SIS IN EUV LITHOGRAPHY. 5. SIS + ashing. Pitch 44 nm. Pitch 60 nm. Pitch 80 nm. 0. 0.5. 1. 1.5. 2. 2.5. 3. 23.9. 33.4. 38.5. 3. sL. E. R, n m. CD, nm. PR pattern.
SEQUENTIAL INFILTRATION SYNTHESIS FOR LINE EDGE ROUGHNESS MITIGATION OF EUV RESIST MARINA BARYSHNIKOVA, DANILO DE SIMONE, BT CHAN, SARA PAOLILLO, PAULINA RINCON DELGADILLO, GEERT VANDENBERGHE KRZYSZTOF KACHEL, WERNER KNAEPEN, JAN WILLEM MAES, DAVID DE ROEST CONFIDENTIAL – INTERNAL USE
MOTIVATION: LINE-EDGE ROUGHNESS REDUCTION IN DSA Standard Flow
Square scan
Rectangular scan
LER = 3.8 nm LWR = 2.7 nm
Flow with SIS
LF
Square scan
Rectangular scan
LER = 2.2 nm LWR = 2.5 nm
MF
HF
No SIS With SIS
•
SIS leads to > 40% improvement in LER and 10% improvement in LWR for 14 nm hp DSA lines
•
LER improvements are in the challenging low to mid frequencies *Presented on SPIE Advanced Lithography 2015 by Arjun Singh 2
PUBLIC
OUTLINE
SIS in EUV lithography Description of the technique LER/LWR smoothening Influence of different process parameter ( including EUV PR material) on smoothening effect of SIS Pattern transfer after SIS
Conclusions
3
PUBLIC
SIS IN EUV LITHOGRAPHY
4
PUBLIC
SIS IN EUV LITHOGRAPHY 3sLER, nm
2.5
PR pattern
AlOx pattern
2
1.5 1
PR pattern
0 23.9
33.4
38.5
23.9
CD, nm
16 14 12 10 8 6 4 2 0
AlOx pattern
CD decrease after SIS and ashing
33.4
38.5
23.9
CD, nm
33.4
38.5
CD_PR, nm
4.5 4
3sLER, nm
Pitch 80 nm
SIS + ashing
0.5
Pitch 60 nm
4 3.5 3 2.5 2 1.5 1 0.5 0
CD_AlOx, nm
3
3sLWR, nm
Pitch 44 nm
3.5
After SIS
3
After Litho
2.5 2 1.5 5
10
15
20
25
30
CD, nm 5
PUBLIC
SIS IN EUV LITHOGRAPHY PITCH 32 NM CAR A
CAR A after SIS and ashing
CD=17.2 nm 3σLWR=4.6 nm 3σLER=2.9 nm
•
CD=6.4 nm 3σLWR=2.2 nm 3σLER=2.0 nm
• SIS allows~30% better LER and ~45% better LWR After ashing lines look less continuous due to local thinning (might be caused by height deviation in original line ) 6
PUBLIC
MAIN CHARACTERISTICS OF SIS IN EUV
SIS allows pattern roughness improvement up to 40-45% The smoothening is prominent on small pitches and has no or negative (depending on SIS
and ashing conditions) effect on big pitches (>= 80nm) SIS is accompanied by CD shrinkage by 40-60%
SIS has limited use for 2D patterns
7
PUBLIC
Material composition
Infiltration
Ashing
8
PUBLIC
INFILTRATION CONDITIONS Experiment #1
Al
Experiment #2
Experiment #3
Al
Al
*Alumina distribution in EUC PR blanket layer after SIS. EDS data
•
Infiltration conditions affect metal oxide distribution with the polymer film 9
PUBLIC
ASHING OF PR AFTER SIS
60 40 20 0
Etching time, sec
30 25
30 20 10
P=44nm
20 15 10 5
0
Original pattern
0
Etching time, sec
6
3s LER/LWR, nm
80
40
CD, nm
Thickness after etching, nm
Amount of polymer removed, %
100
Etching time, sec
3sLER
3sLWR
P=44nm
5 4 3 2 1 0
Etching time, sec
Increase of ashing time
CD 15.9 nm 3sLER 3.4 nm 3sLWR 5.0 nm
CD 26.3 nm
3sLER 3.3 nm 3sLWR 4.7 nm
•
CD 11.6 nm 3sLER 2.9 nm 3sLWR 4.1 nm
CD 11.2 nm 3sLER 2.4 nm 3sLWR 2.9 nm
CD 10.4 nm 3sLER 2.3 nm 3sLWR 2.3 nm
Longer ashing time causes higher shrinkage and better 3sLWR/LER 10
PUBLIC
SCREENING OF DIFFERENT EUV RESIST COMPARISON OF BLANKET PR LAYERS D, a.u.
CAR A
44
1.20
25.6
1.5
CAR 1
20
1.06
30.2
1.0
CAR 2
30
1.04
29.3
15.1
CAR 3
40
1.15
37.2
1.5
PMMA
50
1.11
40.4
1.8
CAR 4
50
1.19
20.7
1.1
40 35 J3030 CAR A
30
Rel. uptake, %
Thickness, Density, Uptake, % nm g/cm3
45
CAR 1
25
CAR 2 CAR 3
20
PMMA
15
CAR 4
10 5 0
D – shows how fast resist can be infiltrated U – maximum uptake when resist is saturated with metal precursor
‘short’
‘long’
Infiltration time
o PMMA has the highest uptake because of high concentration of carbonyl groups o CAR 2 has the highest infiltration coefficient (parameter D) 11
PUBLIC
PR on Si wafer, pitch 44 nm
SCREENING OF DIFFERENT EUV RESIST COMPARISON OF PATTERNED PR LAYERS CAR A
CAR 1
CAR 2
CD= 22.0 nm 3σLWR=3.1 nm 3σLER=3.9 nm
CD= 21.3 nm 3σLWR=3.9 nm 3σLER=4.9 nm
CD= 23.0 nm 3σLWR= 3.7 nm 3σLER=4.6 nm
SIS + ashing
SIS + ashing
SIS + ashing
CD=7.9 nm 3σLWR=1.9 nm 3σLER=2.5 nm
CD=9.1 nm 3σLWR=2.6 nm 3σLER=3.8 nm
CD=9.8 nm 3σLWR=2.4 nm 3σLER=3.6 nm
12
CAR 3
SIS + ashing
PUBLIC
Material composition
Infiltration
Ashing
13
PUBLIC
PATTERN TRANSFER Process #1
Process #2
Stack for pattern transfer: AlOx lines
SiOC-10nm
Pattern transfer after optimized breakthrough step
APF-25nm Si
P=80 nm, CD=40 nm
Alumina distribution within line after SIS and ashing
14
PUBLIC
PATTERN TRANSFER After litho:
After SIS and ashing:
After pattern transfer:
30 25
CD, nm
20 15 10 5 AlOx lines
SiOC-10nm
SiOC-10nm
APF-25nm
APF-25nm
Si
Si
0 After litho
After SIS
After ashing
After PT
Si
• Preliminary results indicate that loss in CD can be partially recovered during the pattern transfer 15
PUBLIC
CONCLUSIONS
1.
2.
3.
4.
SIS allows reduction of line-edge roughness of EUV photoresist by 40% (for the best performing EUV PR LER/LWR after SIS and ashing were 1.9 nm/2.5 nm for P44 and 2.0 nm/2.2 nm for P32). Material composition plays the key role in smoothening with SIS
Decrease of LER/LWR is accompanied by CD shrinkage of ~60%. Shrinkage can be compensated during the pattern transfer Pattern transfer (underplaying stack, processing conditions) needs special development
16
PUBLIC
CONFIDENTIAL – INTERNAL USE