euv lithography industrialization...
Post on 12-Jul-2020
3 Views
Preview:
TRANSCRIPT
Rudy Peeters EUVL Washington DC | Oct 27th, 2014
EUV Lithography Industrialization Progress
• EUV benefit and status at customers
• Towards higher productivity
• Imaging, Overlay, Defectivity
• Summary
27 Oct 2014
Public
Slide 2
Agenda
• EUV benefit and status at customers
• Towards higher productivity
• Imaging, Overlay, Defectivity
• Summary
Agenda
27 Oct 2014
Public
Slide 3
27 Oct 2014
Slide 4
Public
Multi-patterning complexity drives need for EUV
20 n
m (
Imm
ers
ion)
8xMasks
10 n
m (
Imm
ers
ion)
23xMasks
7 n
m (
Imm
ers
ion)
33xMasks
7nm
(E
UV
)
9xMasks
Node
# of litho steps
# OVL Metrology
# max metrology /
litho steps
20 nm
8 (Immersion)
9-11
2
10 nm
23 (Immersion)
36-40
3
7 nm
33 (Immersion)
59-65
4
7 nm
12
3
Layers
9 (EUV)
EUV Lithography will stop the strong increase of litho steps/masks needed
EUV status:
Demonstrated >500 wafers per day at customer sites Slide 5 Public
27 Oct 2014
• More than 500 wafers per day demonstrated
during endurance tests at 2 customer sites
• 7 NXE:3300B systems shipped to customers
• 4 more NXE:3300B systems being integrated
• 4th generation NXE system (NXE:3350B)
integration ongoing
• EUV cleanroom extension is under construction
Slide 6
Public
• EUV benefit and status at customers
• Towards higher productivity
• Imaging, Overlay, Defectivity
• Summary
Agenda 27 Oct 2014
Total produced wafers on NXE:3300B at customer sites Reaching output in 8 months that took NXE:3100 more than 2 years
NX
E:3
10
0
NX
E:3
30
0B
Week since start
27 Oct 2014
Public
Slide 7
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
200001 4 7
10
13
16
19
22
25
28
31
34
37
40
43
46
49
52
55
58
61
64
67
70
73
76
79
82
85
88
91
94
97
10
0
10
3
10
6
10
9
NX
E:3
10
0
NX
E:3
300B
To
tal n
um
ber
of
pro
du
ced
wafe
rs
Wafers per day program
27 Oct 2014
Public
Slide 8
Conversion
efficiency
Dose
margin
Drive laser
power
Optical
transmission
Automation
Collector
lifetime
Droplet
generator
reliability
Drive laser
reliability
Overhead
optimization
Exposure dose
> 500 Wafers
per day
Stage accuracy at
high speed
Laser to
droplet
control
>40W rolled-out on customer NXE:3300B systems
NXE:3300B customer systems
27 Oct 2014
Public
Slide 9
Under clean collector conditions
0
10
20
30
40
50
60
70
80
Customer Customer Customer Customer
EUV
po
we
r (W
atts
)
Courtesy of
System continuously used Stable power over 2 months period of time So
urc
e p
ow
er [
Wat
t]
Acc
um
ula
ted
Gp
uls
e co
un
t
Slide 10
Public
27 Oct 2014
Power of >40W
since start use system System pulse count
Customer system continuous use at power level of >40W
Courtesy of
>40W over 2 months period of time C
olle
cto
r re
flec
tivi
ty [
%]
Slide 11
Public
27 Oct 2014
0
10
20
30
40
50
60
70
80
90
100
No collector degradation seen in 2 months period of time during
continuous use of the system at >40 Watts
Stable collector reflectivity
Stable power over 2 months period of time
Sou
rce
po
wer
[W
att]
Power of >40W
since start use system
Stable ~80W EUV source power at ASML Multiple runs on NXE:3300B stand-alone sources
Meets dose requirement Conversion efficiency >3%
See presentation David Brandt
Slide 12
Public
27 Oct 2014
0
10
20
30
40
50
0
20
40
60
80
0 3 6 9 12 15 18 21 24
Do
se m
argi
n (
%)
Po
we
r (W
)
Time (hours)
Continuous stable source operation at 80 W excellent die yield, ~25% dose margin
Slide 13
Public
27 Oct 2014
24 hours continuous operation at 80W of EUV power with excellent
simulated Die Yield
Die Yield = Simulated % of dies that meet the 1% dose requirement
Wafers per day program: Progress in major areas
27 Oct 2014
Public
Slide 14
Conversion
efficiency
Dose
margin
Drive laser
power
Optical
transmission
Automation
Collector
lifetime
Droplet
generator
reliability
Drive laser
reliability
Overhead
optimization
Exposure
dose
> 500 Wafers
per day
Stage accuracy
at high speed
Laser to
droplet
control
Improvement of Conversion
Efficiency demonstrated
Reduced dose margin with
advanced controls Improvement in collector
lifetime
Doubled source power to ~80
Watts demonstrated
Improved automation
algorithms
Faster resist formulations
demonstrated
See presentation David Brandt
Improved optical design
• EUV benefit and status at customers
• Towards higher productivity
• Imaging, Overlay, Defectivity
• Summary
27 Oct 2014
Public
Slide 15
Agenda
4 nm99.7%Fx: 1.1 nmy: 1.0 nm
(S2F)
Ed
ge
(>
0.1
47
m)
an
d O
utlie
rs (
6sig
S2
F)
rem
ove
d
# of wafers D
CO
[nm
]
Slide 16 Public
27 Oct 2014
Full wafer dedicated chuck overlay of <1.1 nm at
40W source power
• Layers are exposed on the same wafer on the same chuck. • Dedicated chuck overlay is a measure of the overlay performance of the system.
Overlay matching NXE:3300B to Baseliner MMO wafers Full-wafer matched machine overlay <4 nm at 40W, stable over 2 weeks
Wafer – Baseliner controlled
10 nm
99.7%x: 3.3 nmy: 3.0 nm
Customer B: 6par/field; field fingerprint and flyers removed
Lot (3.3, 3.0)
MM
O [n
m]
10 nm
99.7%x: 3.2 nmy: 3.4 nm
6 wafer BMMO S2(F-WEC), 10par/chuck res, W2W outliers removed
Lot (3.2, 3.4)
MM
O [n
m]
1 2 3 4 5 6 0
2.5
5
7.5
10
Aug 4 Aug 8 Aug 17 0
2
4
6
8
27 Oct 2014
Public
Slide 17
Customer A
Wafer – one day after setup
NXE:3300B full wafer Focus Uniformity at >40 W Well below 25 nm performance target
0
5
10
15
20
25
1 2 3 4
Chuck 1 Chuck 2
Fo
cu
s U
nifo
rmity [n
m]
System A B C D
27 Oct 2014
Public
Slide 18
• The focus uniformity performance test is done in resist • Focus sensitive marks are being exposed to obtain full wafer coverage information
Multiple NXE:3300B systems show CDU below 1.5 nm dense and iso lines exposed at 16mJ/cm2 – conventional ill.
22 nm DLs: Full wafer CDU <1.2 nm
22 nm ILs: Full wafer CDU < 1.5 nm
System A System B System C System D System E
CD measurements performed with YieldStar Data corrected for reticle errors and mask shadowing effect
27 Oct 2014
Public
Slide 19
Good Focus and Dose performance stability measurements done over 2 months period at 40W source power
27 Oct 2014
Public
Slide 20
Free running state in preparation for implementation of baseliner control
0
0.25
0.5
0.75
1
1.25
1.5
1.75
2
2.25
2.5
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4
8/20/2014 8/22/2014 8/29/2014 9/7/2014 9/13/2014 9/16/2014
Target
Best Focus stability of 5 nm Max Dose Error < 1% for 99.96%
of exposed fields (2239 of 2240 fields)
Ma
x D
ose
pe
r fie
ld E
rro
r (%
)
25
20
15
10
5
0
-5
-10
-15
-20
-25
7/28/14 8/4/14 8/11/14 8/18/14 8/25/14 9/1/14 9/8/14 9/15/14
Be
st F
ocu
s [n
m]
◆ Z_ch1
■ Z_ch2
Courtesy of Courtesy of
27 Oct 2014
Public
Slide 21
NXE:3300B CD monitoring for 18 nm HP Horizontal lines
at 40W shows stable performance over time
19.5
CD
[n
m]
19.0
18.5
18.0
17.5
17.0
16.5
16.0
15.5
8/20
2.20
Fu
ll Wa
fer C
DU
[nm
]
2.00
1.80
1.60
1.40
1.20
1.00
0.80
0.60 1 1 1 2 3 4
8/22 8/29/14
1 2 3 4 5 6 7 8
9/7/14
1 2 3 4 5 6
9/13/14
1 2 3 4
9/16/14
Wafer - Date
Courtesy of
Achieved >50% exposure dose reduction while keeping
same performance for tip-to-tip and tip-to-space
Tip-to-tip
~30 nm
~59 mJ/cm2 ~24 mJ/cm2
2013 2014 10
15
20
25
30
35
40
45
50
55
60
0 10 20 30 40 50 60
T2
T g
ap
CD
[n
m]
Dose [mJ/cm2]
Tip-to-space
22 nm dense L/S with 20 nm gap
~20 nm
See presentation Eelco van Setten for EUV imaging
~50 mJ/cm2 ~18 mJ/cm2
2013 2014 10
15
20
25
30
35
40
45
50
55
60
0 10 20 30 40 50 60
T2
T g
ap
CD
[n
m]
Dose [mJ/cm2] Conv - Resist C
Quasar - Resist C
Conv - Resist A
Quasar - Resist A
27 Oct 2014
Public
Slide 22
22 nm dense L/S with 20 nm gap
NXE:3300B 10 nm logic imaging at large process window Validation of Tachyon NXE mask 3D model accuracy
-60nm
0nm
60nm
Focus Predicted contours (blue)
match well SEM contours
NXE:3300B, 10 nm logic metal 1 layer example, 45 nm minimum pitch,
Quasar 45, ~20mJ/cm2 Source: ST, 2014
27 Oct 2014
Public
Slide 23
18 nm Dense Lines process window improvement
with FlexPupil optimization
0
2
4
6
8
10
12
14
16
18
20
22
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
EL
[%
]
Depth of Focus [um]
SEC's 18nm DL - Unconstrained (ellipse) PW
Quasar A
Quasar B
LWR = 3.99nm
LWR = 2.95nm
27 Oct 2014
Public
Slide 24
Customer C
Source Mask Optimization and FlexPupil maximizes process
window and minimizes edge placement error (EPE) 27 Oct 2014
Public
Slide 25
• Pattern shift (impacting overlay) due to EUV non-telecentricity and 3D mask topography is reduced by optimizing for EPE
• Source mask optimization (SMO) optimizes FlexPupil shape to maximize process window and minimize EPE
¹Source: Patterns published by Yan Borodovsky of Intel (2012 International Workshop on EUV Lithography Maui, HI)
27 Oct 2014
Slide 26
Public
Front-side reticle defectivity: 10x reduction/year realized Clean scanner remains priority
Ad
de
d p
art
icle
s >
92 n
m
per
reticle
pass
Customer requirement
for full production
without pellicle
2010 2011 2012 2013 2014 2015
0.0001
0.001
0.01
0.1
1
10
100
■ NXE:3100
■ NXE:3300B
Overall PRP <0.004 for system based on >1300 test cycles
Scalability of free-standing pSi films demonstrated and
prototype pSi pellicles successfully tested in EUV tools
11mm x 11mm (~80% EUV transmission)
50mm x 50mm (~82% EUV transmission)
64mm x 106mm ~(86% EUV transmission)
103mm x 122mm (~85% EUV transmission)
Half size prototype
Full size prototype
Open frames used for testing prototype pSi pellicles in EUV tools
27 Oct 2014
Public
Slide 27
All reticles and frames courtesy of Intel
Imaging testing of ½ size pSi pellicle collaborative effort of
ASML and customers 27 Oct 2014
Slide 28
Public
Wilton
Taiwan
½ size pSi film Frame mounted
on reticle
pSi film +
frame
mounted on
reticle
Frame mounted
on reticle
27 Oct 2014
Slide 29
Public
27 nm L/S features successfully exposed with
half-size proto pellicle
SEM image at BE/BF
w/o pellicle
SEM image at BE/BF
w/ pellicle
0.2 µm 0.2 µm
Joint effort of NXE end users and ASML accomplished
imaging results of half size pellicle
CD = 24.2 nm
LWR = 4.0 nm
CD = 24.0 nm
LWR = 3.7 nm
• NXE:3100, NA=0.25, Conventional Illumination
• CDU difference of 0.18 nm
• Pellicle EUV transmission confirmed in imaging data: 85.5% (single pass)
Reticles and frames courtesy of Intel Scanner exposures courtesy of tsmc
• EUV benefit and status at customers
• Towards higher productivity
• Imaging, Overlay, Defectivity
• Summary
27 Oct 2014
Public
Slide 30
Agenda
• >40W stable performance at customers, 80W performance
shown at ASML and being transferred
to customers
Summary: EUV towards production insertion
• Two customers demonstrated 500 wafer
per day capability in endurance tests.
• System performance well within target at
>40W source power
• NXE:3300B imaging and overlay performance is in line with
requirements for 10nm node insertion
• EUV source: Improvements demonstrated in conversion
efficiency, dose margin, automation
and collector lifetime, driving power and availability.
27 Oct 2014
Public
Slide 31
• Defect reduction ~10x per year shown and full-size EUV
pellicle prototype manufactured
Acknowledgements
The work presented today, is the result of hard work and
dedication of teams at ASML, Cymer, Zeiss, and many
technology partners worldwide including our customers
Special thanks to our partners and customers for allowing
us to use some of their data in this presentation
27 Oct 2014
Public
Slide 32
top related