Upload
others
View
41
Download
1
Embed Size (px)
Citation preview
IMEC 2012
ASML NXE:3100 PRE-PRODUCTION EUV SCANNER PERFORMANCE AT IMEC
ERIC HENDRICKX, JAN HERMANS, GIAN LORUSSO, PHILIPPE FOUBERT, PETER DE BISSCHOP, GEERT VANDENBERGHE, KURT RONSE
IMEC 2012
IMEC EUV LITHOGRAPHY TOOL ROADMAP
2006 - 2011 2011 - now 2013 ASML Alpha-Demo tool
40nm 27nm LS 0.25 NA
ASML NXE:3100 – pre production
27nm LS, 22nm LS 0.25 NA
ASML NXE:3300 –production
22, 18nm LS 0.33 NA
Improvements in Resists, masks, CD control, overlay, ...
2 2012 EUVL SYMPOSIUM
IMEC 2012
NXE:3100 Main specifications
▸ Field size: 26x33mm2
▸ NA=0.25 and σ=0.81 ▸ 6 off-axis illumination conditions available ▸ Flare < 8% ▸ MMO vs NXT:1950i < 7nm
24/7 Operation
Interfaced to TEL LITHIUS™ Pro for EUV
XTREME Laser assisted DPP source
3 2012 EUVL SYMPOSIUM
First wafer – May 2011
IMEC 2012
CONTENTS
Availability and productivity CD control Overlay control Outlook and conclusions
2012 EUVL SYMPOSIUM 4
IMEC 2012
0
10
20
30
40
50
60
70
80
90
100
wk41 2011
wk45
wk49
wk1 2012
wk5
wk9
wk13
wk17
wk21
wk25
wk29
wk33
wk37
4 –
wee
k av
erag
e up
tim
e (%
)
5
Source upgrade was needed to stabilize source at 7W configuration NXE:3100 4-week average uptime >50% over the last 15 weeks
Source Upgrade
2012 EUVL SYMPOSIUM
4-WEEK AVERAGE UPTIME OF NXE:3100 COMBINED SCANNER AND SOURCE UPTIME
IMEC 2012 6
Cumulative wafercount now 3000 exposed wafers since tool installation – clear productivity increase since May 2012
0
500
1000
1500
2000
2500
3000
3500
41 2011
45
49
1 2012
5
9
13
17
21
25
29
33
Tota
l waf
er c
ount
Throughput ~ 4 full wafers per hour (batch mode, conventional illumination)
2012 EUVL SYMPOSIUM
TOTAL NUMBER OF EXPOSED WAFERS NXE:3100 CUMULATIVE WAFERCOUNT
IMEC 2012
CONTENTS
Availability and productivity CD control Overlay control Outlook and conclusions
2012 EUVL SYMPOSIUM 7
IMEC 2012
0
2
4
6
8
10
12
23
24
25
26
27
28
29
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
3σ C
DU
(nm
)
Aver
age
CD
(nm
)
Wafer count
23 wafers, 83 fields/wafer, 1 point/field, Hitachi CG-4000
0.7nm CD across batch change
Full batch CDU: Avg. CD V: 24.7nm, 3σ: 1.1nm Avg. CD H: 25.2nm, 3σ: 1.2nm
Total CD control for full wafer batch, full wafer coverage ~1nm 3σ (1 point/field)
H
V
8 2012 EUVL SYMPOSIUM
ACROSS BATCH CD STABILITY OF 27nm V/H LS CD DATA FOR BATCH OF 23 WAFERS – STANDARD PROCESS
J. Hermans 2012 SPIE 8332
IMEC 2012
LONG TERM INTRAFIELD STABILITY OF 27nm V/H LS 5 DATAPOINTS OVER 7 WEEKS TIMEFRAME – STANDARD PROCESS
2 field average, 26x33mm2, 13x19 intrafield sampling, Hitachi CG-5000
Uncorrected intrafield CD 3s ~2.4nm 9
3s 26/06/2012 10/07/2012 20/07/2012 07/08/2012 17/08/2012
V 2.46nm 2.17nm 2.18nm 2.41nm 2.61nm
H 2.53nm 2.56nm 2.37nm 2.26nm 2.29nm
V
H
2012 EUVL SYMPOSIUM
IMEC 2012
LONG TERM INTRAFIELD STABILITY OF 27nm V/H LS 5 DATAPOINTS OVER 7 WEEKS TIMEFRAME – STANDARD PROCESS
Full wafer and full field exposure, 26x33mm2, 3x5 intrafield sampling, Hitachi CG-5000 Average exposure field removed
Intrawafer CD 3s with IF removed ~0.85nm
10
3s 26/06/2012 10/07/2012 20/07/2012 07/08/2012 17/08/2012
V 0.81nm 0.85nm 0.87nm 0.84nm 0.83nm
2012 EUVL SYMPOSIUM
IMEC 2012
0
0.2
0.4
0.6
0.8
1
1.2 S
pher
ical
Com
a
Ast
igm
atism
3-fo
il
(Z5-
>Z37
)
Sph
eric
al
Com
a
Ast
igm
atism
3-fo
il
(Z5-
>Z37
)
Sph
eric
al
Com
a
Ast
igm
atism
3-fo
il
(Z5-
>Z37
)
Sph
eric
al
Com
a
Ast
igm
atism
3-fo
il
(Z5-
>Z37
)
Sep-11 Oct-11 Mar-12 May-12
rms
erro
r (n
m)
▸ Lens aberration signature measurement possible using on-board ilias metrology sensor
11
Lens aberration measurements over 9-month timeframe shows that Lens aberration signature remains well in specification
2012 EUVL SYMPOSIUM
ILIAS LENS ABERRATION METROLOGY RMS ERROR FROM 4 MEASUREMENTS OVER 9 MONTHS
IMEC 2012
Using dipole illumination and FIRM surfactinated rinse, 22nm LS can be resolved with good uniformity across wafer – ready for NXE:3300
12 2012 EUVL SYMPOSIUM
PROCESS OPTIMIZATION AT 22nm V LS CD CONTROL FOR NEW REFERENCE PROCESS
Single wafer, 26x33mm2, 3x5 intrafield sampling, Hitachi CG-5000
P. Foubert 2012 EUVL – P – RE 21
IMEC 2012 2012 EUVL SYMPOSIUM 13
y = -1.3482x + 247.36
0
10
20
30
40
50
0 100 200 300 400 T
hick
ness
(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
Thi
ckne
ss (
nm)
Energy (mJ/cm2)
ML - SEVR140
Eo = 4.32
OoB (%) in SEVR140 resist on NXE:3100 (LDP) = 2.4% Similar measurement on ADT: ADT (LDP) < 0.4%
%100(%)maskAlwithcleartoDosemaskMLwithcleartoDoseOoB ⋅=
MEASUREMENT OF OUT-OF-BAND COMPONENT DOSE-TO-CLEAR OF AL COATED MASK AND ML MASK
IMEC 2012 14
SEVR 140
Mean:27.18nm Mean CD:29.91nm
OOB : 0.32% 3 Sigma CD : 2.40nm
OOB: 2.4% 3 Sigma CD: 2.24nm
2012 EUVL SYMPOSIUM
OoB insensitive resist
Resists can be designed to have lower sensitivity to OoB radiation. CD signature is not affected by higher OoB sensitivity
IMPACT OF OOB COMPONENT ON CD INTRADIE CD MAP FOR OOB SENSITIVE AND INSENSITIVE RESIST
IMEC 2012
CONTENTS
Availability and productivity CD control Overlay control Outlook and conclusions
2012 EUVL SYMPOSIUM 15
IMEC 2012
-4
-3
-2
-1
0
1
2
3
4
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Mea
n (n
m)
Wafer #
X
Y
0
5
10
15
20
25
30
35
40
-4
-3
-2
-1
0
1
2
3
4
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
3σ(n
m)
Mea
n (n
m)
Wafer #
X
Y
NXE:3100 FULL BATCH OVERLAY ON ETCHED REFERENCE WAFER
Very good wafer-to-wafer stability (<1nm mean variation over 22 wafers!), full batch, full wafer coverage and full field size
Reference grid from NXT:1950i Standard wafer alignment 23 wafers, 83 fields/wafer, 26x33mm2, 5x5 points/field Avg. batch correctables removed (10 par model)
Full batch overlay: X: |Mean|+3σ: 13.0nm Y: |Mean|+3σ: 11.7nm
J. Hermans 2012 SPIE 8332 16 2012 EUVL SYMPOSIUM
IMEC 2012
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
Freq
uenc
yOverlay (nm)
X
Y
NXE:3100 BEST ACHIEVABLE MEASURED OVERLAY – CPE AND iHOPC APPLIED
X: |Mean|+3σ: 6.0nm Y: |Mean|+3σ: 5.6nm
Reference grid from NXE:3100, second layer on XT:1900i 1 wafer, 83 fields, 26x33mm2, 17x22 pts/field
Applying 10-parameter, CPE and iHOPC corrections, brings measured overlay down to 6nm |Mean|+3σ
17 2012 EUVL SYMPOSIUM
IMEC 2012
Intrafield distortion signature is stable over 10 months difference signature has 1.4nm 1 sigma
December 2011 September 2012
2012 EUVL SYMPOSIUM
NXE:3100 Intrafield distortion vs. XT:1450 reference grid after removal of 4 correctable intrafield terms 13 x 10 fieldpoints, 4-field average – Intrafield higher order corrections were not used
0
50
100
150
200
250
-5 -4 -3 -2 -1 0 1 2 3 4 5
Cou
nt
Delta Distortion (nm)
X (nm)
Y (nm)
18
Difference
INTRAFIELD DISTORTION MAP STABILITY DISTO SIGNATURE OF 12/11 AND 09/12
IMEC 2012
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76
IF R
esid
uals
C1
(nm
)
Wafer Count
X
Y
3 events when a particle was present on the reticle clamp. In the last 2 events, the particle could be removed without breaking vacuum
2012 EUVL SYMPOSIUM 19
DISTO STABILITY OVER 9 MONTHS INTRAFIELD RESIDUALS FROM DECEMBER 2011 – SEPTEMBER 2012
R. Jonckheere EUVL 2012 Session 3
IMEC 2012
CONTENTS
Availability and productivity CD control Overlay control Outlook and conclusions
2012 EUVL SYMPOSIUM 20
IMEC 2012
CONCLUSIONS
21 2012 EUVL SYMPOSIUM
NXE:3100 Throughput 4 full wafers per hour
(conventional, batch mode)
Uptime 4-week average uptime >50% for 15 weeks
Full batch exposure Demonstrated – 23 wafers with good CD and overlay control
IF CDU of 27nm LS Uncorrected
2.4nm
CD stability Intrafield CD fingerprint stable
Overlay control ~6nm matched to immersion XT:1900i
NXE:3100 has sufficient CD and overlay control to enable pre-production device fabrication in wafer batch mode
IMEC 2012 2012 EUVL SYMPOSIUM 22
M1
M2 Via
Spin-on low-k CVD
low-k
30nm Via
1
10
100
.01 .1 1 5 10 20 30 50 70 80 90 95 99 99.9 99.99
30nm Scale Space
0.03_0.03_0.03_0.01_100-C100.03_0.034_0.03_0.015_100-C130.03_0.036_0.03_0.015_100-C140.03_0.04_0.03_0.015_100-C140.03_0.044_0.03_0.015_100-C140.03_0.048_0.03_0.015_100-C14
Res
ista
cne
per v
ia (O
hm)
Percent
30nm Via
Percent R
esis
tanc
e
per
Link
(Ω
)
30nm Space 34nm Space 36nm Space 40nm Space 44nm Space 48nm Space
shorts
ok
30nm Trench
space width
trench width
First ‘full processing’ wafer could be measured electrically Via/link resistance for 30nm Via is as expected (10-15Ω) Good yield for 30nm Vias with 100 links at relaxed space
DEMONSTRATION OF BACKEND ELECTRICALLY TESTABLE 30NM VIA CHAIN
V. Truffert 2012 SPIE 8325
IMEC 2012 2012 EUVL SYMPOSIUM 23
Unit cell (=4 bit-cells)
~65-75% area scaling compared to 20 nm node
(~328 x 1152 nm = 0.095 µm2)
193i litho
Unit cell size [nm×nm] 248 × 990-1080
Bit-cell area [µm2] ~0.061-0.067
EUV
248 × 810
~0.050
~50% area scaling
Compared to 193i litho-only solution, use of EUV lithography allows for ~50% area scaling of SRAM cell and reduction of critical masks at gate, IM1, IM2, via and metal levels
ONGOING 14NM LOGIC DEVICE FABRICATION EUV ALLOWS AREA SCALING AND REDUCES CRITICAL LAYERS
IMEC 2012
ACKNOWLEDGEMENTS
ASML ▸ Amir Sharomi
▸ Theo Van Den Akker
▸ Andre Van Dijk
▸ Leon Romijn
Xtreme ▸ Volker Kilian
▸ Masaki Yoshioka
▸ Guiseppina Toto
▸ Marc Corthout
Zeiss, TEL
Imec Litho department
24 2012 EUVL SYMPOSIUM