Development of EUV pellicle for reticle defect mitigation - Sematech

Development of EUV Pellicle for Reticle
Defect Mitigation
Yashesh A. Shroff, Michael Goldstein, Bryan Rice,
Pei-Yang Yan, Daniel Tanzil, Sang Lee, K. V. Ravi
October 18, 2006
EUVL Symposium, Barcelona
Intel Corporation

Outline
• Motivation
• Modeling results
– Pitch, Transmission, Uniformity impact
– Aerial image impact
• Experimental results
– At-wavelength measurements
– Full-size pellicle demonstration
– Capping layer study
– Pellicle optical metrology
• Field work results
– CDU analysis at DUV wavelength
• Future work and summary
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 2

Need for an EUV pellicle
•Pellicle-less EUV lithography development is a priority, with risk
– Swapping removable pellicles without generating particles
– Operating within the lithography tool without added defects
– Metrology confirmation of defect free reticles at sub-30nm
•The aim of this research is to create an EUV pellicle as a backup
to pellicle-less operation
•An EUV pellicle is expected to have trade-offs
– Moderate transmission loss
– Minimial uniformity loss
– Minimal contrast loss
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 3

Concept
Square cell mesh
Oct 18, 2006
Hexagonal cell mesh
Stand-off height = 6 mm
Mesh pitch
•
•
Mesh
Mesh
apodizes
apodizes
the
the
illuminator
illuminator
exit
exit
pupil
pupil
and
and
PO
PO
entrance
entrance
pupil.
pupil.
•
•
Mesh
Mesh
film
film
absorbs
absorbs
with
with
both
both
passes.
passes.
•The Concept
– Thin film mounted on a wire-mesh
– Mesh located “far” from reticle plane to defocus defects
– Transmission requires a high percentage open area
– Illumination uniformity requires partial coherence
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 4

Outline
• Motivation
• Modeling results
– Pitch, Transmission, Uniformity impact
– Aerial image impact
• Experimental results
– At-wavelength measurements
– Full-size pellicle demonstration
– Absorption uniformity
– Pellicle optical metrology
• Field work results
– CDU analysis at DUV wavelength
• Future work and summary
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 5

Numerical Model
1. Define mesh transmission:
2. Define illumination spots, integrate
mesh*spot transmission for each and
take the product.
3. Slide location of illumination spot
relative to the mesh to calculate nonuniformity.
1
T ( px,
py)
= ∫TCirc1 dxdy∫TCirc2
dxdy
2 4
π R
Max(
T ) − Min(
T )
ΔT
≡ ±
2 T
Oct 18, 2006
⎡
⎡ x ⎤ ⎤
T ( x,
y)
= UnitStep⎢P
FractionalPart
⎢ ⎥ − CD⎥
×
⎣
⎣ P ⎦ ⎦
⎡
⎡ y ⎤ ⎤⎧x
≥ 0
UnitStep⎢P
FractionalPart⎢
⎥ − CD⎥⎨
⎣
⎣ P ⎦ ⎦⎩
y ≥ 0
1
T = ∫∫T
( x,
y)
dxdy
A
⎡ ⎡σ
NA⎤⎤
R = h Tan⎢ArcSin⎢
⎥⎥
⎣ ⎣ Mag ⎦⎦
2
2
2
[ R − ( x − px)
− ( y ) ]
Circ = UnitStep − py
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 6

Far-field uniformity – Hexagonal mesh
Animation
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 7

0.5*range(T)/〈T〉
Uniformity
Illumination non-uniformity through pitch
(LW=5um)
Square mesh; LW=5um, sigma=0.75, NA=0.25, d=5mm, θ=6deg
2
1
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
50 100 150 200 250 300 350 400 450 500
Pitch (μm)
0
Oct 18, 2006
Square cell mesh
Pitch (μm)
• λ=13.5nm, σ=0.75, NA=0.25
• Mesh-Reticle gap, d: 5mm
• Illumination tilt, θ: 6o 0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Single pass Txm
0.5*range(T)/〈T〉 Uniformity
0
50 100 150 200 250 300 350 400 450 500
Pitch (μm)
0
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 8
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
Hexagonal cell mesh
Hex mesh; LW=5μm, sigma=0.75, NA=0.25, d=5mm, θ=6deg
Pitch (μm)
1 Range ( T)
NonUnifomi ty ≡
×
2 T
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Single pass Txm

Uniformity
3.6
3.2
2.8
2.4
1.6
1.2
0.8
0.4
Illumination non-uniformity through pitch
(LW=10um)
4
2
Square mesh; LW=10μm, sigma=0.75, NA=0.25, d=5mm, θ=6deg
0
50 100 150 200 250 300 350 400 450 500
Pitch (μm)
0
Oct 18, 2006
Square cell mesh
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Single pass Txm
Uniformity
Hex mesh; LW=10μm, sigma=0.75, NA=0.25, d=5mm, θ=6deg
4
1
0
50 100 150 200 250 300 350 400 450 500
Pitch (μm)
0
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 9
3.6
3.2
2.8
2.4
2
1.6
1.2
0.8
0.4
Hexagonal cell mesh
Pitch (μm) Pitch (μm)
Illumination non-uniformity is proportional to mesh line-width
Hexagonal cell based mesh chosen for higher structural rigidity
Selected: Hex cell pitch 250μm; target transmission 90%
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Present work
Single pass Txm

Aerial image modeling
Reticle
Pellicle
mesh
Oct 18, 2006
Entrance
Plane of PO
Illuminator
exit plane
• Mesh extracted from ‘near’ reticle
plane to source and PO
exit/entrance planes
• 3 cases explored: no mesh,
square, and hex mesh
• Mesh offset from reticle = 6mm
• Only PO entrance plane
apodization is considered here.
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 10

LW=10um, Pitch=250um: Hex & Sq comparison
(NA=0.25, sigma=0.75, d=5mm)
No mesh
Square mesh
Hex mesh
Dense pattern:
32/43nm L/S
No mesh Square
mesh
Hex
mesh
No mesh
Square mesh
Hex mesh
Iso pattern:
32/160nm L/S
No mesh Square
mesh
• Oct ~7% 18, 2006reduction
in contrast for iso and dense patterns for hex mesh.
• 7% / 13% reduction in NILS with dense / iso patterns for hex mesh
Hex
mesh
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 11

Outline
• Motivation
• Modeling results
– Pitch, Transmission, Uniformity impact
– Aerial image impact
• Experimental results
– At-wavelength measurements
– Full-size pellicle demonstration
– Capping layer study
– Pellicle optical metrology
• Field work results
– CDU analysis at DUV wavelength
• Future work and summary
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 12

Full-size pellicle
•Si membrane on 70LPI mesh
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 13

Transmission
Transmission characteristics for full-size mesh
• Near EUV • Across-field uniformity at focus
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
12 12.5 13 13.5 14 14.5 15
Oct 18, 2006
Wavelength (nm)
T=0.581
Scanner
WL
Transmission
0.4
-80 -60 -40 -20 0 20 40
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 14
0.6
0.58
0.56
0.54
0.52
0.5
0.48
0.46
0.44
0.42
X-Position (mm)
Pellicle transmission at 13.57nm scanner wavelength is about 58%,
double pass yields 33.6%.

Capping layer optimization: Transmission vs. stress
4 capping layers investigated: Si oxidation mitigation
0.6
0.5
0.4
Ti/Si/Ti: 20A/767A/20A
0.3
-5 0 5
Oct 18, 2006
22Å Ti
Line scan (at focus)
0.6
0.5
0.4
11Å Ru 22Å Ru 33Å Ru
Ru/Si/Ru: 11A/819A/11A
0.3
-5 0 5
0.6
0.5
0.4
Position(mm)
Ru/Si/Ru: 22A/810A/22A
0.3
-5 0 5
Line scan of pellicles with 300 x 50μ2m wide EUV beam at focus at ALS (LBL)
0.3
-5 0 5
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 15
0.6
0.5
0.4
Ru/Si/Ru: 33A/829A/33A
Stable Ru capping layer helps optimize membrane stress (tensile) at high transmission
Note: At-focus scan. Average transmission matters.

0.65
0.55
0.45
0.35
Capping layer study: Transmissivity near EUV
0.6
0.5
0.4
0.3
0.25
12 12.5 13 13.5 14 14.5 15
Wavelength (nm)
Oct 18, 2006
y
Ti/Si/Ti: 20A/767A/20A
Ru/Si/Ru: 11A/819A/11A
Ru/Si/Ru: 33A/829A/33A
Ru/Si/Ru: 22A/810A/22A
0.25
12 12.5 13 13.5 14 14.5 15
Wavelength (nm)
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 16
0.65
0.6
0.55
0.5
0.45
0.4
0.35
0.3
Si: 150nm
Si: 100nm
Mesh + membrane transmission @ scanner wavelength:
22A Ti cap: 55%
11A Ru cap: 57%
22A Ru cap: 52%
33A Ru cap: 59%
100nm Si: 58%
150nm Si: 51%

Mesh metrology: Si membrane study
3” sample
150nm Si on 80% transmissive mesh
Actinic measurement
Peak transmission = 53%
Elemental decomposition via XPS indicates
need for capping layer.
C
Oct 18, 2006
SiO
XPS depth profile
Si
O
At-wavelength (13.5nm)
in-focus scan
-2
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Ra=7.9nm
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 17
SiO
Ni
C
2
1.5
1
0.5
0
-0.5
-1
-1.5
150nm Si / 80% Mesh
0.5
0.45
0.4
0.35

T (%)
Full-scale pellicle optical metrology
Non-destructive pellicle membrane thickness measurement is
necessary to comply with high volume production requirements.
• An optical metrology model is developed and calibrated using (n,k).
• Rapid pellicle membrane thickness 2-D map generation capability.
Film thickness
Uniformity map
40
35
30
25
20
15
10
5
Thickness of membrane:
1164 Å
0
200 300 400 500 600 700 800 900 1000
Oct 18, 2006
Measured Transmittance
Calculated transmittance
(with Ni absorption and interference scattering)
Wavelength (nm)
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 18

0.7
0.6
0.5
0.4
0.3
0.2
0.1
Si deposition on Mo/Si ML
110nm Si on MoSi ML Reference MoSi ML
Extract 110nm Si txm
0
12.5 12.75 13 13.25 13.5 13.75 14 14.25 14.5
Oct 18, 2006
Si, Mo/Si ML: Reflectivity near EUV
wavelength (nm)
110nm Si/MoSi ML refl
Reference Mo/Si ML refl
0.5
12.5 12.75 13 13.25 13.5 13.75 14 14.25 14.5
Note: Calculated Si and SiO2 transmissions are based on CXRO data
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 19
0.9
0.85
0.8
18% discrepancy
0.75
0.7
0.65
0.6
0.55
Si: Txm near EUV
wavelength (nm)
Scanner wavelength
Extracted: 110nm Si (norm)
Calc: 110nm Si txm
Calc: 110nm Si+3nm Ox txm

Towards improving pellicle membrane
transmission
• Measured Si membrane
transmission is lower than
expected for crystalline Si with
surface oxidation.
• Need to alter deposition
conditions to improve intrinsic Si
transmission.
• Magnetron sputtered Si proven
for high transmission at EUV
wavelength. See chart
(Source: Eric Gullikson, ALS)
Oct 18, 2006
0.0
10 15 20 25 30 35 40 45 50
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 20
Transmission
1.0
0.8
0.6
0.4
0.2
80% txm at
13.5nm
CXRO Si filter
100 nm Si
100 nm Si + 6 nm SiO2
Wavelength (nm)

Outline
• Motivation
• Modeling results
– Pitch, Transmission, Uniformity impact
– Aerial image impact
• Experimental results
– At-wavelength measurements
– Full-size pellicle demonstration
– Capping layer study
– Pellicle optical metrology
• Field work results
– CDU analysis at DUV wavelength
• Future work and summary
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 21

DUV exposure with blank mesh: Full scan speed
Experiments run on an ASML XT 1400*
23x26mm field, 11x51 point sampling
Mesh transmission 91%, LW = 18μm, pitch = 300μm
Reticle transmission 90%, stand-off height = 5mm
NA = 0.93, σ = 0.75
Best dose for same CD is 10% higher for exposures with mesh pellicle. This is
consistent with a mesh transmittance of 91%.
Sensor at wafer-plane averages out mesh impact.
Blank reticle line scan
Oct 18, 2006
Sensor
0
-0.006 -0.004 -0.002 0 0.002 0.004 0.006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 22
1
0.8
0.6
0.4
0.2
-0.2
Reticle (no mesh)
Reticle (with mesh)
Difference
* Thanks to Nanda Samarakone and Jim Hunter of ASML for supporting this work.

CDU analysis – full field CD uniformity
50
-
49
3.1
28
5.3
27
-
69
-
51
3.8
48
2.5
29
2.1
26
2.1
8
-
70
Oct 18, 2006
-
68
3.4
52
2.6
47
2.7
30
1.9
25
1.9
9
2.3
7
-
71
3.5
67
5.2
53
2.4
46
2.2
31
2.6
24
2.5
10
2.2
6
2.7
72
3.0
66
4.4
54
2.7
45
2.7
32
1.9
23
1.8
11
2.4
5
2.6
VERTICAL
73
3.4
65
4.6
55
2.7
44
2.8
33
67.4
22
2.9
12
2.5
4
3.2
74
3.6
64
5.6
56
2.9
43
2.9
34
25.3
21
3.4
13
2.5
3
2.8
75
3.3
63
6.3
57
3.1
42
43.2
35
2.8
20
2.7
14
2.4
2
3.3
76
-
62
4.6
58
4.5
41
3.1
36
2.6
19
2.2
15
2.4
1
-
61
-
59
6.0
40
4.3
37
3.4
18
3.6
16
-
3σ Edge Fields Vert
Max 0.0
Ave −
Min 0.0
Spec ≤ 7
60
-
39
5.5
38
7.0
17
-
3s Full Fields Vert
Max 67.4
Ave 5.2
Min 1.8
Spec ≤ 4.2
VERTICAL
With pellicle Without pellicle
CD uniformity difference between wafers with and without
mesh is negligible – on average 0.2nm difference observed.
68
66
64
62
60
58
56
54
52
50
48
46
44
42
40
38
No
3σ Edge Fields Vert
Max 0.0
Ave −
Min 0.0
Spec ≤ 7
3s Full Fields Vert
Max 67.4
Ave 4.4
Min 1.7
Spec ≤ 4.2
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 23
50
-
49
3.4
28
6.0
27
-
69
-
51
4.6
48
2.4
29
1.8
26
1.9
8
-
70
-
68
3.5
52
2.9
47
2.4
30
1.7
25
2.0
9
1.9
7
-
71
3.5
67
4.9
53
2.8
46
2.0
31
2.2
24
2.7
10
1.9
6
2.7
72
3.1
66
4.0
54
2.2
45
2.3
32
1.9
23
1.9
11
2.5
5
2.7
73
3.2
65
4.1
55
2.8
44
2.6
33
67.4
22
2.3
12
2.4
4
2.6
74
3.9
64
4.8
56
2.5
43
2.2
34
24.6
21
3.0
13
2.2
3
3.0
75
3.8
63
5.7
57
3.0
42
2.0
35
2.0
20
2.3
14
2.4
2
3.3
76
-
62
4.8
58
4.5
41
3.0
36
2.7
19
2.2
15
2.5
1
-
61
-
59
6.2
40
4.3
37
3.5
18
3.6
16
-
60
-
39
4.8
38
6.4
17
-
69
67
65
63
61
59
57
55
53
51
49
47
45
43
41
39
No

Summary
• A new model of illumination uniformity for pellicles has been created. Results
are consistent with experimental data.
– Mechanical stability in mesh is an important factor – low stress and high structural
rigidity with decreasing line-width are needed.
– Modeling results indicate linear dependence of LW to illumination uniformity.
Hexagonal mesh with 250μm pitch chosen. Current line-width of 18μm will ultimately
drive to 5μm.
• Full-scale pellicle created and at-wavelength transmission measured.
– 100% yield in membranes. Process is reliable.
– Single-pass EUV transmission at 58%
• Capping layer optimization shows 3.3nm Ru can improve membrane stress
without compromising EUV transmission.
• Successful field run with ASML’s XT1400 tool by mounting a blank mesh on
patterning reticle. Negligible impact on CDU, process latitude at DUV.
• Created a model to measure pellicle uniformity using non-destructive optical
techniques.
• Future work:
– Quantify reticle defect mitigation using EUV pellicles
– Improve single-pass transmission to 84% with implementation of suitable capping
layer solution with improvement in sputter deposition process, reduction in film
thickness, and higher mesh transmission.
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 24

Thanks to…
•Eric Gullikson at the Advanced Light Source EUV beamline (LBL)
for at-wavelength transmission measurements.
•ASML (particularly, Nanda Samarakone and Jim Hunter) for
supporting mesh CDU imaging study at DUV with XT1400.
•Brian Coombs (Intel) for ECD measurements of wafers.
•MATTEC (Materials Technology) at Intel for AFM and elemental
depth profiling of membranes via XPS.
•Ravi Mullapudi at Tango Systems (San Jose, CA) for helping
develop reliable membranes on challenging substrates.
•Committed support from senior management at Intel for funding
and providing resources for the project.
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 25

Backup
Oct 18, 2006
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 26

Si / Ru transmission
Si thickness (nm)
Oct 18, 2006
150
100
50
0
0 1 2
Ru thickness (nm)
3 4
DoE: Cap layer optimizations
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 27
0.95
0.9
0.85
0.8
0.75
0.7

XPS metrology: Ti/Si/Ti profile
Oct 18, 2006
C
Capping layer protects Si from oxidation, improving transmission
Marginal cap layer oxidation observed
Si
O
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 28
O
Si-O
C

CDU analysis for mesh impact at DUV
V-H Bias (nm)
2.0
1.5
1.0
0.5
0.0
-0.5
Process window curves with and without pellicle are
comparable.
Best dose for same CD is 10% higher for exposures with mesh
pellicle. This is consistent with a mesh transmittance of 91%.
Oct 18, 2006
H-V bias for 100nm Iso lines
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Cross-slit position by column
H-V +scan without pellicle
H-V +scan with pellicle
Y. Shroff, et. al., EUV Pellicle Development ~ EUVL Symosium, 2006 ~ 29