Microdroplet laserplasma sources for EUV Lithography

1
Micro-droplet laser-plasma sources for EUV
Lithography
Martin Richardson, Laser Plasma Laboratory,
College of Optics Photonics, CREOL
FPCE, University of Central Florida, Orlando
January 17, 2005
2
The Laser Plasma Laboratory at CREOL

• C-S Koay
• S. George,
• K. Takenoshita,
• R. Bernath,
• J. Duncan,
• Zoubir,
• T. Anderson,
• Ji Choi,
• N. Barbieri,
• C. Brown,
• S. Terrawattanasook

C. Keyser S. Grantham K. Gabel G. Shriever M.
Kado Feng Jin D. Salzmann E. Fu jiwara N. Vorobiev
M. Al-Rabban Dept Physics, Qatar University H. A
Scott Lawrence Livermore National Laboratory,
Livermore, CA V. Bakshi Sematech, Austin, Texas,
USA
3
The EUV Soft X-ray regiona black part of the
spectrum

• Few sources all weak
• Few detectors
• Inefficient optics
• No applications

Until now!
4
The EUV is being opened up by
5
Lithographythe technology that produces todays
computers
Process
0.130 mm process
Pentium 4
6
Lithographydrives the computer chip industry
Computer chip development has been governed for
30 years by Moores Law
7
Minimum Feature Sizeon a computer chip
This size has decreased from optical to EUV
wavelengths
8
Todays Computer Chipsare made with UV laser
light sources
Lithography machine
The wavelength of these UV sources is too large
for tomorrows lithography
9
Tomorrows Lithographywill use EUV sources
Wavelength ?13 nm (EUV)

• Year 2010

• Completely new optics for lithography

• Use mirrors instead of lenses

10
Tomorrows Lithographywill use EUV sources
New type of lithography machine
11
Many New Technologiesto be developed
LPL is making developments in the illuminator
12
The limits of Xenon as an source for 13.5 nm
radiation
Of all the ionized states of Xenon
Xe17
Xe8
Xe9
Xe16
Xe15
Xe10
Xe14
only one state, Xe10, provides emission
at 13 nm.
Xe11
At optimum plasma temperature, 40 are Xe10 ..
and lt 5 of these ions emit at 13.5 nm . of
these only half emit within a 2 band ...we
cannot expect CE much greater than 1
Xe13
Xe12
13
For the last 3 years we have made the case for
tin as an EUV source..
14
In Antwerp we made the case for tin as an EUV
source..
15
In Antwerp we made the case for tin as an EUV
source..
16
We proposed mass-limited droplet technology many
years ago
35 mm
Small diameter droplet targets. Target needs to
be fully ionized by end of laser pulse.
Stability of droplets in target region 3
mm. Plasma located far from nozzle. No nozzle
erosion.
Number of source atoms should be equal to number
of radiators.
High repetition-rate, low cost target
generation (20-100kHz).
towards droplet per shot regimeno excess target
material in chamber
17
We adopt a combined theory experimental
approach to source development
Experimental Studies
Theoretical codes
hydro codes
Irradiation conditions
Plasma Expansion parameters
Quantitative energetics
Radiation code
Plasma diagnostics
Atomic databases
Ionization radiation dynamics
High resolution spectroscopy
In-band metrology
Out-of-band-band metrology
US Patent 6,377,651, B1
18
Two laser systems used in these investigations.
LASER 1 - PRECISION 1 Hz NdYAG LASER Max Energy
1.7 J , Pulse duration 11.5 ns Stability
5, Min focused spot size 35 mm (f10cm
lens) LASER 2 - 100 Hz NdYAG Max Energy 0.35
J, Pulse duration 10 ns
19
Interferometer design
Image Plane
Wollaston Prism
Polarizer1
Polarizer2
Lens
laser
?
plasma
p?
f
b

• Imaging
• M17.75 Resolution 5mm
• Fringe spacing
• ? 532 nm Lens f7.5 cm
• 8.7mrad b 37.9 cm
• i 12.1 mm

20
Plasma density profile from interferometry
0 ns
4 ns
- 4 ns
focused laser beam
21
Comparison of electron density data and hydrocode
calculations
22
Precision 13.5 nm metrology of the droplet source
FC2 measurements June 2004
CE 2.2 at 13.5 nm Radial illumination
leads to higher CE
CE 2.0 with commercial 100 Hz laser
CE 2..........
23
Code predictions for optimum irradiation condition
Sn ion distribution and hydrodynamic calculations
predict optimum intensity for max CE to be 1.0
x 1011 W/cm2

180
Hydro code MED103 Te as function of laser
intensity, for tin-doped (28tin by mass)
spherical target.
160
140
120
100
Te, max (eV)
80
60
40
20
0
1.0E10
1.0E11
1.0E12
1.0E13
Intensity (W/cm2)
24
Still higher values of CE possible..
FC2 used to measure CE from solid tin
In-band CE gt 5.5 with solid tin at 13.5 nm
25
Source size well within Roadmap requirements .
26
Precision spectroscopy of the EUV emission.
Transmission Grating Spectrograph Flat-Field
Grazing incidence Spectrograph
Two EUV instruments
The 4d -4f UTA is the brightest feature in the
radiation from the plasma
(see poster paper SoP39 by S. George et al)
27
Spectral measurements made with two spectrographs
TRANSMISSION GRATING SPECTROGRAPH
FLAT-FIELD SPECTROGRAPH
plasma
Schwanda, Eidmann M. Richardson, J. X-Ray Sci.
Tech, 4, 8-17, (1993)
transmission grating
Rev. Sci. Instr, 72 (1), 2001, 108-118

• Spectrograph Parameters
• Gold plated Hitachi grating
• RG 5.649 m
• Line spacing 1200 lines/mm (nominal)
• Angle of incidence ? 87
• slit width 70?m
• at 10nm ?? 0.04 nm R 250

CCD sensitivity
Rev. Sci. Instr, 68 (9), 1997, 3301-3306
28
Precision spectroscopy of the EUV emission.
Transmission Grating Spectrograph Flat-Field
Grazing incidence Spectrograph
Two EUV instruments
The 4d -4f UTA is the brightest feature in the
radiation from the plasma
(see poster paper SoP39 by S. George et al)
29
Precision spectroscopy examines ion structure of
UTA
I 1.2 x 1011 W/cm2
Subtle differences in shape of UTA variation of
ion distribution
I 6.3 x 1011 Wcm-2 Spot size 35 µm
30
Progress in modeling EUV emission
(see poster paper SoP43 by M. Al-Rabban et al.)
31
Progress in modeling EUV emission
(see poster paper SoP43 by M. Al-Rabban et al.)
32
Progress in modeling EUV emission
(see poster paper SoP43 by M. Al-Rabban et al.)
33
With tin - a wavelength benefit for CE
A small adjustment in the wavelength of choice,
from 13.5 nm to 13.6 nm, significantly increases
the CE.
34
Quantitative ion spectrometry now completes our
understanding of plasma dynamics
Faraday Cup Ion Probe
-60V
Oscilloscope
Thomson Parabola Spectrometer
Droplet target
50 150 mm
Ion spectrometer (Electro Static Ion Energy
Analyzer ESIEA)
NdYAG
Laser beam
Analyzer
TOF field free path
Aperture
DC HV 0 2 kV
100 mm
Plasma
Channel Electron Multiplier
Apertures
MCP Phosphor screen
E B field
Oscilloscope
Ion energy analyzer
Target chamber
35
Identifying ion species by ion spectral analysis
Ion spectrometer signal
Ion mass (M/Z) spectra
Conversion
O3
H
Cl3
O4
E/Z304eV
O2
Cl2
O
Cl
Sn
Sn2
e electron charge R radius of analyzer
path Mp proton mass l CEM distance from
source Eelectric field
All spectral lines are identified
36
Energy (E/Z) vs Mass (M/Z) Mapping on Ion
spectral analysis
Ion spectrometer signal
Ion mass (M/Z) spectra
Interpolation
Intensity 2.4 x 1011W/cm2
Entire ion spectral map obtained for each plasma
condition
By scanning analyzed energy up to E/Z values 5
keV
37
Ion energy distribution by Ion spectral analysis
Ion spectral map
Ion energy distribution (Number of ions at target)
Counting charge

• Steps
• Measured signal V
• electrons flow in load resistance
• ions reaching CEM in analyzed energy window
• ions in unit energy
• ions at target

Ion energy distribution obtained for individual
ion species
Ion spectrum characteristic of thermal plasma.
No fast ion component
(see poster paper SoP37 Takoneshita et al)
38
Ion probe analysis of the effect of the repeller
field
Ion probe signal comparison
Repeller field is installed in front of Ion probe
Laser energy at target 160mJ
Water target
640V
Laser energy at target 240mJ
Fastest ions are stopped by the repeller field.
39
Ion spectral analysis of the effect of the
Repeller Field
Effect of the repeller field on each ion species
Ion kinetic energy 380eV
O
No field
306V
Cl
Sn
Repeller field is installed in front of Ion
spectrometer
408V
Sn-doped target
Ion spectrometer
Individual ion species stopped by the field
strength of the equivalent ion kinetic energy
40
Our approaches to study and minimize particle
emission
41
Our approaches to study and minimize particle
emission
42
Our approaches to study and minimize particle
emission
43
Implications of droplet plasma for stepper source
No power ceiling to this technology Higher powers
available with higher El, fs, hc
44
Threat factors to address
OUT-OF-BAND COLLECTOR HEATING AND WAFER
EXPOSURE 13.5 nm emission band brightest
feature First mirror thermal loading lt 10x source
power
COLLECTOR LIFETIME Minimal source of
ions Repeller field inhibits ions AND
particles Multiple mitigation technologies
Replaceable collector optics
45
Threat factors to address
OUT-OF-BAND COLLECTOR HEATING AND WAFER
EXPOSURE 13.5 nm emission band brightest
feature First mirror thermal loading lt 10x source
power
COLLECTOR LIFETIME Minimal source of
ions Repeller field inhibits ions AND
particles Multiple mitigation technologies
Replaceable collector optics
46
The next steps..
High power source demonstration combine our
source with a high rep. rate laser
Improved modeling, plasma physics and target
design higher CE, controlled radiation dynamics
Ion and particle dynamics. Mirror erosion studies
Long-term source stability
Addition of collectors C1 optic from the ETS,
and others..
Northrop- Grumman donation UCF now owns all the
TRW source equipment, optics and IP Implement
Xe droplet technology,
47
Acknowledgments
David Attwood, Eric Gullickson (Center for X-Ray
Optics) Gerry OSullivan (University College,
Dublin, Ireland) Steve Grantham (NIST) Etsuo
Fujiwara (Himijei Institute, Japan) Greg
Shimkaveg, Somsak Teerawattenasook, Joshua Duncan
(SoOptics-CREOL) Santi van der Westen, Caspar
Bruineman Fred Bijkerk (FOM The
Netherlands) and many others