pellicle and thin film short-wavelength metrology...
TRANSCRIPT
© RWTH Aachen
Serhiy Danylyuk
RWTH Aachen University
21.02.2016, San Jose, CA
PELLICLE AND THIN FILM SHORT-WAVELENGTHMETROLOGY
© RWTH Aachen
Pellicle requirements
Page 2
*C. Zoldesi, et al., “Progress on EUV pellicle development,” Proc. SPIE 9048 (2014)
Pellicle requirement HVM Target
EUV transmission 90% single pass
Spatial non-uniformity of the transmission <0.2%
Angular non-uniformity of the transmission <300 mrad
Dynamic heat load 5.6 W/cm2 (250 W IF)
© RWTH Aachen
Pellicle illumination conditions NXE:3300
Page 3
NAmask = 0.0825 illumination angles (Y-axis) are between 1.27° and 10.73°
© RWTH Aachen
Pellicle illumination conditions NXE: High NA
Page 4
illumination angles in X direction are between -7.9(-6.6°) and 7.9°(6.6°)illumination angles in Y direction are between 2.1°(1.8°) and 9.9°(10.2°)
NAwafer = 0.55 NAmask_x ~ 0. 1375 (0. 115), NAmask_y ~ 0.068(0.073)
MAGx= 4(4.8)MAGy= 8(7.5)
© RWTH Aachen
Ray Tracing Analysis
Page 5
Defects on the pellicle can lead to a non-uniform transmission through the pellicleat different reticle coordinates
The target specification states a transmission of 90% with <0.2% uniformity
Via ray tracing, the influence on theillumination profile of differently sizeddefects is investigated
NA = 0.33/4
pellicle
reticleprojectionoptics
defect
2.5 mm
reti
cle
coo
rdin
ate
pellicle
defect
© RWTH Aachen
Influence of defect size
Page 6
Different types of defectsare investigated
Circular absorber defects
100% absorption
50% absorption
circular hole defects(100% transmissioninstead of 90% pellicletransmission)
The defect size (diameter ofcircular defect) is varied
© RWTH Aachen
Effective pellicle pixel model
0 20 40 60 80 10088
89
90
91
Tra
nsm
issio
n [
%]
Defect diameter [µm]
100% absorption defect
50% absorption defect
10% transmission defect
±0.2% zone
Page 7
𝑇 = 𝑇𝑝𝑒𝑙𝑙𝑖𝑐𝑙𝑒 − 𝑇𝑑𝑒𝑓𝑒𝑐𝑡 ×𝐴𝑟𝑒𝑎𝑑𝑒𝑓𝑒𝑐𝑡
𝐴𝑟𝑒𝑎𝑝𝑒𝑙𝑙𝑖𝑐𝑙𝑒_𝑝𝑖𝑥𝑒𝑙
𝐴𝑟𝑒𝑎𝑝𝑒𝑙𝑙𝑖𝑐𝑙𝑒_𝑝𝑖𝑥𝑒𝑙 ≈ 𝑑 × (tan 𝜃𝑚𝑎𝑥 − 𝜃𝑚𝑖𝑛)
NXE 3300 : 420-450µm (2.5 mm stand-off)
NXE High NA: X-direction (~700(600)µm)Y-direction (~350 (375)µm)
For off-axis illumination pellicle pixel has to be multiplied with pupil fill factor (PFR), this can reduce the allowed defect size to sub-10µm
In the worst case (100% absorbing) defects with <20µm diameter size is allowed per effective pellicle pixel (350-700µm)
For pellicle transmission uniformity metrology even for aggressive dipoles there is no need to resolve pellicle uniformity better than 50µm
© RWTH Aachen
Transmission uniformity measurements
Page 8
Intensity after pellicle
Uniformity noise floor s=0.13% for the choosen data stack (10 pictures)
4 5 6 7 8 925
30
35
40
45
50
Tra
nsm
issio
n[%
]
Position [mm]
0.2% band
3,80 3,82 3,84 3,86 3,88 3,90
0,0
0,1
0,2
0,3
0,4
0,5
Tra
nsm
issio
n
Position [mm]
Knife-edge resolution 48µm
Membrane is fixed at 6 mm distance to the EUV-sensitive CCD with 13µm pixels
© RWTH Aachen
Pellicle OoB reflectivity
*Park et al, Proc. SPIE 9048, Extreme Ultraviolet (EUV) Lithography V, 90480S (2014)
© RWTH Aachen
Pellicle OoB reflectivity
50 100 150 200 250 300 350 400
0
10
20
30
40
50
60
Pelli
cle
OoB
reflectivity [%
]
Wavelength [nm]
a-C (25 nm)
B4C (4nm)/a-C(25 nm)/B
4C (4nm)
IMD5 model
50 100 150 200 250 300 350 400
0
10
20
30
40
50
60
Pe
llicle
Oo
B r
eflectivity [
%]
Wavelength [nm]
p-Si (65 nm)
a-C (25 nm)
IMD5 model
50 100 150 200 250 300 350 4000
10
20
30
40
50
60
Pe
llicle
Oo
B r
eflectivity [
%]
Wavelength [nm]
B4C capped p-Si
B4C capped a-C
IMD5 model
50 100 150 200 250 300 350 4000,0
0,1
0,2
0,3
0,4
0,5
0,6
IMD5 model
Re
fle
ctivity
Wavelength [nm]
p-Si (25 nm)
p-Si (60 nm)
n,k data: a-C (CXRO+Tarrio et al (1998)+Windt et al (1988)+Palik (1991))p-Si (CXRO+Palik (1985))
© RWTH Aachen
Pellicle OoB reflectivity
50 100 150 200 250 300 350 4000
10
20
30
40
50
60
Pelli
cle
OoB
reflectivity [%
]
Wavelength [nm]
uncapped p-Si
Si3N4 capped p-Si
Mo capped p-Si
B4C capped p-Si
4nm double side caps
IMD5 model
• OoB reflectivity of a pellicle can exceed 50%
• This may have influence on the performance of the scanner
• Capping layers can influence the effect strongly
• OoB reflectivity characterization is necessary
Currently there are no tools for the region between 20 nm and 200 nm
no “fresh” n,k data for modelling either
OoB specification?
© RWTH Aachen
EUV Reflectometer @ Aachen
EUV Source Sample chamber
1st grating 2nd grating
2nd CCD
grating
sample CCDgrating
referenceslit-
aperture
multi angle system (1°-15°)wavelength range 9-17 nm
mirrorAngular dependencies allow to extract refractive index data at each wavelength
© RWTH Aachen
Multi-angle measurement results
Excellent sensitivity to thickness and composition
Layered systems investigation
Page 13
9 nm target thickness of HfO2
HfO2.075 (d=10 nm, ρ=8.43 g/cm3)S iO2(d=0.26 nm)
For complex stacks independent thickness measurement (XRR, ellipsometry, TEM) may be required to decrease ambiguity of the model
© RWTH Aachen
Near-edge EUV Reflectometry
Page 14
Absorption edges of elements in working wavelength range
Element Binding Energy (eV)
Corresponding Wavelength (nm)
Electron state
4 Be 111,5 11,12 K 1s
12 Mg 88,7 13,98 L1 2s
14 Si 99,82 12,42 L2 2p1/2
15 P 136 9,12 L2 2p1/2
37 Rb 113 10,97 M4 3d3/2
38 Sr 136 9,12 M4 3d3/2
57 La 105,3 11,77 N4 4d3/2
59 Pr 115,1 10,77 N4 4d3/2
60 Nd 120,5 10,29 N4 4d3/2
61 Pm 120 10,33 N4 4d3/2
62 Sm 129 9,61 N4 4d3/2
63 Eu 133 9,32 N4 4d3/2
89 Ac 80 15,50 O4 5d3/2
90 Th 92,5 13,40 O4 5d3/2
91 Pa 94 13,19 O4 5d3/2
92 U 102,8 12,06 O4 5d3/2
Grazing incidence angle 5°
Direct elemental sensitivity
Less ambiguous modelling
Near-edge fine structure
© RWTH Aachen
Near-edge EUV Reflectometry
Page 15
Near-Edge X-Ray Absorption Fine Structure (NEXAFS)
[*] M. Banyay, Surface and Thin Film Analysis by Spectroscopic Reflectometry with EUV Emitting Laboratory Sources, PhD (RWTH Aachen, 2011).[**] E. O. Filatova, A. A. Sokolov, E. Yu. Taracheva, I. V. Bagrov, Technical Physics Letters, Vol. 35, Issue 1, Jan 2009, pp. 70-72.[***] E. O. Filatova, A. S. Shulakov, and V. A. Luk’yanov, Phys. Solid State 40 (7), July 1998.
Determination of chemical bonds and local symmetry Quick qualitative interpretation (fingerprinting)
© RWTH Aachen
SUMMARY
Pellicle transmission spec resolution should not be below 50µm
Metrology with this reolution is not a challenge
OoB pellicle reflectivity may become a problem and should be controlled
New reliable n,k data are necessary to model the performance, preferable measured directly on a pellicle
Hard to measure in VUV wavelength region, no reflectometers orellipsometers in the range between 20 and 200 nm
Reflectometry (and n,k) measurements near the absorption edges (e.g. near Si L-edge @12.4nm) are often NEXAFS-modified and difficult tomodel precisely