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