IMEC 2013
ASML NXE:3100 PRE-PRODUCTION EUV SCANNER PERFORMANCE AT IMEC
ERIC HENDRICKX, JAN HERMANS, GIAN LORUSSO, PHILIPPE FOUBERT, MIEKE GOETHALS, RIK JONCKHEERE, GEERT VANDENBERGHE, KURT RONSE
IMEC 2013
CONTENTS
Introduction
NXE:3100 stability
Outlook to 3300
Conclusions
2013 EUVL SYMPOSIUM - HENDRICKX 2
IMEC 2013
IMEC EUV LITHOGRAPHY TOOL ROADMAP
2006 - 2011 2011 - now 2014
ASML Alpha-Demo tool
40nm � 27nm LS0.25 NA
ASML NXE:3100 –pre production
27nm LS, 22nm LS0.25 NA
ASML NXE:3300 –production
22, 18nm LS0.33 NA
2013 EUVL SYMPOSIUM - HENDRICKX 3
IMEC 2013
NXE:3100
Main specifications
▸ Field size: 26x33mm2
▸ NA=0.25 and σ=0.81
▸ 6 off-axis illumination conditions available
▸ MMO vs NXT:1950i < 7nm
Interfaced to TEL LITHIUS™ Pro - EUVDischarge Produced Plasma source
2013 EUVL SYMPOSIUM - HENDRICKX 4
SUSS MicroTecMaskTrack Pro
EUV TechnologiesOutgassing tool
IMEC 2013
CONTENTS
Introduction
NXE:3100 stability
Monitoring scheme
Productivity
CD control
Overlay control
Out of band radiation
Outlook to 3300
Conclusions
2013 EUVL SYMPOSIUM - HENDRICKX 5
IMEC 2013
MONITORING PROCEDURE ON NXE:3100
2013 EUVL SYMPOSIUM - HENDRICKX 6
FEM
CD
U/o
verl
ay
= NXE:3100
Test\logging Target Parameters
Power Productivity Power at IF, RS, WS
FEM wafer CD control 27nm LS BE, BF, EL, DOF
Uniform wafer CD control 27nm LS CDU
Overlay control
Grid, IF residuals
Ilias Lens stability Lensaberration terms
Al mask exposure
Out of band radiationstability
% OoB
IMEC 2013
NXE:3100 PRODUCTIVITYCUMULATIVE WAFERCOUNT
▸ 24/7 operation
▸ DPP source 2013 average power at IF 4.9W
▸ Average power at waferstage 310 microWatt/mm
▸ Average 2013 throughput 2-3 full wafers per hour
▸ Average system uptime 2013 ~52%
Cumulative wafercount of exposed wafers now exceeds 6000 wafers on NXE:3100
2013 EUVL SYMPOSIUM - HENDRICKX
0
1000
2000
3000
4000
5000
6000
7000
41
2011
49 5
2012
13 21 29 37 45 1
2013
9 17 25 33
Tota
l w
afe
r co
un
t
7
IMEC 2013
NXE:3100 PRODUCTIVITYCOLLECTOR LIFETIME AND IMPACT ON POWER
▸ NXE:3100 DPP source was operated using the same swap flange (collector mirror + debris mitigation) for 9 months
▸ Due to low power on system, it was then decided to replace the swap flange (collector + debris mitigation system)
▸ Both power at IF and power at waferstage were recovered, improving productivity
▸ Post-mortem confirmed collector erosion and Snaccumulation in debris mitigation system
2013 EUVL SYMPOSIUM - HENDRICKX 8
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
0
50
100
150
200
250
300
350
400
450
500
Mar-13 Apr-13
IF power (W
)
Slit Integrated power
(microWatt/m
m) WS power
IF power
DPP collector mirror lifetime exceeded 6 months under normal operation conditions
Power at IF and at waferstageReplacement of swap flange
IMEC 2013
NXE:3100 PRODUCTIVITYLENS TRANSMISSION
▸ Lens transmission can be estimated by power measurement at waferstage and at reticle stage
▸ After 1 year of use, cycles of lens cleaning were started using on-board cleaning and lens transmission improved by ~16%
2013 EUVL SYMPOSIUM - HENDRICKX 9
Cleaning cycle
WS/
RS
po
wer
rat
io
Lens transmission only showed limited impact of one year system use First demonstration of scanner on-board lens cleaning in the field
96%
100%
104%
108%
112%
116%
120%
IMEC 2013
CD CONTROL OVER 21 MONTHSBEST ENERGY 27NM H AND V LS FROM 12/2011 – 08/2013
2013 EUVL SYMPOSIUM - HENDRICKX 10
10.0
11.0
12.0
13.0
14.0
15.0
16.0
17.0
0 20 40 60 80 100 120
Best energy (mJ/cm2)
Wafer Number
H
V
Best dose of 27nm H and V LS showed a gradual drift over the first 10 months of operation – was solved by more frequent calibration of NXE:3100 WS sensors to reference
IMEC 2013
CD CONTROL27NM V LS CD STABILITY OVER 10 MONTHS
2013 EUVL SYMPOSIUM - HENDRICKX 11
07/09/2013
10/08/2013
08/07/2013
0
1
2
3
4
5
6
7
8
9
10
20
21
22
23
24
25
26
27
28
29
30
1 11 21 31 41 51 61
CD 3s (nm)
Average CD (nm)
CD
CD 3s
CD3s - IF
After repair of sensor drift, average CD, intrafield signature,and across wafer CD signature are stable over time
IMEC 2013 2013 EUVL SYMPOSIUM - HENDRICKX
OVERLAY CONTROLINTRAFIELD RESIDUALS FROM DECEMBER 2011 –AUGUST 2013
12
November 2011 September 2013
Intrafield distortion signature is stable over 22 months
First layer: XT:1450Second layer: NXE:3100After removal of 10 par OVL correction model
IMEC 2013
1 11 21 31 41 51 61 71 81 91 101 111 121 131
IF R
esi
du
als
C1
(n
m)
Wafer Count
x
In 22 months, only 5 events when a particle was present on the reticle clamp during monitor. In the last 4 events, the particle could be removed without breaking vacuum
2013 EUVL SYMPOSIUM - HENDRICKX 13
y
OVERLAY CONTROLINTRAFIELD RESIDUALS FROM DECEMBER 2011 –AUGUST 2013
First layer: XT:1450Second layer: NXE:3100After removal of 10 par OVL correction model
Rik Jonckheere, session 2
IMEC 2013 2013 EUVL SYMPOSIUM - HENDRICKX 14
OVERLAY CONTROLGRID RESIDUALS FROM 12/2011 – 09/2013
5
6
7
8
9
10
11
12
13
14
15
1 11 21 31 41 51 61 71 81 91 101 111 121
Grid residual 3s (nm)
3s Grid Residual Y (nm)
3s Grid Residual X (nm)
Wafer grid residual 3s is stable to within +- 1nm and has a stable signature
First layer: XT:1450Second layer: NXE:3100After removal of 6 par OVL correction model
Wafer BS particle
November 2011
September 2013
IMEC 2013
0
1000
2000
3000
4000
5000
6000
7000
-20
-18
-16
-14
-12
-10 -8 -6 -4 -2 0 2 4 6 8
10
12
14
16
18
20
Fre
qu
en
cy
Overlay (nm)
X
Y
NXE:3100 BEST ACHIEVABLE MEASUREDOVERLAY – CPE AND iHOPCAPPLIED
X: |Mean|+3σ: 6.0nmY: |Mean|+3σ: 5.6nm
Reference grid from NXE:3100, second layer on XT:1900i1 wafer, 83 fields, 26x33mm2, 17x22 pts/field
Applying 10-parameter, CPE and iHOPC corrections, brings measured overlay down to 6nm |Mean|+3σ
2013 EUVL SYMPOSIUM - HENDRICKX 15J. Hermans 2012 SPIE 8332
IMEC 2013
NXE:3100 LENS ABERRATIONSRMS ERROR FROM 5 MEASUREMENTS OVER 10 MONTHS
2013 EUVL SYMPOSIUM - HENDRICKX 16
0
0.2
0.4
0.6
0.8
1
1.2
Spherical Coma Astigmatism 3-foil Z5-Z37
rms lens aberration (nm)
Lens aberration measurement over 9-month timeframe shows that aberration signature remains well in specification
UncorrectedCorrected
IMEC 2013 2013 EUVL SYMPOSIUM - HENDRICKX
y = -1.3482x + 247.36
0
10
20
30
40
50
0 100 200 300 400Thickness (nm)
Energy (mJ/cm2)
AL - SEVR140
Eo = 183.47
y = -38.478x + 166.41
0
10
20
30
40
50
2 2.5 3 3.5 4 4.5 5
Thickness (nm)
Energy (mJ/cm2)
ML - SEVR140
Eo = 4.32
OoB (%) in SEVR140 resist on NXE:3100 (DPP) = 2.4%
%100(%)maskAlwithcleartoDose
maskMLwithcleartoDoseOoB ⋅=
MEASUREMENT OF OUT-OF-BAND COMPONENTDOSE-TO-CLEAR OF AL COATED MASK AND ML MASK
17
IMEC 2013
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
01/04/2012 10/07/2012 18/10/2012 26/01/2013 06/05/2013 14/08/2013
OoB (%)
Date
OoB measurements shows variability in 1.6-3.5% range (DPP source), which exceeds the reproducibility of the test (0.02% 3s)Not correlated to major interventions on system
MEASUREMENT OF OUT-OF-BAND COMPONENTDOSE-TO-CLEAR OF AL COATED MASK AND ML MASK
2013 EUVL SYMPOSIUM - HENDRICKX 18
IMEC 2013
CONTENTS
Introduction
NXE:3100 stability
Monitoring scheme
Productivity
CD control
Overlay control
Out of band radiation
Outlook to 3300
Conclusions
2013 EUVL SYMPOSIUM - HENDRICKX 19
IMEC 2013
2012 POR TEST ON NXE:330022NM LS PROCESS CD UNIFORMITY
▸ 22nm LS, NXE:3300, conventional illumination
▸ 2012 POR \ NXE:3100 monitor mask (CDU27_1)
2013 EUVL SYMPOSIUM - HENDRICKX 20
2012 POR tested on NXE:3300 with 22nm LS on 3100 mask -showed IF signature (mask) and process limits at 22nm LS
3 x 5 samplingRaw data – includes mask signature
IMEC 2013
NXE:3300 MONITOR MASK ABSORBER WIDTH UNIFORMITY ON MASK
2013 EUVL SYMPOSIUM - HENDRICKX 21
0
0.5
1
1.5
2
2.5
3
3.5
43
2n
m H
32
nm
V
27
nm
H
27
nm
V
27
nm
H
27
nm
V
22
nm
H
22
nm
V
27
nm
H
27
nm
V
22
nm
H
22
nm
V
22
nm
H
22
nm
V
CDU32 CDU27_1 CDU27_2 CDU22
Mask IntrafieldCD 3s (4x, nm)
ADT 3100 3300
Mask intrafield CD uniformity for 22nm LS at <2nm 3s at 4x
44nm pitch CH (1x)44nm pitch LS (1x)
40nm pitch CH (1x)
34nm pitch CH (1x)
IMEC 2013
▸ NXE:3100 exposure
▸ Wafer coated on TEL Lithius Pro
▸ Dipole 60-X illumination, 20.5 mJ/cm2
▸ Full wafer and full field exposure
▸ CD measured in 3 x 5 field positions, including field edges
▸ Raw data reported – split up in IF and across wafer signature
Total1.67nm 3s
Intrafield1.50nm 3s
Intrafield subtracted0.82nm 3s
2013 EUVL SYMPOSIUM - HENDRICKX 22
EUV RESIST LS PERFORMANCE22nm LS CDU – RESIST C
IMEC 2013 2013 EUVL SYMPOSIUM - HENDRICKX 23
NXE:3300 PROCESS NEW CANDIDATE52NM PITCH CH CD UNIFORMITY AFTER LITHO-ETCH
Litho
Etch
Mean CD: 25.34 nmLCDU 3σ : 2.70 nm
Mean CD: 31.16 nm
LCDU 3σ : 3.79 nm
Contact hole uniformity 3s across wafer improves to 2.70nm 3s through resist etch
31nm CH at 52nm pitch printed to 31nm after lithoCD distribution over 32 fields, 75 CH per field, 27.25 mJ/cm2
5.8nm etch bias1.1nm 3s reduction
Ming Mao et al, 2013 MNE
IMEC 2013
CONCLUSIONS
242013 EUVL SYMPOSIUM - HENDRICKX
NXE:3100
Throughput More than 6000 wafers exposed Collector mirror influences power level~6 months collector mirror lifetime
CD control Stable CD and CD map
OVL control Grid and IF residuals are stable, but can be impacted by particles on mask and wafer BSBest demonstrated is ~6nm (matched to XT1900i)
NXE:3100 has sufficient CD and overlay control to enable pre-production device fabrication in wafer batch modeNXE:3300 preparation for track, resist, and mask ongoing
NXE:3300
Resist Candidate resist tested on NXE:3100 – 22nm VLS 1.67nm 3s
Mask New mask fabricated with improved uniformity
Track Accepted TEL Lithius Pro-Z track (09/2013)
IMEC 2013
ACKNOWLEDGEMENTS
ASML▸ Amir Sharomi
▸ Theo Van Den Akker
▸ Jan-Willem Van Der Horst
▸ Andre Van Dijk
▸ Charles Pigneret
▸ Leo Heusschen
▸ Tom Jonckers
▸ Guiseppina Toto
▸ Sjef Haagmans
▸ Oktay Yildrim
▸ Roel Knops
▸ Mircea Dusa
▸ Timon Fliervoet
2013 EUVL SYMPOSIUM - HENDRICKX 25
TEL▸ Kathleen Nafus
▸ Koichi Matsunaga
▸ TEL support team