Lens Engineering Development Department
Core Technology Center
EUV scattering from
Mo/Si multilayer coated
Mirrors
Kazutami Misaki and Noriaki Kandaka
Nikon Corporation
E-mail: [email protected]
International EUVL Symposium
18-20 October 2010, Kobe JAPAN
Abstract
Mo/Si multilayer coated mirrors are employed in EUVL as main optical element.
Scattering due to surface and interface imperfections crucially affects the
throughput and image contrast of EUV optics. Unlike the scattering from a single
surface, scattering from multilayer mirrors is very complicated. Specifically, we
have to consider the effect of interfere between each layer (i.e., optical path
difference of the scattered light). We have performed measurements of EUV
scattering from Mo/Si multilayer mirrors with various conditions: substrates,
coatings, wavelengths, and configurations. In particular, we measured the angleresolved scattering not only in the “reflecting plane” which includes incident and
reflected beam, but also in the direction perpendicular to the “reflecting plane”. As
a result, they were distinctly different. In other words, 2-dimensional scattering
distribution was examined and found to be anisotropic around the specular peak,
as expected by the effect of multilayer interfere. As approaching normal incidence
this discrepancy should be small, however, it is not negligible even at the incident
angle of 10 degree. EUV reflectivity and the surface power spectral density are
also examined in conjunction with the scattering distribution. The relationship of
these properties and the performance of Mo/Si mirrors are discussed in detail.
EUVL Symposium, 18-20 October 2010
p.2
Scattering from a single surface
Born approximation for scattering from a single surface
2
π
1 dI = 16 R ･ PSD(f)
I0 dΩ λ 4
・・・(1)
λ: wavelength
R: Reflectivity
I0: incident Intensity
PSD: Power Spectral Density of a surface
agrees with the measured tail in the case of
small roughness, near normal incidence and
small scattering angle.
E. Gullikson, Proc. SPIE 3331, pp. 72-80
EUVL Symposium, 18-20 October 2010
p.3
Scattering from multilayer
ref
lec
t
t
ligh
nt
ide
inc
ed
lig
ht
We have to take into account the effect of interfere
between each layer (i.e., optical path difference of
the scattered light).
θ
scattered light : every direction
i.e., various state (phase shift)…
phase coherent ⇒ high scatter
phase lag increase ⇒ attenuated
N. Kandaka, et al., EUVL symposium 2004
EUVL Symposium, 18-20 October 2010
p.4
Measured samples
4 samples with
different level of
smoothness:
mid-level
sample C
1.E+05
1.E+04
PSD_ave[nm^4]
rough
sample A
sample B
PSD after ML depo.
measured by AFM (in house)
smooth
sample D
(ML deposited
on Si-wafer)
1.E+03
1.E+02
1.E+01
1.E+00
1.E-01
1.E-02
1.E-04
1.E-03
1.E-02
1.E-01
Hz[/nm]
EUVL Symposium, 18-20 October 2010
p.5
Effect of interfere between each layer
Scattering
1.E+00
tail deduced (calculated) from PSDs
−
−
−
−
intensity [a.u.]
1.E-01
1.E-02
sample A
sample B
sample C
sample D
1.E-03
a single surface: Eq. (1)
1.E-04
1.E-05
effect of multilayer
1.E-06
1.E-07
0
5
10
15
scattering angle [deg]
EUVL Symposium, 18-20 October 2010
※ clear difference
20 at large angle
p.6
Measurement facility: PF (KEK) BL-12A
KEK Photon Factory (PF)
Synchrotron facility : BL-12A
strong intensity
Detector with variable applied voltage
dynamic range: 6 order of magnitude
Courtesy of KEK
suitable to study scattering
monochromator
BL-12A
measurement chamber
EUVL Symposium, 18-20 October 2010
p.7
intensity [a.u.]
A (cal.)
A (mes.)
intensity [a.u.]
Measurements and calculations
5
10
15
20 0
B
(cal.)
scattering angle [deg]
B (mes.)
0
5
10
15
scattering angle [deg]
intensity [a.u.]
intensity [a.u.]
0
EUVL Symposium, 18-20 October 2010
20 0
C (cal.)
C (mes.)
5
10
15
20
(cal.)
scattering angle D
[deg]
D (mes.)
5
10
15
scattering angle [deg]
20
in good agreement
p.8
Scattering tail with non-peak wavelength
measured results
peak -3Å
129A (-3A)
132A (peak)
134A (+2A)
Reflectivity
intensity
peak +2Å
peak
126 128 130 132 134 136
wavelength [Å]
-15
-10 -5
0
5
10
scattering angle [deg]
15
multilayer effect clearly observed
EUVL Symposium, 18-20 October 2010
p.9
2-D distribution of scattered light
reflected beam
specular peak
incident
beam
φ=
0
Y
θ
θ’
φ : azimuthal angle
φ
φ=
scattered 180
θ light
X
Generally, scattering tail is
examined in the “reflecting
plane” (X-Y plane, φ0 & φ180).
What about other directions ?
Z
Considering optical path difference, the scattered light
shows anisotropic distribution around the specular
peak (i.e., rotational asymmetry).
EUVL Symposium, 18-20 October 2010
p.10
Calculated 2-D distribution
relative attenuation coeff.
calculation: phase difference ⇒ attenuation
1.2
anisotropic
0
1.0
30
0.8
60
0.6
90
0.4
120
0.2
0.0
in the
reflecting
plane
150
incident angle = 10°
0°
10° 15°
5°
scattering angle
180
20°
φ : azimuth
For more rigorous treatment and another approx. models…
e.g., “D. G. Stearns et al. 1998”, “S.Schroeder et al. 2010”
p.11
EUVL Symposium, 18-20 October 2010
Calculation: various incident angle
0
relative attenuation coeff.
30
60
15°
°
0°
°
90
120
150
180
20°
°
5°
°
0°
10°
°
0°
10°
10°
20°
As approaching normal incidence
the discrepancy decreases,
however, it is not negligible even
20° at the incident angle of 10 degree.
EUVL Symposium, 18-20 October 2010
p.12
2-D scattering measurement configuration
（in the “reflecting plane”）
）
（orthogonal direction）
）
angle-scan
c
in
detector
rot. axis
sample
rot. axis
m
a
e
b
t
n
e
id
“normal”
detector
rot. axis
am
be
nt
de
ci
in
sample
rot. axis
angle-scan
sample set
inclined
detector
EUVL Symposium, 18-20 October 2010
p.13
Measured 2-D distribution
Φ 90 & 270
Φ 0 & 180
Φ 90 & 270
Φ 0 & 180
Φ 90 & 270
Φ 0 & 180
Φ 90 & 270
Φ 0 & 180
-20
-15
-10 -5
0
5
10
scattering angle [deg]
EUVL Symposium, 18-20 October 2010
sample A
sample B
sample C
sample D
obviously different
depending on the
azimuthal angle, φ
15 φ 0 & 180 : asymmetry
φ 90 & 270 : symmetry
p.14
Comparing measurements and calculation
Based on “φ 0” profile and calculated azimuthal dependence,
we estimated profiles in other directions.
Φ 90 & 270 measured
Φ 0 & 180 measured
Φ180 estimated
Φ90 estimated
Φ270 estimated
-15
-10 -5
0
5
10
scattering angle [deg]
EUVL Symposium, 18-20 October 2010
some discrepancy
accorded at large
scattering angle
validated the concept
of multilayer effect in
15
scattering
p.15
Scattering loss estimation
8%
7%
6%
5%
4%
3%
2%
1%
0%
deduced from measured tail
A
B
C
D
0
5
10
scattering loss
scattering loss
estimated from PSDs
15
scattering angle [deg]
sample A ： 6.6%
sample B ： 6.6%
sample C ： 1.4% well-agree !
sample D ： 0.3%
EUVL Symposium, 18-20 October 2010
8%
7%
6%
5%
4%
3%
2%
1%
0%
0
5
10
15
scattering angle [deg]
sample A ： 7.8% (~8%)
sample B ： 5.7% (~6%)
sample C ： 1.9% (~2%)
sample D ： 0.4% (~0.5%)
p.16
EUV reflectivity
measured EUV reflectivity
A
B
C
D
70
60
reflectivity [%]
refer to “sample D”,
50
A ∆ R = -6.5% -6.3% -7.4%
B ∆ R = -5.6% -6.3% -5.3%
C ∆ R = -1.0% -1.1% -1.5%
estimated from PSDs
40
deduced from
scattering
measurements
30
20
10
0
12 14 16 18 20 22 24
incident angle [deg]
EUVL Symposium, 18-20 October 2010
◎ good agreement
by ~1% accuracy !
p.17
Summary
Considering scattering from multilayer mirrors,
we have to take into account the effect of
interfere between each layer.
2-dimensional scattering distribution was
experimentally examined and found to be
anisotropic around the specular peak. This
discrepancy is not negligible even at the incident
angle of 10 degree.
The scattering loss of EUV reflectivity can be
estimated from PSD to an accuracy of a few
percent.
EUVL Symposium, 18-20 October 2010
p.18
References
・ E. Gullikson, Proc. SPIE 3331, pp. 72-80 (1998)
“Scattering from normal-incidence EUV optics”
・ N. Kandaka, et al., EUVL symposium 2004 (2004)
“Measurement of EUV scattering from Mo/SI multilayer mirrors”
・ D. G. Stearns, et al., JAP 84, pp. 1003-1028 (1998)
“Nonspecular x-ray scattering in a multilayer-coated imaging system”
・ Sven Schroeder, et al., Applied Optics 49, pp. 1503-1512 (2010)
“Angle-resolved scattering and reflectance of extreme-ultraviolet
multilayer coatings: measurement and analysis”
EUVL Symposium, 18-20 October 2010
p.19
Acknowledgements
We would like to express our gratitude to…
- staffs of KEK Photon Factory (PF) synchrotron facility
giving us good opportunity to examine the scattering
- our colleagues (team members)
preparing the samples, multilayer coatings,
useful discussion and suggestion,
valuable and perpetual contribution
We hope to achieve steady development of multilayer and
brilliant future of EUVL project !!
EUVL Symposium, 18-20 October 2010
p.20