Upload
dangthuy
View
224
Download
3
Embed Size (px)
Citation preview
Oct. 21, 2009 2009 EUVL Symposium @ Prague, Czech Republic
Flare Calculation on EUV Optics
Masayuki Shiraishi*Tetsuya Oshino*, Katsuhiko Murakami*, Hiroshi Chiba**
* 2nd Development Department** Optical Design DepartmentDevelopment HeadquartersPrecision Equipment Company
Precision Equipment Company
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 2
OverviewOverview
Flare treatmentDefinition and notationConventional methodNew method
Multilayer extinct effectObscuration of scatterCombine PSFs not via system-PSDTwo PSFs, for flare and TIS
Results of flare evaluationPO #1 – Kirk flare test, flare correction (OPC)PO #2 – Kirk flare test
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 3
OverviewOverview
Flare treatmentDefinition and notationConventional methodNew method
Multilayer extinct effectObscuration of scatterCombine PSFs not via system-PSDTwo PSFs, for flare and TIS
Results of flare evaluationPO #1 – Kirk flare test, flare correction (OPC)PO #2 – Kirk flare test
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 4
Definition and notationDefinition and notationImage Intensity
{ }
∫∫∞
′′′−+−=
⊗+−=
⊗+−=
xdxIxxPSFxITIS
xIxPSFxITIS
xIxPSFxTISxI
SC
SC
SC
rrrrr
rrr
rrrr
)()()()1(
)()()()1(
)()()()1()(
00
00
0δ PSFSC
(1–TIS)δ
PSFFlare
I(x): Flared image intensityI0(x): Original intensityTIS: Total integrated scatterPSFSC: PSF contributing flare; component reaching wafer
∫∫∫∞∞
⋅== drrPSFrxdxPSFTIS SCSC )(2)( 00 πrr
PSFSC0: PSF contributing flare; component as scattered, PSFSC≠
Uniform intensitydecrease Image blur
derived from PSDs
TIS
PDFxITIS ⋅+−= )()1( 0r
DC-flare modelF: DC flarePD: pattern density
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 5
Definition and notationDefinition and notationFlare
Kirk flare (KF)
∫∫ −=fieldbright
00 )()( xdxxPSFxF SC rrrr
ID(xD): Intensity at dark pad xDIB(xB): Intensity at bright area xB
⎪⎪⎩
⎪⎪⎨
⎧
−+−=
−=
=∫∫
∫∫
fieldbright
fieldbright
max
min
)()1()(
)()( ,
xdxxPSFTISxI
xdxxPSFxI
IIKF
BSC
BB
DSC
DD
B
Drrrr
rrrr
0
1
x
Inte
nsity
IDIBFlare effect on image can be expressed using TIS & PSFSC.
“Kirk flare”, which can be experimentally observed as an intensity ratio (= dimensionless), does not equal to “flare.”
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 6
OverviewOverview
Flare treatmentDefinition and notationConventional methodNew method
Multilayer extinct effectObscuration of scatterCombine PSFs not via system-PSDTwo PSFs, for flare and TIS
Results of flare evaluationPO #1 – Kirk flare test, flare correction (OPC)PO #2 – Kirk flare test
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 7
Conventional methodConventional methodFlare calculation from power spectral density (PSD)
)(,),(),( 662211 fPSDfPSDfPSD L
∑=
=6
1
2 )()(i
iiii fPSDfSystemPSD αα
∫ ⋅= drrFrFlare )(2π
equivalent system PSD on pupil
PSF by scattering
Mirror PSDs
⎟⎠⎞
⎜⎝⎛
⎟⎠⎞
⎜⎝⎛=≡
zrSystemPSD
zrPSFrF SC
λλπ 2
24)()(
Problems:- No scatter extinction effect- No obscuration effect- Linear conversion twice(small angle approximation)
- PSF domain limited at afinite distance (not infinity)
- Same PSF for flare & TIS
zrf i
ii ≈≈θλ
∫∞
⋅= drrFrTIS )(2π
r in bright area
Not adequate at a larger scatter angle
Not adequate at alarger scatter angle
“Flare” componentshould be calculatedfrom another PSF,not as-scattered.
This infinite integrationcannot be carried outunder linear domainconversion.
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 8
OverviewOverview
Flare treatmentDefinition and notationConventional methodNew method
Multilayer extinct effectObscuration of scatterCombine PSF not via system-PSDTwo PSFs, for flare and TIS
Results of flare evaluationPO #1 – Kirk flare test, flare correction (OPC)PO #2 – Kirk flare test
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 9
New method: TISNew method: TISTIS calculation from power spectral density (PSD)
)(,),(),( 662211 fPSDfPSDfPSD L
⎟⎟⎟
⎠
⎞
⎜⎜⎜
⎝
⎛
+′⎟
⎟⎠
⎞⎜⎜⎝
⎛′=
22
2
20 4)()(
rz
rDPSz
rrFi
ii
iiλλ
πβ
∫∞
⋅= drrPSFrTIS SC )(2 0π
Mirror PSFs
∑=
=6
1
00 )()(i
iSC rFrPSF
Mirror PSDs
PSF by scattering, component as scattered
Non-linear conversion,not via the SystemPSD.
)()()( iiiiii ExtfPSDfDPS θ⋅=′ Scatter extinction by ML
)sin( iif θλ =
Consideration:- Scatter extinction effect included- Approximation released partially(esp. between finite and infinite)
- Not via system PSD
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 10
New method: FlareNew method: FlareFlare calculation from power spectral density (PSD)
)(,),(),( 662211 fPSDfPSDfPSD L
⎟⎟⎟
⎠
⎞
⎜⎜⎜
⎝
⎛
+′′⎟
⎟⎠
⎞⎜⎜⎝
⎛′′=
22
2
24)()(
rz
rDPSz
rrFi
ii
iiλλ
πβ
∫ ⋅=areabright in
)(2r
SC drrPSFrFlare π
∑=
=6
1)()(
ii
SC rFrPSF
)()()()( iiiiiiii ObsExtfPSDfDPS
Mirror PSDs
Mirror PSFs
PSF by scattering, component reaching wafer
θθ ⋅⋅=′′
Scatter extinction by ML
Non-linear conversion
Obscuration
)sin( iif θλ =
Flare component on wafer
Consideration:- Obscuration effect included- Separate PSF for flare and TIS
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 11
Multilayer extinct effectMultilayer extinct effectScatter extinction by interference in multilayers (MLs) E.M.Gullikson, SPIE 3331
EUV scatter on MLEUV reflection on ML
spec
ular
scattered
substrate
multilayersincidence
incidence incidence
phase matched,peak reflectivity phase
mismatched,extinct
phase matched,peak reflectivity
spec
ular
specular
phase
mism
atched,
extinct
log (scatter), no ML ML extinct coeff.log (extinct * scatter)
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 12
Obscuration of scatterObscuration of scatterObscuration ratio (of component reaching wafer to as-scattered component)
Wafer
static field
Mirrorscattered light,not reaching wafer
scatter loss
scattered light,reaching wafer(within static field)
scattered light, reaching wafer(outside static field)
scatter ray tracing5,000,000 rays were thrown away
from the light source per condition.Internal obscuration was modeled.About 1,500,000 rays reached the
exposure field on the reticle plane.Each ray created scatter light on
each PO mirror in random direction at each scatter angle (including specular lights, of course 100% of them can reach wafer).
Total ~1 billion rays were thrown away. “Obscuration ratios”(how many scatter rays, out of rays leaving reticle, reach wafer) for each scatter angle for each PO mirror were obtained.
internalobscuration
Internal obscuration: apertures, mechanics, mirror coating areas and a flare stop.
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 13
Combine PSF not via system PSDCombine PSF not via system PSDConventional conversion
M1 PSDM2 PSDM3 PSDM4 PSDM5 PSDM6 PSD
SystemPSDon pupil Flare PSF
TISImage Calc.,Kirk Flare, ...
New conversionML extinct
M1 PSF’TIS
Image Calc.,Kirk flare, ...
Obscuration
M1 eq.PSD’
PSFSC
M1 PSF’’M1 eq.PSD’’
PSFSC0
M6 PSF’M6 eq.PSD’
M6 PSF’’M6 eq.PSD’’
......
......
linear linearsame PSF
non-linear
different PSF
equivalent PSD Mirror PSF Flare PSF
M1 PSDM2 PSDM3 PSDM4 PSDM5 PSDM6 PSD
not via systemPSD
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 14
Two PSFs, for flare and TISTwo PSFs, for flare and TISTIS and PSFSC are derived from PSFSC0
∫∫∞
≡ xdxPSFTIS SC rr)(0
Wafer
static field
Mirrorscattered light,not reaching wafer
scatter loss
scattered light,reaching wafer(within static field)
scattered light, reaching wafer(outside static field)
internalobscuration
0SCPSFPSF as scattered:created from mirror PSDs
TIS is obtained by integrationof this to infinity.
SCPSF
total integration(constant)ap
ply
all o
bscu
ratio
nsPSF as reaching wafer:applied all obscurations’effect to PSFSC0
convolution with I0(x). characteristic of PO
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 15
OverviewOverview
Flare treatmentDefinition and notationConventional methodNew method
Multilayer extinct effectObscuration of scatterCombine PSFs not via system-PSDTwo PSFs, for flare and TIS
Results of flare evaluationPO #1 – Kirk flare test, flare correction (OPC)PO #2 – Kirk flare test
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 16
EUV1 PO #1 flare evaluationEUV1 PO #1 flare evaluationKirk-flare measurement By courtesy of Selete
Kirk flare measurement and calculation agree well with each other by ~1% error.
Kirk flare
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
0.1 1 10dark pad's radius [um]
Kirk
Fla
re [%
]
Calculation (outer 26x2mm)Experiment (outer 26x2mm)Calculation (outer 200um dia.)Experiment (outer 200um dia.)
For the smallest pad, the apparent flare seems to be affected also by wavefront error. (~1%)
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 17
EUV1 PO #1 flare evaluationEUV1 PO #1 flare evaluationPrinted CD variation By courtesy of Selete
CD variation with flare canbe precisely predicted usingflare PSF and TIS.
H. Aoyama et al., Jpn. J. Appl. Phys., 48 (2009) 056510.on dark area
on clear area
calculation also including mask CD and resist blur effect
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 18
EUV1 PO #1 flare evaluationEUV1 PO #1 flare evaluationFVC evaluation on PL test site By courtesy of Selete/Toshiba
H. Aoyama et al., Proc. SPIE 7271 (2009) 72712D.
Flare variation compensation (FVC) based on flare PSF and TIS has +/– 2nm CD error in the range of 22%~65% pattern density.
hp 35nm
22% 36% 50% 65%Pattern density:
corn
erce
nter
Mask target size with FVC
FVC evaluation
Raw PSF and multi-fractal function
Top view of SEM images
full chip
flare mapping& mask resizingwith EDA
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 19
EUV1 PO #2 flare evaluationEUV1 PO #2 flare evaluationKirk-flare measurement By courtesy of Intel
Calculated Kirk-flare value is 7.4% at the center of field.
This variation includes mask contrast variation of ~1%and the illumination non-uniformity.
*Additional details are on Intel’s poster presented by Yashesh Shroff on EUV1 static imaging.(“Static Test Evaluation of EUV1 Full-Field Exposure Tools”)
2um pad in bright field
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 20
SummarySummary
New calculation method of flare PSF and TIS is developed.
Using flare PSF and TIS obtained by our new method,
actual flare behavior can be well predicted.
Parameters characteristic of PO are flare PSF and TIS,
which should be separately provided.
(TIS should not be obtained by integrating provided PSF.)
Acknowledgements
H. Aoyama, K. Tawarayama, Y. Tanaka and I. Mori (Selete), providingprecious flare evaluation data on Kirk-flare tests and on PL test site evaluation.Y. Nakajima and R. Inanami (Toshiba), PL test site design and evaluation.A. Myers, Y. Shroff (Intel), providing precious flare evaluation data.M. Nakajima (Nikon), throwing billions of scatter rays for months.
NIKON CORPORATION2nd Development department
Nikon Corporation Oct. 21, 2009 2009 International Symposium on Extreme Ultraviolet Lithography 21
Thank you.