Transition between fragmentation and permeable outgassing ...
NIST programs in photoresist outgassing and optics damage
Transcript of NIST programs in photoresist outgassing and optics damage
NIST programs in photoresistoutgassing and optics damage
C. Tarrio, S. B. Hill, S. Grantham, N Faradzhev, T. B. Lucatorto
National Institute of Standards and Technology
Outline
• Resist outgassing measurements
• Optics lifetime testing
• Witness sample testing
• Results from initial witness-plate tests
• Summary
Semi-quantitative method: Combining resist data to rank damage potential
Measure resist outgassing and composition
Measure carbon growth for all significant constituent molecules
Assume partial pressure proportional to amount of each constituent
Assume total carbon growth is sum of growth expected from each component
Weaknesses:
Extrapolation of growth rates to stepper conditions is problematic
Assumes each molecule acts independently from other molecules and background water vapor
Outgassing measurements
5
An example of 6 resist compositional analyses
Total outgas
Significant components
observed
Isobutene Diethyl hexaneBenzene Diphenyl sulfideAcetone HexaneToluene Tert-butylbenzene
6
NIST optics damage facility
Load lockBe filter
Sample
• Average, in-band (13.5 nm ± 2%) intensity 4-8 mW/mm2 at sample• Expose in controlled partial pressure of admitted gases• Bake to 150º C• Base pressure ~2×10-10 Torr• C-deposition measured by XPS and spectroscopic ellipsometry
• Logarithmic pressure dependence continues down to 10-10 Torr
• Surface coverage: Yakshinskiy, et al, Proc. SPIE 7271-36 (2009).
carbon growth rate
Both surface coverage and carbon growth are logarithmic with pressure
8
Extrapolation to lower pressures: potential errors
• More recent work shows mitigation by ambient water vapor must also be considered at very low contaminant partial pressures.
• Intensity scaling also less likely to be linear at lower pressures…
Langmuir
Pure lin
ear
log(p/p0)
px
linear
Extrapolation from only 1 decade of high-pressure data
tert-butylbenzene
C10H14
tert-butylbenzene
C10H14
~p/(1
+a×p)
Witness-plate facility (BL-1B)
•Incorporates grazing incidence broadband (6 - 18 nm) mirror to demagnifyEUV beam on witness plate.
• Sufficient intensity to provide mass limited deposition rates.
• In situ power measurement to ensure accurate resist dose.
• Single-wavelength, Null-field Ellipsometric Imaging System (NEIS) monitors deposition in real-time during exposure.
• Wafer translation and rotation and spot oversampling to insure uniform wafer exposure.
• Narrow band 13.5 nm radiation to expose 200 mm wafer in <1hr.
• Integrated glovebox permits rapid cycle time for samples.
• Base pressure of the vacuum ~ 10-9 Torr.
Witness-plate facility (BL-1B)
Witness-plate
MirrorLight from SURF III
SamplechamberGlovebox
Witness-plate facility (BL-1B)Cut away view of the exposure chamber
Wafer
Witness plate is at intersection of focused EUV beam from synchrotron and NEIS laser beam. Second mirror of focused beam path relays narrow band EUV to the resist.
Witness plate
Relay mirror
Test sequence flow chart
Transport to SURF
H-atom clean WS
Co-expose resist and WS
Transport to SE
SE for carbon thickness
(1)
Post-expH-atom clean
XPS on WSCorrelate at%
to DR/R(2)
Measure Eo of resist
(1) Post exposure DR/R must be ≤2% as determined by carbon thickness measured by spectroscopic ellipsometry(2) Post cleaning (non-cleanable) DR/R ≤0.16% determined by
XPS at%
transport to XPS
EUV spot and spectra for BL1B witness-plate facility
1 mm
50 W/cm2
25 W/cm2
0 W/cm2
Witness-plate
2 mm
17 mW/cm2
8.5 mW/cm2
0 mW/cm2
Wafer plane
0
0.05
0.1
0.15
0.2
0.25
0.3
0
0.015
0.03
0.045
0.06
0.075
0.09
5 10 15 20
Witness-plate Relay mirrorWafer
Po
wer
(m
W/n
m @
300
mA
)
wavelength (nm)
Incident power
NEIS system provides real-time measurement of carbon contamination
Time evolution of line profile
Time evolution of center thickness
•The null-field ellipsometric imaging system (NEIS) thickness monitor provides a real-time measurement of carbon accumulation on the irradiated sample.
•Center thickness evolution normalized to ex situ spectral ellipsometry.
Dose-to-Clear (E0) measurement
•E0 measurement necessary for a proper wafer exposure.
•Turntable action of wafer motion provides simple method for E0 determination.
•Resist processing and thickness measurement provided by Center for Nanoscale Science and Technology (CNST)
0
5
10
15
20
25
30
35
40
2.5 3 3.5 4 4.5 5
E0 run
Thi
ckne
ss (
nm)
Dose (mJ/cm2)
Pre- and post- processing facilities for witness-plate testing at NIST
•Atomic hydrogen cleaning facility for pre- and post- exposure cleaning of witness-plates.
•Scanning spectroscopic ellipsometer for evaluation post-exposure contamination thickness.
•X-ray photoelectron spectroscopy system for post-exposure elemental analysis of contamination from witness-plate test.
•EUV reflectometer available for reflectivity pre- and post-exposure scans of witness-plates.
Three types of metrology correlate WS results
SE
2 4 6 8
1
2
3
4
5
2 4 6 8
XPS
EUVReflectometry
60.6
60.8
61
61.2
61.4
61.6
4 6 8 10 12Position (mm)
Position (mm)
Pos
ition
(m
m)
Pos
ition
(m
m)
Pos
ition
(m
m)
Position (mm)
Position (mm)
Position (mm)
R (
%)
Thi
ckne
ss (
nm)
Thi
ckne
ss (
nm)
Witness plate results show acceptable reproducibility
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
A B C D
Thi
ckne
ss (
nm)
Resist
4 resists for qualification tests:
•Repeatability adequate for qualification purposes.
• Rates varied over a factor of three from the least to the most contaminating.
•No elements other than carbon detected above the ~ 1% atomic concentration level.
Summary
• NIST has constructed a witness plate testing facility with the goal to meet the requirements for resist outgas testing for the ASML NXE
• In situ ellipsometric imaging provides the first real-time measurement of evolution of contaminant across EUV intensity distribution
• On-campus wafer processing, XPS, SE and EUV reflectometry available at NIST
• The NIST facility will soon be available to provide testing on selected key resist for resist developers pending qualification of post-exposure cleaning
• Multiple tests on several resists show good repeatability with sufficient process control
• Comparison between NIST and ASML results indicate that there is good correlation between EUV and electron beam witness-plate tests