euv source for high volume manufacturing: performance at ... · 2017 source workshop, dublin,...
TRANSCRIPT
2017 Source Workshop, Dublin, Ireland, November 7th
EUV Source for High Volume Manufacturing:
Performance at 250 W and
Key Technologies for Power Scaling
Igor Fomenkov
ASML Fellow
• Background and History
• EUV Imaging
• Principles of EUV Generation
• EUV Source: Architecture
• EUV Sources in the Field
• Source Power Outlook
• Summary
Outline
2017 SW
Public
Slide 2
Background and History
2017 SW
Slide 3
Public
Why EUV? - Resolution in Optical Lithography
2017 SW
Public
Slide 4
Critical Dimension:
𝐶𝐷 = 𝑘1 ×
𝑁𝐴
Depth of focus:
𝐷𝑂𝐹 = 𝑘2 ×
𝑁𝐴2
k: process parameterNA: numerical aperture: wavelength of light
KrF-Laser: 248nm
ArF-Laser: 193 nmArF-Laser (immersion): 193 nm
EUV sources: 13.5 nm
EUV sourceWafer stage
Reticle stage
theoretical limit (air): NA=1
practical limit: NA=0.9
theoretical limit (immersion):NA ≈ n (~1.7)
k1 is process parameter
traditionally: >0.75
typically: 0.3 – 0.4
theoretical limit: 0.25
2017 SW
Slide 5
Public
EUV development has progressed over 30 yearsfrom NGL to HVM insertion
’84
1st lithography
(LLNL, Bell Labs,
Japan)
USA
40 nm hp
70 nm
L&S
Japan
80 nm
160 nm
NL
28 nm Lines and spaces
USA
5 mm
ASML starts
EUVL research
program
ASML ships 1st
pre-production NA
0.25 system
NXE:3100
ASML ships 1st
NA 0.33 system
NXE:3300B
NL
13 nm L/S
ASML ships 1st
HVM NA0.33 system
NXE:3400B
NL
7 nm and 5 nm
node structures
’85 ’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 ’97 ’98 ’99 ’00 ’01 ’02 ’03 ’04 ’05 ’06 ’07 ’08 ’09 ’10 ’11 ’12 ‘13 ’14 ’15 ’16 ’17 ’18
NL
19 nm Lines and spaces
ASML ships
2 alpha demo tools:
IMEC (Belgium) and
CNSE (USA)
USANL NL NL
0
5
10
15
20
25
30
35
40
45
50
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
CD
(n
m)
EUV resist: 4x resolution improvement in ten years 12nm half pitch resolved with non-CAR resist on 0.33 NA EUV
ADT, NXE:3100, NXE:33x0, NXE:3400B as measured by ASML/ IMEC,
Exposure Latitude > 10% and / or Line Width Roughness < 20%, Dose ≤ 35mJ/cm2
ADT NXE:3100 NXE:33x0 NXE:3400B
Non-CAR resist
Half Pitch: 12nm
Slide 6
2017 SW
Public
High-NA EUV targets <8nm resolutionRelative improvement:5X over ArFi, 40% over 0.33 NA EUV
2005 20151995 201019900.01
0.10
1.00
20202000
Year of introduction
Major technology step
(e.g. source, mirror)
EUV 0.33EUV >0.5
EUV 0.25
Engineering optimization of numerical
aperture resulting in a resolution step
comparable to historical wavelength
transitions
Slide 7
2017 SW
Public
i-Line 365 nm
PAS 2500/40
0.4 NA 700nm
PAS 5500/300
0.57 NA 250nm
KrF 248 nm
ArF 193 nm
AT:1100
0.75 NA 90nm
ArFi 193 nm
XT:1700i
1.2 NA 45nm
EUV 13.5 nm
NXE:3400
0.33 NA 13nm
2017 SW
Slide 8
Public
EUV roadmapSupporting customer roadmaps well into the next decade
Overlay [nm]55 WPH 125 WPH 185 WPH145 WPH
NXE:3300B
NXE:3350B
NXE:next
High NA
NXE:3400B2017
2015
2013
Introduction
products under study
7
3.5
3
<3
<2At customer upgradable
New platform
High-NA Field and Mask Size productivityThroughput >185wph with anamorphic Half Fields
Fast stages enable
high throughput
despite half fields
Thro
ugh
pu
t [3
00
mm
/hr]
Source Power/Dose [W/(mJ/cm2]
Throughput for various source powers and doses
0
20
40
60
80
100
120
140
160
180
200
0 5 10 15 20 25 30 35
500 Watt
30 mJ/cm2
0.33NAWS, RS current performance
WS 2x, RS 4x
HF
High-NA anamorphic
FF
2017 SW
Slide 9
Public
250 Watt
20 mJ/cm2
EUV Imaging – NXE:3400B
2017 SW
Slide 10
Public
Overlay
Imaging/Focus
Productivity
NXE:3400B: 13 nm resolution at full productivitySupporting 5 nm logic, <15nm DRAM requirements
Resolution 13 nm
Full wafer CDU < 1.1 nm
DCO < 1.4 nm
MMO < 2.0 nm
Focus control < 60 nm
Productivity ≥ 125 WPH
Overlay set up
Set-up and modeling
improvements
Reticle Stage
Improved clamp flatness
for focus and overlay Projection Optics
Continuously Improved
aberration performance
New Flex-illuminator
Sigma 1.0 outer sigma,
reduced PFR (0.20)
Leveling (Optional)
Next generation UV LS
reduced process dependency Wafer Stage
Flatter clamps, improved
dynamics and stability
SMASH-X prepared
Metro frame prepared
for Smash-x
Slide 11
2017 SW
Public
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1,600,000
20
14
05
20
14
10
201
415
20
14
20
201
425
20
14
30
20
14
35
201
440
20
14
45
20
14
50
201
503
20
15
08
201
513
20
15
18
20
15
23
201
528
20
15
33
20
15
38
201
543
20
15
48
201
553
201
605
20
16
10
201
615
20
16
20
20
16
25
201
630
20
16
35
20
16
40
201
645
20
16
50
201
703
20
17
08
20
17
13
201
718
20
17
23
20
17
28
201
733
20
17
38
To
tal n
um
be
r o
f w
afe
rs e
xp
os
ed
Week
2017 SW
Slide 12
Public
>1.4M wafers exposed on NXE:3xx0B at customer sitesCurrently 15 systems running in the field. First system was shipped Q1 2013
>1.4 Million EUV Wafers at Customers
AL 2015
AL 2016
AL 2017
2017 SW
Slide 13
Public
Significant progress in system availability is recognized
by our customers
Source: Intel and TSMC presentations
at EUVL Symposium 2016, and SPIE 2017.
Slide 14
13nm LS and 16nm IS: full-wafer CDU 0.3 nmmeets 5 nm logic requirements, with excellent process windows
13
nm
LS
lea
fsh
ap
ed
ipo
le16 n
m I
Ss
ma
ll-c
on
ve
nti
on
al
(s_
ou
t=
0.5
)
non-CAR resist
CAR resist
2017 SW
Public
NXE:3400B illuminator: increased pupil flexibility at full throughput
Field Facet Mirror
Pupil Facet MirrorIntermediate
Focus
2017 SW
Public
Slide 15
New flex illuminator on NXE:3400B13nm resolution without light loss at 20% pupil fill ratio
First illuminators qualified and
currently being integrated in a system
The animation shows the 22 standard
illumination settings. They are measured in the
illuminator work center, using visible light and a
camera on top of the illuminator
2017 SW
Public
Slide 16
Slide 17
Two-fold approach to eliminate reticle front-side defects
2017 SW
Public
1. Clean scanner
Reflected
illumination
EUV Reticle (13.5nm)
Reticle
2. EUV pellicle
particle
Pellicle
Without Pellicle With Pellicle
Slide 18
Pellicle film produced without defects that print
2017 SW
Public
Improvement from Q3 2016 to now
# o
f defe
cts
Defect size in um
Zero defects
450
400
350
300
250
200
150
100
50
0
> 40
> 25-40
> 10-25
2017 SW
Slide 19
Public
ASML pellicle confirmed for use in NXE:3400B to at least 140W
Y-nozzle cooling can extend pellicle to >205W
NXE:3400B @ 140W
Power ramp in 4 steps: 95W, 115W, 125W, 140W
22nm PRP-i reticle with pellicle
0
20
40
60
80
100
120
140
160
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
So
urc
e p
ow
er
@ i
f (
W)
Wafer number
Pellicle removed from
scanner for analysis
Slide 20
DGL membrane as spectral filterlocated at Dynamic Gas Lock (DGL)
suppresses DUV and IR, plus removes outgassing risk to POB2017 SW
Public
DGL membrane (~ 50 x 25 mm)
>100x DUV suppression
>4x IR suppression
Effective DUV and IR suppression
EUV: Principles of Generation
2017 SW
Slide 21
Public
Laser Produced Plasma Density and TemperatureNishihara et al. (2008)
EUV
LPP
Ion density ~ 1017 – 1018 #/cm3
Temperature ~ 30 -100 eV
Slide 22
2017 SW
Public
Fundamentals: EUV Generation in LPPLaser produced plasma (LPP) as an EUV emitter Slide 23
2017 SW
Public
30 micron diameter tin droplet
Focused
Laser light
electrons
tin ions
“ejecta”microparticles
tin vapor
1. High power laser interacts with liquid tin producing a plasma.
2. Plasma is heated to high temperatures creating EUV radiation.
3. Radiation is collected and used to pattern wafers.
Tin Laser Produced Plasma
Image
EUV
EUV
EUV
EUV
Dense hot
Plasma
Plasma simulation capabilitiesMain-pulse modeling using HYDRA
Main-pulse
shape
1D simulations are fast and useful
for problems that require rapid
feedback and less accuracy
3D:
+ real asymmetric
profile
2D and 3D
simulations are run
for the full duration
of the Main pulse.
Results include
temperature,
electron density,
spectral emission,
etc.
Target
Sn target using a real
irradiance distribution
1D:
real pulse shape
2D:
+ symmeterized
beam profile
Electron density (top half) with laser light (bottom half)
500μm500μm500μm
500μ
m
Po
we
r (a
rb. U
nits)
Time
2017 SW
Public
Slide 24
Simulation of the EUV source
The plasma code’s outputs
were processed to produce
synthetic source data. The
comparison to experiments
helps to validate the code and
understand it’s accuracy.
Measured Shadowgrams
Simulated Shadowgrams
Simulated EUV spectra
Emission anisotropy
Collector rim
Reflected laser modelingC
onve
rsio
n E
ffic
iency (
%)
Target Diameter (µm)
Flu
en
ce
(J/c
m2
/S/m
m)
Wavelength (nm)
Simulation
Conversion Efficiency
Slide 25
2017 SW
Public
EUV Source: Architecture
and Operation Principles
2017 SW
Slide 26
Public
• Key factors for high source power are:High input CO2 laser power
High conversion efficiency (CO2 to EUV energy)
High collection efficiency (reflectivity and lifetime)
Advanced controls to minimize dose overhead
Vessel
With Collector, Droplet
Generator and Metrology
Focusing Optics
Ele
ctr
on
ics
Ele
ctr
on
icsControllers for Dose
and Pre-pulse
*∝EUV power
(source/scanner interface, [W]) CO2 power [W]
Conversion
Efficiency [%]1 – Dose
Overhead [%] *
Pre-PulseIs
ola
tor
PreAmp PASeed
op
tics
op
tics PAo
pti
cs
op
tics PAo
pti
cs
Power Amplifier
Beam Transport SystemFab Floor
Sub-Fab FloorMaster Oscillator
PA op
tics
Pre-pulse
requires seed
laser trigger
control
2017 SW
Slide 27
Public
LPP: Master Oscillator Power Amplifier (MOPA)
Pre-Pulse Source Architecture
NXE:3XY0 EUV Source: Main modulesPopulated vacuum vessel with tin droplet generator and collector Slide 28
2017 SW
Public
Droplet GeneratorDroplet Catcher
EUV Collector
Intermediate Focus
2017 SW
Slide 29
Public
EUV Source: MOPA + Pre-Pulse
Pre-pulse transforms tin droplet into
“pancake/mist” that matches
CO2 main pulse beam profile
MOPA = Master Oscillator Power Amplifier
PP = Pre-Pulse
MP = Main Pulse
Droplet (<30um)CO2 beam
Pre-pulse
CO2 beam
Main pulseExpanded target matched to main pulse
>5% conversion efficiency achieved
Pre-Pulse
Droplet stream
~80 m/s
Dropletz (mm)
Pre-pulse laser
Expands the droplet and prepares the Sn target
Main-pulse laser
Heats and ionizes the Sn target to produce EUV light
Movie: Backlight shadowgrams
from a 3300 MOPA+PP source
Main Pulse
2017 SW
Slide 30
Public
Forces on Droplets during EUV Generation
High EUV power at high repetition rates drives requirements for higher speed droplets with large space between droplets
2017 SW
Slide 31
Public
Droplet Generator: Principle of Operation
• Tin is loaded in a vessel & heated above melting point
• Pressure applied by an inert gas
• Tin flows through a filter prior to the nozzle
• Tin jet is modulated by mechanical vibrations
Nozzle
Filter
ModulatorGas
Sn
0 5 10 15 20 25 30-10
-5
0
5
10
Dro
ple
t p
ositio
n, mm
Time, sec
140 mm 50 mm 30 mmPressure: 1005 psi Frequency: 30 kHz Diameter: 37 µm Distance: 1357 µm Velocity: 40.7 m/s
Pressure: 1025 psi Frequency: 50 kHz Diameter: 31 µm Distance: 821 µm Velocity: 41.1 m/s
Pressure: 1025 psi Frequency: 500 kHz Diameter: 14 µm Distance: 82 µm Velocity: 40.8 m/s
Pressure: 1005 psi Frequency: 1706 kHz Diameter: 9 µm Distance: 24 µm Velocity: 41.1 m/s
Fig. 1. Images of tin droplets obtained with a 5.5 μm nozzle. The images on the left were obtained in
frequency modulation regime; the image on the right – with a simple sine wave signal. The images
were taken at 300 mm distance from the nozzle.
Short term droplet position stability σ~1mm
16 mm
2017 SW
Slide 32
Public
Droplet Generator: Principle of Operation
Multiple small droplets coalesce together to form larger droplets at larger separation distance
Large separation between the droplets by special modulation
2017 SW
Slide 33
Public
Droplet Generator: Principle of OperationLarge separation between the droplets by special modulation
Tin droplets at 80 kHz and at different applied pressures.
Images taken at a distance of 200 mm from the nozzle
1.5 mm
Increasing
Droplet Generator
Pressure
2017 SW
Slide 34
Public
Collector Protection by Hydrogen Flow
Sn droplet / plasma
H2 flow
Reaction of H radicals with Sn to form SnH4, which can be pumped away.Sn (s) + 4H (g) SnH4 (g)
• Hydrogen buffer gas (pressure
~100Pa) causes deceleration of ions
• Hydrogen flow away from collector
reduces atomic tin deposition rate
Laser beam
IF
Sn catcher
DG
EUV collectorTemperature controlled
• Vessel with vacuum pumping to
remove hot gas and tin vapor
• Internal hardware to collect micro
particles
2017 SW
Slide 35
Public
EUV Collector: Normal Incidence
• Ellipsoidal design
• Plasma at first focus
• Power delivered to exposure tool at second focus (intermediate focus)
• Wavelength matching across the entire collection area
Normal Incidence GradedMultilayer Coated Collector
Slide 36
Public
Productivity increases via source availabilitySecured EUV power is matched with increasing availability
Productivity = Throughput(EUV Power) Availability
Source power
from 10 W to > 250 W
Drive laser power from 20 to 40 kW
Conversion efficiency (CE) from 2 to 6% (Sn droplet)
Dose overhead from 50 to 10%
Optical transmission
Source availability
Drive laser reliability
Droplet generator reliability & lifetime
Automation
Collector protection
2017 SW
EUV Power= (CO2 laser power CE transmission)*(1-dose overhead)
Raw EUV power
EUV Sources in the Field
2017 SW
Slide 37
Public
2017 SW
Slide 38
Public
Productivity targets for HVMSource contribution to productivity
Conversion
efficiency
Dose
margin
Drive laser
power
Laser to
droplet
control
Optical
transmission
Overhead
optimization
Exposure
dose
Stage accuracy
at high speed
Automation
Collector
maintenance
Droplet
generator
maintenance
Drive laser
reliability
>1500 WPD
in 2016
Productivity
2017 SW
Slide 39
Public
EUV Source operation at 250W with 99.90% fields meeting dose spec
Operation Parameters
Repetition Rate 50kHz
MP power on droplet 21.5kW
Conversion Efficiency 6.0%
Collector Reflectivity 41%
Dose Margin 10%
EUV Power 250 W
Open Loop Performance
BaselineImproved Isolation
NXE scanner productivity above 125 wafers per hourNXE:3400B at 207W, 126 WPH
NXE:3400B ATP test: 26x33mm2, 96 fields, 20mJ/cm2
Slide 40
2017 SW
PublicT
hro
ug
hp
ut
[Wa
fers
pe
r h
ou
r]
110
100
90
80
70
60
50
40
30
20
10
02014Q1
2014Q2
2014Q3
2014Q4
2015Q3
2015Q4
2016Q2
2016Q4
2017Q1
NXE:3300Bat customers
NXE:3350BASML factory
NXE:3350Bat customers
130
120
2017Q3
NXE:3400BASML factory
2017Q3
NXE:3400Bat customers
NXE:3400BASML factory
(proto)
0
20
40
60
80
100
120
140
2014 2015 2016 2017target
2018target
Thro
ugh
pu
t [W
PH
]at
20
mJ/
cm2
Productivity roadmap towards >125 WPH in placeSlide 41
Public
Source power increase Source power increase
Transmission improvement
Faster
wafer swap
Transmission
improvement
Source power scaling
continues to support
productivity roadmap
Slide 41
2017 SW
Public
resulting in
~8 wph gain
>250W is now demonstrated,
shipping planned end of 2017
Progress for 2017: 250W demonstrated
∝EUV power
(source/scanner interface, [W]) CO2 power [W]
Conversion
Efficiency [%]*1 – Dose
Overhead [%] *
Increase average and peak laser power
Enhanced isolation technology
Advanced target formation technology
Improved dose-control technique
2017 SW
Public
Slide 42
System
Integration
08 09 10 11 12 13 14 15 16 17 18
Year
Dose c
ontr
olle
d E
UV
pow
er
(W)
0
25
50
75
100
125
150
175
200
225
250
Research
Shipped
Shipment in 2017
2017 SW
Slide 43
Public
Power Amplifier Chain Increases CO2 Power Good beam quality for gain extraction and EUV generation
Key technologies:1. Drive laser with higher power capacity 2. Gain distribution inside amplification chain3. Mode-matching during beam propagation4. Isolation between amplifiers5. Metrology, control, and automation
0
5000
10000
15000
20000
25000
30000
35000
40000
0 50 100 150 200 250 300 350
PA
s o
utp
ut
(W)
Seed power (W)
Power Amplifiers
PA1
PA2
PA3
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 20180
25
50
75
100
125
150
175
200
225
250
3100 NOMO (shipped)
3100 MOPA (research)
3100 MOPA+PP (research)
3300 MOPA+PP (shipped)
3400 MOPA+PP (research)
3400 MOPA+PP (Shipping 2017)
Do
se
co
ntr
oll
ed
EU
V p
ow
er
(W)
Year
0
0.5
1
1.5
2
2.5
Laser Gain (ratio to 3300 source)
3400
laser gain3300
laser gain
MOPA + Pre-pulse
Seed System with high
power pre-amplification
Scaling laser power requires laser isolation advances
From NXE:3300 to NXE:3400, enhanced isolation gives stable >2x increase laser gain
Unstable laser onset without enhanced isolation
MOPA + Pre-pulse
Seed System
with pre-amplification
Beam Transport &
Final Focus
VesselNXE:3300B Drive Laser
4-stage power amplificationImproved thermal
management
NXE:3400 Drive Laser
High-power 4-stage power amplification
+Enhanced
isolation
2017 SW
Public
Slide 44
0
1
2
3
4
5
6
7
Conversion Efficiency vs target type
Convers
ion e
ffic
iency (
%)
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 20180
1
2
3
4
5
6
3100 NOMO (shipped)
3100 MOPA (research)
3100 MOPA+PP (research)
3300 MOPA+PP (shipped)
3400 MOPA+PP (research)
3400 MOPA+PP (Shipping 2017)
Convers
ion E
ffic
iency (
%)
Year
Introduction
of prepulse
Advanced target
formation with
enhanced
isolation
Enhanced isolation leads to >205W EUV powervia advanced target formation for high CE
2017 SW
Public
Slide 45
Open Loop Performance >30%With enhanced
isolation
Before enhanced
isolation
Reduced
Sigma
3.9mJ5.1mJ
Benefits of enhanced isolation:
• Higher, stable CO2 laser power lower dose overhead
• High conversion efficiency operation higher pulse energy
Dose
ove
rhe
ad
requ
ire
d
to m
ee
t d
ose
sp
ec
Without
enhanced
isolation
30%
20%
10%<10.5%
~30%
With
enhanced
isolation
His
togra
m
His
togra
m
Enhanced isolation improves EUV performance
2017 SW
Public
Slide 46
Comparing two dose-control techniques at 210W:higher in-spec power with improved dose-control technique
0 500 1000 1500 2000 2500 3000 3500 40000
2
4
Time [s]
Dose E
rror
[%]
0 500 1000 1500 2000 2500 3000 3500 4000200
210
220
Time [s]
Pow
er
(M
ean)
[W]
0 200 400 600 800 1000 1200200
210
220
Time [s]
Pow
er
(M
ean)
[W]
0 200 400 600 800 1000 12000
2
4
Time [s]
Dose E
rror
[%]
Previous dose-control technique Improved dose-control technique
One Hour In Spec
209W, 61.6% die yield 210W in-spec
On the same EUV source, back-to-back performance comparing
previous and improved dose-control techniques demonstrates
higher in-spec power can be delivered with reduced overhead
2017 SW
Public
Slide 47
2017 SW
Slide 48
Public
Productivity targets for HVMSource contribution to productivity
Conversion
efficiency
Dose
margin
Drive laser
power
Laser to
droplet
control
Optical
transmission
Overhead
optimization
Exposure
dose
Stage accuracy
at high speed
Automation
Collector
maintenance
Droplet
generator
maintenance
Drive laser
reliability
>1500 WPD
in 2016
Productivity
Third generation Droplet Generators: average lifetime increased
Run time ~ 2700 hours
Long runtime and high reliability70% reduction in average maintenance time
2017 SW
Public
Slide 49
Performance
parameterGen III
Key Features Restart & Refill capable
Run-time~780 hrs (Ave)
(2700 hrs max)
Start-up yield >95%
Availability 95%
Droplet
diameter27 µm
Runtime of the Droplet Generator
5
10
20
0
500
1000
1500
2000
2500
3000
2016 Q2 2016 Q3 2016 Q4 2017 Q1 2017 Q2 2017 Q3
Serv
ice
tim
e in
hrs
(in
clu
de
s B
tim
e)
Ru
nti
me
s in
Ho
urs
BIII (Field Best)
BIII Average
Service Time(Btime Included)
2017 SW
Slide 50
Public
Hydrogen gas central to tin management strategy
Hydrogen performs well
for all these tasks!
Requirements for buffer gas:
➢ Stopping fast ions (with high
EUV transparency)
➢Heat transport
➢ Sn etching capability
High heat capacity High thermal conductivity
High EUV transparency
2017 SW
Slide 51
Public
Primary debris
Primary debris – directly from plasma and before
collision with any surface:
➢ Heat and momentum transfer into surrounding gas
o Kinetic energy and momentum of stopped ions
o Absorbed plasma radiation
➢ Sn flux onto collector
o Diffusion of stopped ions
o Sn vapor
o Sn micro-particles
Sn
Sn
Sn vapor (diffusion debris)
Sn particles
Sn+ Fast Sn ions (line of sight debris)
Main
Pre
Droplets
2017 SW
Slide 52
Public
3D measurement of fast tin ion distributions Faraday cups measure tin ion distributions
LASER
Ion measurements inform H2 flow requirements for source
Target direction
Laser direction
FC direction
Faraday cup
Red = moreBlue = less
2017 SW
Slide 53
Public
Tin ion distributions
Data are used for optimization of H2 flow in the source
2017 SW
Slide 54
Public
Microparticle debris from plasmaDark-field scattergraph imaging
dropletz (mm)
dropletx (mm
) and/or droplety (2 mm
)
2025
3035
242628303234363840
droplets
MP+PP
0
0.2
0.4
0.6
0.8
1
1.2
Fraction of pulses without microparticle debris
Research 3100
2017 SW
Slide 55
Public
Plasma-generated self-cleaning
Elemental hydrogen (H*) reacts with tin (Sn) to form Stannane (SnH4) which is gaseous and is pumped out of the vessel. Sn (s) + 4H (g) SnH4 (g)
2017 SW
Slide 56
Public
Collector Lifetime Continues to Improve>100 Gpulse to 50% EUVR
Collector reflectivity loss over time reduced to <0.4%/Gp
0 20 40 60 80 100 120 140 160 180 2000
10
20
30
40
50
60
70
80
90
100
110
Co
llecto
r R
efle
ctivity, %
No
rm.
Gigapulses
EUVL 2016
EUVL 2017
EUV Source Power Outlook
2017 SW
Public
Slide 57
2017 SW
Slide 58
Public
Collector protection secured up to 250 WCollector protection demonstrated on research tool
0
1
2
3
4
5
6
7
8
9
EU
V a
t IF
(m
J)
Main pulse peak power at plasma (MW)
2017 SW
Slide 59
Public
Research progress in EUV sourceDemonstrated EUV pulse energy of 7.5mJ
Short burst EUV
demonstration
on research
platform
400W research platform
Average of 800
sequential bursts
3400 EUV
source
discussed in this
presentation
375W in-burst at 50kHz
Slide 60
Public
Summary: EUV readiness for volume manufacturing15 NXE:3XY0B systems operational at customers
Significant progress in EUV power scaling for HVM
- Dose-controlled power of 250W
- EUV CE of 6%
CO2 development supports EUV power scaling
- Clean (spatial and temporal) amplification of short CO2 laser pulse
- High power seed system enables CO2 laser power scaling
Droplet Generator with improved lifetime and reliability
- >700 hour average runtime in the field
- >3X reduction of maintenance time
Path towards 400W EUV demonstrated in research
- CE is up to 6 %
- In-burst EUV power is up to 375W
2017 SW
Acknowledgements:Slide 61
Public
Alex Schafgans, Slava Rokitski, Michael Kats, Jayson Stewart
Andrew LaForge, Alex Ershov, Michael Purvis, Yezheng Tao, Mike Vargas, Jonathan Grava
Palash Das, Lukasz Urbanski, Rob Rafac Joshua Lukens, Chirag Rajyaguru
Georgiy Vaschenko, Mathew Abraham, David Brandt, Daniel Brown
Cymer LLC, an ASML company, 17075 Thornmint Ct. San Diego, CA 92127-2413, USA
Mark van de Kerkhof, Jan van Schoot, Rudy Peeters, Leon Levasier, Daniel Smith, Uwe Stamm, Sjoerd Lok, Arthur
Minnaert, Martijn van Noordenburg, Joerg Mallmann, David Ockwell, Henk Meijer, Judon Stoeldraijer, Christian
Wagner, Carmen Zoldesi, Eelco van Setten, Jo Finders, Koen de Peuter, Chris de Ruijter, Milos Popadic, Roger
Huang, Roderik van Es, Marcel Beckers, Niclas Mika, Hans Meiling, Jos Benschop, Vadim Banine and many others.
ASML Netherlands B.V., De Run 6501, 5504 DR Veldhoven, The Netherlands
2017 SW
Acknowledgements:Slide 62
Public
2017 SW