Adsorptive Retention Performance Index
for NGL Process Filters
Michael Sevegney, Toru Umeda, Shuichi Tsuzuki, Toru Numaguchi
2011 International Symposium on Lithography Extensions
October 21, 2011
Better Lives. Better Planet.SM
This presentation is the work product of Pall Corporation and
no portion of this presentation may be copied, published,
performed, or redistributed without the express written
authority of a Pall Corporate Officer.
© 2011 Pall Corporation
2
Introduction
Filter rating (e.g. 10 nm, 5 nm, etc.)
• Rating of a filter product
is determined by
Sieving
Surfactant also
minimizes adsorptive
mechanisms
• In actual process fluids,
removal is achieved by
Sieving
Adsorption
3
Performance for defect reduction in litho
process
Introduction
Contact time extended
Nylon 6,6 filter
Nylon 6,6 membrane filters
HDPE membrane filters
60
50
40
30
20
10
0
Rating of filter products / nm
Conceptual
Conceptual rendering
rendering of
of defect
defect reduction
reduction performance
performance against
against filter
filter ratings.
ratings.
4
Introduction
•• Adsorption
Adsorption between
between filter
filter media
media and
and defect
defect sources
sources (②)
(②) can
can be
be affected
affected
by
by adsorption
adsorption between
between solvent
solvent and
and filter
filter media
media (①)
(①) and
and affinity
affinity of
of solvent
solvent
to
to defect
defect sources
sources (③).
(③).
•• Therefore,
Therefore, adsorption
adsorption effect
effect of
of defect
defect sources
sources in
in filtration
filtration should
should be
be
examined
examined in
in each
each process
process fluid
fluid system.
system.
Solvent
③
Defect
sources
①
②
Filter
media
Liquid
Gel or Solid
5
Schematic Summary of the Study
Scope of current presentation
Quantification
of adsorbing
performance
Commercial
results
Comparison
Search
dispersible
nanoparticle
in process
fluid
Filtration
testing
Understand
adsorption
mechanism
Simulated
test particle
selection
Investigate
adsorbing
parameters
using
simulated
particle
Establish
performance
index for
adsorption
enhanced
filtration
products
Predict appropriate
filtration product for new
chemical system
6
Test Particle Selection
•
Test design summary
Test fluid
Actual fluid or its
solvent
Test filter
Actual filter products
Test particle
Defect source is usually not
available
Need to identify a test particle
that exhibits adsorptive
performance similar to real
defects.
7
Test Particle Selection
Property of real defect source found in photoresist
Microbridge is dominant defect on resist pattern that can be reduced using
point-of-use filtration.
Microbridge precursor adsorbs onto nylon 6,6 membrane rather than HDPE
membrane.
Removal efficiency / %
•
Microbridge removal efficiency in
193 nm lithography (L/S 90 nm hp).
Umeda, T., et. al, “Defect reduction by using
point-of-use filtration in new coater/developer”,
Proc. SPIE 7273, 72734B (2009).
Nylon 6,6 20 nm
HDPE 10 nm
8
Test Particle Selection
•
Simulation particle candidate: Metal nanoparticles
Adsorbing property can be varied changing ligand species
Example: Pt-Decylamine (amine, C10)
δ+
H
Pt
Pt
N
δ-
Core metal particle
which determines
particle diameter
Protective ligand
which determines
adsorptive property
Core metal nanoparticle
diameter: 2 - 4 nm
9
Test Particle Selection
•
Simulated particle selection in previous studies
Base solvent
Cyclohexanone
PGMEA
Results
Pt-Decylamine showed
adsorptive retention similar
to microbridge pre-cursors
Unable to differentiate retentive
performance of Pt-Decylamine in
Nylon 6,6 vs. HDPE membranes
References
Umeda, T., et al,
“Determine a performance
index of the filter taking
adsorption into account
using metal nanoparticle
aimed at 1X generation”
SEMATECH ISLE (2010)
Umeda, T., et al, “Performance
index determination for lithography
fluid filters using metal nanoparticle”
JSAP 57th Spring Meeting (2011)
10
Test Particle Selection
•
Log Pow: Octanol-water partition coefficient
Specific adsorption property of nylon 6,6 is assumed to be based on its hydrophilicity
Log Pow describes hydrophobicity of the ligand; used to predict adsorptive property of
the test particle
[
solute ] octanol
Log Pow = Log
[solute ] un−ionized water
Hydrophilic group
O
C
CH2
4
O
H
C
N
H
N
CH2
6
n
Bond-line structure of nylon 6,6
11
Test Particle Selection
•
Test nanoparticles
In the current study, removal efficiency was determined as a function
of Log Pow of ligands.
Core nanoparticle
(4 nm)
Ligand
Log Pow
Pd
Decylamine (C10)
4.1
Pd
Octylamine (C8)
3.1
Pd
Heptylamine (C7)
2.5
12
Test Particle Selection
Results
Similar microbridge removal tendency was achieved using lower-Log-Pow ligand-Pd
nanoparticle.
In PGMEA, Pd-Heptylamine was found to be a good candidate particle for simulating
microbridge pre-cursors.
100%
Nylon 6,6
80%
Removal efficiency
•
Octylamine
Decylamine
60%
40%
HDPE
In PGMEA
Log Pow=0.27
20%
Heptylamine
0%
2
2.5
3
3.5
Log Pow of ligands
4
4.5
13
Plan to Quantify Adsorptive Retention
•
Apparent adsorption kinetics
Adsorption performance of each filter product can be quantified using apparent
adsorption kinetics in the simulated filtration test system.
k : adsorbing rate constant
−
dC
= kC n
dt
C : metal nanoparticle concentration
n : adsorbing reaction order
t : contact time
14
Summary
•
New performance index is needed to indicate real filter performance
which includes adsorptive effect.
•
Substituted metal nanoparticles were found to simulate adsorptive
retention of microbridge precursors.
o
Pt–Decylamine / Cyclohexanone
o
Pd–Heptylamine / PGMEA
•
Adsorption performance of the filtration product to the simulation particle
can be quantified using apparent adsorption kinetics.
•
As a next step, rate equation constants (n, k) will be determined for
various fluid systems in order to establish an adsorptive retention
performance index for NGL process filters.
15
Appendix
16
Sieving Retention of 3 nm Nanoparticles
Table sieving retention of 3 nm nanoparticles in no adsorptive condition.
Membrane
Nanoparticle / size
Solvent
Nylon 6,6 20 nm
Au-Hexanethiol/3 nm
PGMEA:PGME 95:5
HDPE 10 nm
Pt-Decylamine/3 nm
Cyclohexanone
Removal
efficiency
0%
3.6%
17
Dipole moment
δ+
H
Metal nanoparticle
Metal nanoparticle
N
δ-
S
Ligand
Ligand
Amine
No adsorption to HDPE
Thiol
No adsorption to Nylon 6,6
High adsorption to Nylon 6,6
18
www.pall.com/green
19