Transcript Thorium Based Thin Files as EUV Reflectors

Thorium Based Thin Films
as EUV Reflectors
Jed Johnson
Brigham Young University
Reflectors in EUV range
EUV range is about
100-1000Å
General Challenges:
- hydrocarbon buildup
- absorption
- high vacuum needed
Complex index of
refraction: ñ=n+ik
Applications of EUV Radiation
Thin Film or Multilayer Mirrors
EUV Lithography
EUV Astronomy
Soft X-ray Microscopes
The Earth’s magnetosphere in the EUV
Images from www.schott.com/magazine/english/info99/ and www.lbl.gov/Science-Articles/Archive/xray-inside-cells.html.
Creating Thin Films
• Ions from an induced argon plasma bombard a target. Atoms are
then ejected from the target and accumulate as a coating on the
substrate.
Measuring Reflectance
Data is taken primarily at the ALS (Advanced Light Source) at
LBNL in Berkeley, CA. Accelerating electrons produce high
intensity synchotron radiation.
Why Actinides?
Delta vs. beta scatter plot at 4.48 nm
Note: Nickel and its neighboring 3d elements are the
nearest to uranium in this area.
ñ r  n  ik  1    i
  1  n,
 k
Periodic table
Beta vs. Delta + Beta
30.4 nm (41 eV)
Thorium vs. Uranium
Why such a large difference in optical properties?
Thorium (11.7 g/cm^3) is less dense than uranium
(19.1 g/cm^3).
Calculated Reflectance
vs. energy (eV) at 5 deg
1
0.9
0.8
Reflectance
0.7
Gold
Ir
Ni
U
Thorium
0.6
0.5
0.4
0.3
0.2
0.1
0
0
100
200
300
Photon Energy
400
500
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Reflectance
Reflectance
Measured reflectances of UOx, NiO on Ni, and Ni on quartz at 5 degrees from 2.7-11.6 nm
2.5
3
NiO on Ni
4.5
5
4.4
Ni on Quartz
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Reflectance
Reflectance
UOx
3.5
4
Wavelength (nm)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
6.6
7.1
UOx
7.6
8.1
Wavelength (nm)
NiO on Ni
Ni on Quartz
8.6
4.9
5.4
5.9
6.4
Wavelength (nm)
UOx NiO on Ni Ni on Quartz
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
8.4
8.9
UOx
9.4
9.9
10.4
Wavelength (nm)
NiO on Ni
10.9
Ni on Quartz
11.4
However….
The mirror’s surface will be oxidized.
At optical wavelengths, this oxidation is
negligible. It is a major issue for our thin films
though.
Problems with Uranium
Immediate oxidation to UO2. (10 nm in 5 min)
Further oxidation to U2O5 is less rapid. (5 – 10
nm in six to 12 months)
Can even proceed to UO3!
Lower density = lower reflectance
Contrasting two types of oxidation
Reflectance of Naturally Oxidized and
Reactively Sputtered UO2
1
Reflectance
0.8
0.6
0.4
0.2
0
3
5
7
9
11
Wavelength (nm)
13
UO18-Naturally oxidized UO2 [i]
Lunt UOx on UO2-Reactively Sputtered [ii]
The mirrors end up the same!
15
17
A Possible Alternative: Thorium
Only one oxidation state:
ThO2. We know what we
have!
The densities of UO2 (about
11 g/cm3) and ThO2 (9.85
g/cm3) are similar.
Rock stable: Highest melting
point (3300 deg C) of any
known oxide.
Calculated Reflectance
vs. energy (eV) at 10 deg
Reflectance, S polarization at 10 degrees of various materials
0.9
0.8
Reflectance
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
100
200
300
400
Energy in eV
Au
Ni
ThO2
UO2
500
600
First Thorium Reflectance Data
(Nov. ‘03 ALS)
M easured Reflectance of Th02 at 10 degrees
0.8
0.7
0.6
Reflectance
0.5
0.4
0.3
0.2
0.1
0
0
5
10
15
20
25
30
Wavelength (nm)
2.16-2.8 nm
2.7-4.8 nm
4.4-6.8 nm
12.4-18.8 nm
17.2-25.0
22.5-32.5
6.6-8.8 nm
8.4-11.6 nm
11.0-14.0 nm
35
Measured and Calculated
Reflectance at 10 deg
Measured and Calculated Reflectance of Th02 at 10 degrees
0.9
0.8
0.7
Reflectance
0.6
0.5
0.4
0.3
0.2
0.1
0
0
5
10
15
20
25
30
35
40
Wavelength (nm)
2.16-2.8 nm
2.7-4.8 nm
4.4-6.8 nm
6.6-8.8 nm
8.4-11.6 nm
11.0-14.0 nm
12.4-18.8 nm
17.2-25.0
22.5-32.5
calc. AFM CXRO S polarized
45
“Zoomed in”
(and nm  eV)
Thorium oxide
1.00
0.90
0.80
0.70
calc ThO2 42nm
Reflectance
0.60
thick ThO2
measured
0.50
meas
meas
0.40
meas
0.30
0.20
0.10
0.00
70
75
80
85
90
95
eV
100
105
110
115
120
Higher Energies
Thorium oxide
1.00
0.90
0.80
0.70
calc ThO2 42nm
thick ThO2
Reflectance
0.60
measured
meas
0.50
meas
meas
meas
0.40
meas
meas
0.30
0.20
0.10
0.00
100
120
140
160
180
200
eV
220
240
260
280
300
Recap of Differences
At some points, the measured curve lags
around 15 eV behind predicted curve.
Some regions of theoretically high
reflectance are drastically and inexplicably
low.
Now the important question: WHY???
Einstein’s Atomic Scattering Factor
Model
Photons are scattered
principally off electrons.
More electrons = higher
reflection.
Assumption: condensed
matter may be modeled
as a collection of noninteracting atoms. In the
higher energy EUV,
chemical bonds shouldn’t
contribute. (except near
threshold regions)
Can the ASF model be applied in
the visible light range?
Silicon (opaque)
and oxygen
(colorless gas)
combine to form
SiO2 (quartz).
Clearly the
chemical bonds
have a dramatic
effect on the
compound’s
properties.
Where then is the ASF model
valid?
At some point,
ASF model and
measured data
should converge.
Unpublished
BYU study: SiO2
plots never
converged up to
300 eV.
So, what is the explanation?
3 possibilities, none of which is totally
convincing.
Thorium oxide
1.00
0.90
0.80
0.70
calc ThO2 42nm
Reflectance
0.60
thick ThO2
measured
0.50
meas
meas
0.40
meas
0.30
0.20
0.10
0.00
70
75
80
85
90
95
eV
100
105
110
115
120
Possibility #1
Bad experiment! Data has never been so clean
though and the features are clear.
Beamline 6.3.2 coordinator at ALS has no
explanation.
M e asure d Re fle ctance of Th02 at 10 de gre e s
0.8
0.7
0.6
Reflectance
0.5
0.4
0.3
0.2
0.1
0
0
5
10
15
20
25
30
Wavelength (nm)
2.16-2.8 nm
2.7-4.8 nm
4.4-6.8 nm
12.4-18.8 nm
17.2-25.0
22.5-32.5
6.6-8.8 nm
8.4-11.6 nm
11.0-14.0 nm
35
Possibility #2
The sputtered
film wasn’t
pure thorium.
Possibly an
alloy?
EDX w/ SEM
indicates 
Carbon and Oxygen
Cutting fluid residue left on target?
Thorium carbide?
Hydrocarbon contaminant?
Carbon Impurities in silicon? (EDX “sees through”)
Surface XPS only sees Th.
Bottom Line: none of
these small
contributions could
have caused a drop
from ~70% to ~10%
reflectance.
Possibility #3
Maybe chemistry IS playing a larger role in this
region than previously expected.
Could the atomic scattering model be
somewhat incorrect in this range?
What lies ahead…
Thoroughly clean target and substrate.
Produce additional oxidized Th samples.
Verify November measurements. If
reproducible, formulate explanation.
Conclusions
Thorium shows definite promise as a good,
durable reflector in the EUV.
Contingent upon follow up trials, it is possible the
Atomic Scattering Factor model needs revision in
the EUV.
Acknowledgments
BYU XUV Research Group colleagues
Dr. David D. Allred
Dr. R. Steven Turley
BYU Physics Department Research
Funding
Hydrocarbon Contaminants
Airborne hydrocarbons accumulate on mirror
surfaces.
Buildup Rates
Spectroscopic Ellipsometry indicates the
thickness of deposits.
Hydrocarbon Buildups Lower
Reflectance
Reduced Reflectance with Hydrocarbon Thickness.
Theoretical change in reflectance vs. grazing angle and
organic thickness. (at λ=40.0 nm)
Preparing a Standard Contaminant
DADMAC (polydiallyldimethyl-ammonium chloride) is used as the
standard contaminant which coats the surface.
Salt concentration affects shielding and eventual thickness of
DADMAC layer.
Four Methods of Cleaning Tested
Opticlean®
Oxygen Plasma
Excimer UV Lamp
Opticlean® + Oxygen Plasma
Opticlean®
Procedure:
Applied with brush, left to dry,
peeled off (DADMAC comes too)
Results:
2 nm polymer residue left
(ellipsometry)
XPS revealed the components of
Opticlean® (F,O,Si,C), but not
heavier metals used in thin films.
Prominent thin-film lines: U-380
eV, V-515 eV, Sc-400 eV.
No surface damage (SEM)
Oxygen Plasma Procedure
Oxygen plasma is formed
between two capacitor
plates by inducing a radio
frequency (RF) electric
current across the plates.
High energy ions
mechanically break up
molecular bonds of the
surface molecules and
blast them off surface.
Atomic oxygen in the plasma
readily reacts with the
surface contaminants,
breaking them up into
smaller and more volatile
pieces which easily
evaporate.
Oxygen Plasma Results
2
Change in thickness (Å)
Contaminants are
removed rapidly.
Concerns:
 Top graph indicates
increase in thickness
over time… oxidation
 Bottom graph confirms
growing layer is NOT
hydrocarbons. There is
no XPS carbon peak.
Results of O2 Plasma Exposure
0
-2 0
2
4
6
8
10
12
-4
-6
-8
-10
-12
Minutes In Plasma
14
16
18
UV Lamp Theory
High energy photons
break up hydrocarbon
bonds. Volatile fragments
leave the surface.
UV produces oxygen
radicals which react with
oxygen gas to form
ozone. The reactive
ozone oxidizes
contaminants and they
evaporate.
UV Results
UV Lamp
Change in Thickness (Ǻ)
4.5 Å DADMAC layer
eliminated rapidly,
followed by slow
oxidation.
XPS shows no carbon
peak.
Concern: silicon
doesn’t appear to
oxidize, but mirror
coatings such as U
and Ni do.
2
1
0
-1 0
-2
-3
-4
-5
-6
-7
-8
5
10
15
Time Under UV Lamp (s)
20
25
Opticlean® + Plasma
Very effective: Removes both
large and small particles.
Drawback: Procedure is long and
specialized equipment required.
Recommendations
1. For rigorous cleaning, Opticlean® +
Plasma is most effective
2. UV Lamp shows potential for ease
and quickness, but heavy oxidation
can ruin surfaces
3. Further Study: Which surfaces will
oxidize (from UV) and how much?