View
5
Download
1
Category
Preview:
Citation preview
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
Recommended