Potential Challenges of Metal-Oxide Based

Photoresist and Subsequent Rework Removal

Benjamin L. Clark1, Michael Kocsis1, Michael Greer1,

Andrew Grenville1, Takashi Saito2, Koichi Hontake2,
Lior Huli2, Richard Farrell2, David Hetzer2,
Hiroie Matsumoto2, Andrew Metz2, Shan Hu2, Fitrianto2,
Jeffrey Smith2, Shinichiro Kawakami3,
Koichi Matsunaga3, Masashi Enomoto3,
Jeffrey Lauerhaas4, Anthony Ratkovich4, David DeKraker4

Inpria Corporation1
TEL Technology Center, America, LLC2
Tokyo Electron Kyushu Limited 3 / TEL FSI, Inc.4
April 19th 2016
EUV Photo Resist Design Principles
Small Molecular
Building Blocks

Photocondensed
Molecular
Metal Oxides

High EUV Robust Etch &

Absorbance Mechanical Properties
EUV Photo Resist Development Strategy

Patterning Fab Integration

Performance (Lab-to-Fab)

Working with equipment, materials, consortia,

university, and device manufacturer partners
Transition From Lab-to-Fab
 Inpria Y-series photoresists contain organometallic complexes, which are
soluble in commonly used fab solvents
 Demonstrating compatibility with fab equipment and processes is critical for
integration in the fab
 Demonstrate:
– Zero cross-contamination
– Film coating and uniformity
– Film defectivity analysis
– Metal residue detection on Si backside and EBR region
– Etch selectivity relative to conventional CAR
– Ability to rework without surface degradation
• Particles
• Metals
 Cross-contamination check
Experiment
schematic: Monitor wafers,
Monitor wafers,
processed through track, Coated 50 wafers wafers with processed through track,
PRE-resist install Inpria Y-series resist POST-resist install

Both sides of all monitor

wafers analyzed with
two techniques:
TXRF (397 pts) VPD-ICP-MS

TXRF detection limit

Typical
fab spec
 Coating Metrics – Defectivity
 Defectivity improvement in resist manufacturing
– Comparison between two resists synthesized with standard and improved filtration
without POU filter
1.20 12000
1.00
Film defectivity @ > 90nm

1.00 10000 Standard

Standardresist
filtration

Frequency
0.80 8000 Aggressive
Improvedfiltration
filtration
0.60 6000
99 %
0.40 4000
w/o POU filter
0.20 2000
0.01
0.00 0

90-106
106-123
123-139
139-155
155-171
171-188
188-204
204-220
220-236
236-253
253-269
269-285
285-301
301-318
318-
Standard filtration
Standard filtration Improvedfiltration
Aggressive filtration

Resist
Defect size [nm]

Improved filtration methods during resist manufacturing

helps reduce film defectivity
 Coating Metrics – Defectivity
 Defectivity improvement by Coater / Developer
– Comparison between standard dispense system and FEF (Filtration
Enhanced Function) with POU filter
1.20 1000.0
1.00
Film defectivity @ > 90nm

Standard dispense
1.00 800.0

Frequency
IDS
FEF
0.80 600.0
0.60
77 %
400.0
0.40
0.23 200.0
w/ POU filter
0.20
0.0
0.00

90-106
106-123
123-139
139-155
155-171
171-188
188-204
204-220
220-236
236-253
253-269
269-285
285-301
301-318
318-
Standard dispense IDS
FEF
Coater/developer Defect size [nm]

FEF on Coater / Developer system also helps to

reduce film defectivity
 TactrasTM

Etch Rate Test

Etch rate comparison under standard Si-ARC/OPL etch condition
Inpria YA series
Conventional CAR resist Inpria YA series
with hard bake
Initial After etch Initial After etch Initial After etch
Shrunk by
X-SEM images bake

X300k

PR thickness 49 nm 30 nm 15 nm 9 nm 9 nm 7 nm
Initial – after ⊿19 nm ⊿6 nm >3X ⊿2 nm >9X

Inpria resist obtained more than 9X higher etch resistance compared to

conventional EUV CAR resist under typical Si-ARC / OPL etch condition,
same processing time.
TEL ORIONTM –hp for Rework
• Unique closed chamber with novel ViPRTM
process using SPM + Steam

• Aggressive strip process to remove resist

and other organic containing layers with
minimum process time & chemical use

60
Challenge Removal 1e15 / 2 keV Implanted Resist

Time to Strip (sec)

50
High dose (>E15) implant DUV 40
Tri-layer (Si-ARC up to 43% Si) 30
20
Amorphous C 10
Plasma Doped (PLAD) 0
Std SW SPM Steam Injected ViPR+
Strip
 ORIONTM-hp
Rework Test
• Resist strip test with steam injected SPM + SC1
– Steam Injected SPM time varied; SC1 time fixed

Resist Detection Limit Post rework

component Method (E10 atoms/cm2) 10 sec 20 sec 30 sec 60 sec
Metal TXRF 5 <5 <5 <5 <5
• Inpria resist stripped successfully with steam injected SPM
– Metal level below detection limit with ≥ 10sec process time
– Defects at baseline levels with ≥ 20sec process time
Conclusions
 Inpria resist compatibility with CLEAN TRACKTM LITHIUS ProTM -
EUV coater/developer has been demonstrated

 Inpria resist stripped successfully with steam injected SPM

 Metal level below detection limit with ≥ 10sec process time

 Defects at baseline levels with ≥ 20sec process time

Acknowledgements
 Special thanks to ASML for NXE3300 exposures

 TELTM personnel at ASML and IMEC

 TEL ES / SPS personnel at TEL Technology Center, America, LLC

/ TEL FSI, Inc.
Thank you