 2020 8

I.A. Both Electron-Trapping and Hole-

Chemistry Can Independently Create Acid

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07

Acid

/nm

3

Absorbed EUV Photons/nm3

FQY = 5.1 ± 0.1

FQY = 3.1 ± 0.1

Highest dose: 36 mJ/cm2

15 wt.%

Electron-Trapping

~60%Hole-Chemistry

~40%

~40%

~60%

Narasimhan, Grzeskowiak, Denbeaux & Brainard SPIE 2017

IEUVI-TWG 2020 9

6 H+/Photon

What Mechanism(s)?

• If e- only: Need 6-10 e-

• If p+ only: Need 6-10 p+

• If both e- and p+: Need 6-10 e-/p+ Pairs

• If either e- or p+: Need 3-5 e-/p+

How Many

e/p+ Pairs?

Quantum Yield

Triangle

Our Group

Kozawa & Tagawa

4. Narasimhan & Brainard SPIE 2017

Supports this conclusion

I.A. How Many Electrons per Absorbed Photon?

Yields in Chemistry

are ~70-100%

IEUVI-TWG 2020 10

IB. How Far Do They Go?

E-Beam Depth Studies: Experiment and Modeling

Vary Dose

& Voltage

Thickness Loss

(Ellipsometry)

Depth of Reactions1

Bake and Develop

• We used top down exposures and measure the depth to represent

the lateral electron travel away from the EUV absorption site.

• We studied 2000, 700, 250 and 80 eV electrons.

Narasimhan, Grzeskowiak, Denbeaux

& Brainard JM3 (2015)

IEUVI-TWG 2020 11

Experiment:

Thickness Loss of Open-Source Resist

0

10

20

30

40

50

60

0.01 0.1 1 10 100 1000

Th

ickn

ess L

oss (

nm

)

Dose (µC/cm2)

2000 eV 700 eV

250 eV

80 eV

15 wt.%

1.5 wt% TBAL

Film Thickness

Direct Electron Exposure

IEUVI-TWG 2020 12

I.B. Histograms from Modeling 100,000 Electrons

vs. EUV Absorption Site

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0.2

0 10 20 30 40 50 60

Num

ber

of E

nerg

y L

oss E

vents

per

Ele

ctr

on p

er

0.1

nm

Depth in Resist (nm)

80 eV

250 eV

700 eV

2000 eV

Fully-Stochastic Monte

Carlo Simulation

Program: LESiS

Originally created by Leo Ocola

and rewritten by our group:

Narasimhan JM3 (2015).

IEUVI-TWG 2020 13

I.B. Thickness Loss Simulation Results

0

10

20

30

40

50

60

0.01 1 100 10000

Thic

kness L

oss (

nm

)

Dose (μC/cm2)

700 eV

LESiS

Model

Experiment

0

10

20

30

40

50

60

0.01 1 100

Thic

kness L

oss (

nm

)

Dose (μC/cm2)

2000 eV

LESiS

Model

Experiment

Threshold set to match 700 eV simulation and experimental data

Low-Energy

Transitions

IEUVI-TWG 2020 14

I.B. Thickness Loss Simulation Results

0

5

10

15

20

25

0.01 1 100 10000

Thic

kness L

oss (

nm

)

Dose (μC/cm2)

80 eV

LESiS

Model

Experiment

0

5

10

15

20

25

30

0.01 1 100 10000

Thic

kness L

oss (

nm

)

Dose (μC/cm2)

250 eV

LESiS

Model

Experiment

Threshold set to match 700 eV simulation and experimental data

IEUVI-TWG 2020 15

I.B. What is Missing from Our Model?

Electron Trappinge- + PAG H+

(DE = 2-3 eV)

e- e-(4) Plasmon

Generation:

DE ≈ 3-24 eV

e-e-

e-(2) Ionization:

DE = 10-12 eV

Binding Energy Energy-

Loss

Events

5 eV

Count when

Electrons

“Fall below 5 eV”

Ionization

Plasmon

(CSDA)Continuous

Slowing-Down

Approximation

Thanks to Liam Wisehart

IEUVI-TWG 2020 16

0

10

20

30

40

50

60

0.01 1 100 10000

Thic

kness L

oss (

nm

)

Dose (µC/cm2)

0

10

20

30

40

50

60

0.01 1 100 10000

Thic

kness L

oss (

nm

)

Dose (µC/cm2)

I.B. Thickness Loss Simulation Results

A better model when 3 or 5 eV transitions are included.

No

Transition

3 eV

Transition

5 eV

Transition

No

Transition

Experiment 5 eV

Transition

3 eV

Transition

Experiment

250 eV 80 eV

Threshold set to match 700 eV simulation and experimental data

Low-Energy

Transitions

IEUVI-TWG 2020 17

I.C. Do Multiple e-/p+ Pairs Coexist in Time and Space?

I.D. Can an e- Fall into Another hn Hole?

EUV

hn = 92 eV

e-

EUV

hn = 92 eV

e-

In order to answer this question we must know:

1) The arrival rates of photons (We Know)

2) The cross-section for electron/hole recombination (vs. e- energy) (Don’t Know)

3) The lifetimes of the electrons and holes (Don’t Know)

IEUVI-TWG 2020 18

I.C. Electron Energy vs. Travel Time

10-16

10-14

10-12

10-10

10-8

0

10

20

30

40

50

60

70

80

90

Time (s)

Ele

ctr

on E

nerg

y (

eV

)

200 EUV photons on OS2

Average time between

arrival of photons in

100 x 100 nm area

*

*Photoelectron

escaped into vacuum

100 nm

Modelled

by LESiS

Can’t Co-Exist; Can’t Interact

IEUVI-TWG 2020 19

I.E. Conclusions about Mechanisms

• How many electrons are made?Our opinion: 3-5 e-

• By what mechanisms is acid generated?Our opinion, both electron-trapping and hole-initiated reactions occur independently.

• How far do electrons travel?2-5 nm. Our modeling matches our experimental results.

Better question: How far do they travel and still react?

• Do Multiple e-/p+ Pairs Coexist in Time and Space?No—by at least 3 orders of magnitude

• Can an e- Fall into Another hn’s Hole?They don’t coexist, so probably not.

Five Key Questions:

IEUVI-TWG 2020 20

II. 2011-2012: Oregon State and Cornell University

Hafnium-Oxide Resists

OSU/Inpria1

36-nm h/p lines

12 mJ/cm2

Cornell2

Improve resist stochastics by incorporating metals with

high EUV absorptivity (Thackeray).3

1. Stowers et al., SPIE, 2011.

2. Trikeriotis et al., SPIE, 2012.

3. Thackeray JM3, 2011.

20-nm lines

IEUVI-TWG 2020 21

Cobalt Platinum

& Palladium

Tin

Antimony

Inpria

&

Cornell:

Cardineau et al.; SPIE 2014

Molecular Organometallic Resists for EUV (MORE)Our group wrote a proposal to Intel to look at the rest of

the periodic table.

IEUVI-TWG 2020 22

II. Motivations for Metal-Based Resists(Inpria Resist Design Principles)

High selectivity etch

directly into SOC

Polymer

MOx

Cluster

Small Building Blocks

High Absorbance

High Material Homogeneity

Low

Electron

Blur

High Etch Selectivity

IEUVI-TWG 2020 23

II. Wide Process Window

for Inpria SnOx Resists on NXE-3300

DRAM: 40 nm Pitch

Dense Pillars

DtS: 52 mJ/cm2

LCDU: 2.4 nm

DOF > 140 nm

ELmax: >30%

Logic: 13 nm HP L/S

DtS: 33 mJ/cm2

DOF > 200 nm (10% EL)

ELmax: 22%

LWR: 3.4 nm

Large Process Window

Printable to < 10 nm L/S

IEUVI-TWG 2020 24

0

0.5

1

1.5

2

2.5

3

3.5

20253035404550

Ro

ug

hn

es

s (

nm

)

Feature Size (nm)

50 nm 35 nm 25 nm 22 nm

Dose = 300 mJ/cm2

LWR

LER

LER = Line Edge Roughness

LWR = Line Width Roughness

Del Re, et al., JM3 (2015)

II. Mono-Nuclear Tin Carboxylates

with Remarkable LER

Imaged at PSILER = Line Edge Roughness

LWR = Line Width Roughness

IEUVI-TWG 2020 25

JP-20

Development: Hexanes 20s

Dose = 5.6 mJ/cm2

Development: Water 20s

Dose = 9.2 mJ/cm2

We discovered this resist system in 2014. One of its remarkable features is

that development in either water or hexanes yields negative-tone imaging.

II. Antimony MORE Resists

Passarelli, et al. JM3 (2015)

IEUVI-TWG 2020 26

• Inpria started with HfOx resists and is now manufacturing SnOx resists,

that are under evaluation in fabs around the world.

• CNSE has explored multiple platforms, targeting highly absorbing

metals.

• The Major Advantages of Metal-Containing Resists Are:

− Increased absorbance for better photon stochastics.

− Single-component systems for better homogeneity.

− Smaller molecular size.

− Huge etch resistance.

II. Metal-Containing Resists: Overview

IEUVI-TWG 2020 27

III. Can Metal-Based Resists Replace CAR’s in

Some Applications?

I think it is inevitable.

If So, Why Hasn’t It Happened Already?

It’s Complicated…• Metal-Fab Integration Issues

• Redesign of Manufacturing Protocols

• Many New EUV Exposure Mechanisms

But I don’t think CAR to MOx is as complicated as 193-nm to EUV.

IEUVI-TWG 2020 28

III. Industrial Experience:

CAMP vs. Metal-Containing Resists

1.Ito, Willson, Frechet CAMP Pat. App.

2.Kunz et al. SPIE

3.Brainard - Resist to EUV-LLC

4.Stowers - MNE

36 years x 90% Market/Research 10 years x 3%

Market/Research

32 Industry-Years 0.3 Industry-Years

DUV 193-nm EUV

19821

19922

19983

20084

Metal-

Containing

Resists

Roughly 100:1

Industrial Experience

of CAMP vs. Metal

IEUVI-TWG 2020 29

Sematech is Gone….

Fundamental Understanding of:

• Mechanisms of Traditional CAMP Resists Remain Poorly Understood.

• New Metal Resists Need to be Discovered.

• Each new Metal Resist will have a Separate Mechanism.

Billions of $$ have been spent to develop the Physics and Engineering of EUV.

This ndustry needs to find a way to support research in the chemistry of EUV

resists.

III. What is the Best Way for EUV Resists to Reach

their Full Potential?

IEUVI-TWG 2020 30

IV. Acknowledgements

Inpria

Stephen Meyers

Jason Stowers

Andrew Grenville

BMET Team

Patrick Naulleau

Chris Anderson

Sematech

Mark Neisser

Stefan Wurm

Chandra Sarma

SUNY Polytechnic Institute (CNSE)

Professor Greg Denbeaux

Students:

Amrit Narasimhan Steven Grzeskowiak

Michael Murphy Brian Cardineau

James Passarelli Jodi Hotalen

Paul Scherrer Institut (EUV Exposures)

Michaela Vockenhuber

Yasin Ekinci

IEUVI-TWG 2020 31

Appendix

IEUVI-TWG 2020 32

II. How Many Electrons: Chemical Mechanisms

Thackeray18 and Kozawa2 think that EUV resists need both phenolic

polymers and electron trapping to generate acid.

(18) Thackeray et al. SPIE 2013

(2) Kozawa et al.SPIE 2010

No acid?

IEUVI-TWG 2020 33

Possible Mechanism of Electron-Trapping

IEUVI-TWG 2020 34

II. Fully-Stochastic Monte Carlo Simulation

Program: LESiS1

Photoelectron:EUV hn

e-

Yeh et al.5 (1985)

e-e-

e-Ionization:

Gryzinski et al.6 (1965)

Vriens et al.7 (1964-69)

e- e-Plasmon

Generation:

Ferrel8 (1956)

Quinn et al.9 (1962)

Elastic Scattering: e-DE = 0

Mott & Massey (1965)

Input Data and Theory from

‘56-’85 gas phase experiments.

Key Assumptions:

• The gas phase work applies to

EUV.

• Plasmons do not generate

electrons.

• Continuous Slowing Down does

not generate electrons

.

1. Originally created by Leo Ocola

and rewritten by our group: A. Nara-

simhan, et al., JM3 14 (4), 043502

(2015).

IEUVI-TWG 2020 35

II. Fully-Stochastic Monte Carlo Simulation

Program: LESiS1

1. Originally created by Leo Ocola and

rewritten by our group: A. Narasimhan, et

al., JM3 14 (4), 043502 (2015).

.

Secondary e-

Secondary e-

Input e-

(80 eV)

2 nm

2.5

nm

IEUVI-TWG 2020 36

V. Can Metal-Based Resists Replace

CAR’s in Some Applications?

Gregg Gallatin:

RLS trade-off

Gallatin, EUV 2007 Symposium

3

1

REQLER

Sensitivity

LER

Reso

luti

on

Sensitivity

LER

Reso

luti

on

Reso

luti

on

Absorbance

Target transmission for EUV resists is 50%.

James Thackeray, SPIE 2011 Plenary Presentation

At film thickness of 20 nm, PHS will only stop 10% of photons.

IEUVI-TWG 2020 37

I.C. Electron Travel Time (LESiS)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.E-15 1.E-12 1.E-09 1.E-06 1.E-03 1.E+00

Pro

babili

ty o

f an E

lectr

on T

ravelli

ng T

his

T

ime B

efo

re B

ein

g T

rapped

Travel Time before Cutoff(s)

Distribution of Electron Travel Time

104 EUV photons on

OS2 for each data set

Time between EUV

Pulses

Average time between

arrival of photons in

100 x 100 nm area

5 eV

Cutoff

8 eV

Cutoff3 eV

Cutoff

100 nm

Modelled

by LESiS