Accurate Models for EUV Simulation and Their Use for ... - Sematech

Accurate Models for EUV
Simulation and Their Use for
Design Correction
Gian Francesco Lorusso, Jan Hermans, Bart Baudemprez,
Eric Hendrickx (IMEC)
U. Klostermann, Stephen Jang, Lena Zavyalova, Jacob
Sorensen, Weimin Gao, Kevin Lucas (Synopsys)

Goals
• The ASML EUV ADT became
operational at IMEC since June
2008
• Experimental CD data became
available to validate model and
simulator accuracy
• Full Resist Model (Calibration and Verification)
• OPC Model (Calibration and Verification)
2009 International Symposium of Extreme Ultraviolet Lithography
Gian Francesco Lorusso
© imec 2009 2

Outline
• Full Resist Model Calibration and Verification
– Introduction
– Calibration Data and Input Parameters
– Model Fit and Output Parameters
– Flare, CD and Pitch Validation
– CDU Analysis and Simulation
• OPC Model Calibration and Verification
– Data
• Description
• Proximity through Flare
• MEEF
• Precision and Stability
– Flare
– Shadowing
– Modeling
• Flare Map Accuracy
• Flow and RMS Error
• 2D Model Verification
• Conclusions
2009 International Symposium of Extreme Ultraviolet Lithography
Gian Francesco Lorusso
© imec 2009 3

Full Resist Model Calibration and Verification:
Calibration Data and Input Parameters
• Reticle
– Die: 1 slit position, 1 scan position
• Features
– CD 32, 36, 40nm
– Pitch 1:1, 1:5
– Vertical and Horizontal orientation
• Layout
– FEM 7 x 9
• Resist
– ShinEtsu SEVR-59 (65nm thickness)
• Exposure
– NF = 0nm, DF = 50nm
– NE = 17 mJ/cm2 , DE = 1 mJ/cm2 • Metrology / Analysis
– Hitachi CG4000
FEM’s 32, 36, 40nm, DL and ISO, H and V
2009 International Symposium of Extreme Ultraviolet Lithography
• Optical
– λ = 13.6nm, σ = 0.5, NA = 0.25
• Mask
– Mask CD data
• Flare
– 6.3% (Calculated using Zeiss PSF)
• Resist
– Traditional CAR
Gian Francesco Lorusso
© imec 2009 4

CD (mm)
CD (mm)
RMS (nm)
Full Resist Model Calibration and Verification:
Model fit and Output Parameters
0.050
0.045
0.040
0.035
0.030
0.025
0.045
0.040
0.035
0.030
0.025
0.020
2.0
1.0
0.0
32nm V DL
-0.15 -0.10 -0.05 0.00
Focus (mm)
0.05 0.10 0.15
-0.15 -0.10 -0.05 0.00 0.05 0.10 0.15
Focus (mm)
32nm V ISO
V H V H V H
32nm LS 36nm LS 40nm LS
13 mJ/cm2
14 mJ/cm2
15 mJ/cm2
16 mJ/cm2
17 mJ/cm2
18 mJ/cm2
19 mJ/cm2
20 mJ/cm2
13 mJ/cm2
14 mJ/cm2
15 mJ/cm2
16 mJ/cm2
17 mJ/cm2
18 mJ/cm2
19 mJ/cm2
20 mJ/cm2
DL
ISO
2009 International Symposium of Extreme Ultraviolet Lithography
• Lateral Diffusion length of 18nm
– Value in line with expectations
• Large deprotection rate k1
– Deprotection rate was increased
for this resist to achieve lower
dose
• Bossung Asymmetry
– The model reproduces well
Bossung tilt and focus shift as a
3D effect
Overall Fit RMS ~ 1nm
Gian Francesco Lorusso
© imec 2009 5

Full Resist Model Calibration and Verification:
Flare, CD and Pitch Validation
(See slide 12 for
exposure conditions)
RMS (nm)
5
4
3
2
1
0
design vs wafer
model vs wafer
H V H V H V
Mid flare Low flare High flare
• Average model error ~ 0nm
• Maximum model error ~ 2nm
• Remaining tendencies
– Model error is not random through pitch
Overall Validation RMS ~1nm
2009 International Symposium of Extreme Ultraviolet Lithography
average, min and max error (nm)
3
2
1
0
-1
-2
-3
H V H V H V
Mid flare Low flare High flare
Gian Francesco Lorusso
© imec 2009 6

Full Resist Model Calibration and Verification:
CDU Analysis
23 fields NF
Mask CDU32
65nm SEVR59
Wafer and Mask metrology
25 slits, 4 scans
Intradie
FW CDU HV = 1.88nm
0.82nm
1.25nm
Mask
32nm FW CDU NF HV = 1.88nm
2009 International Symposium of Extreme Ultraviolet Lithography
0.94nm
Intradie after REC
FW CDU HV = 1.62nm
Reticle
Die
Wafer
Gian Francesco Lorusso
© imec 2009 7

Average CD
Full Resist Model Calibration and Verification:
CDU Analysis
32.0
31.5
31.0
30.5
30.0
29.5
Mask
-12.88
After REC
-9.70
-6.52
Best focus (nm)
-3.34
40
20
0
-20
-40
-60
-80
-0.15
-12.88
3.03
Slit position (mm)
-8.64
6.21
-4.40
Focus
-0.15
4.09
Slit position (mm)
CD V
CD H
9.39
12.57
-12.88
BF H
BF V
8.33
12.57
After Focus
-9.70
-6.52
-3.34
-0.15
3.03
Slit position (mm)
6.21
Shadowing Intensity
CD V
CD H
9.39
Intra-die CDU can be fully explained by mask, focus, shadowing
and intensity fingerprints
2009 International Symposium of Extreme Ultraviolet Lithography
12.57
-12.88
After Shadowing
-9.70
-6.52
-3.34
-0.15
3.03
Slit position (mm)
6.21
CD V
CD H
9.39
12.57
-12.88
After Intensity
-9.70
-6.52
-3.34
-0.15
3.03
Slit position (mm)
6.21
9.39
Gian Francesco Lorusso
© imec 2009 8
12.57

Full Resist Model Calibration and Verification:
CDU Simulation
Simulator Input: Measured mask CD, slit intensity, best focus and mask azimuth, SEVR59 full resist model
Experimental H Simulated H Experimental V Simulated V
Simulated H CD (nm)
32.5
32.0
31.5
31.0
30.5
30.0
29.5
29.0
y = 0.913x + 2.922
R 2 = 0.7448
28.5
29.0 29.5 30.0 30.5 31.0 31.5 32.0 32.5
Experimental H CD (nm)
33.5
33.0
32.5
32.0
31.5
31.0
30.5
30.0
29.5
2009 International Symposium of Extreme Ultraviolet Lithography
Simulated V CD (nm)
y = 0.9327x + 1.9051
R 2 = 0.788
29.0
29.5 30.0 30.5 31.0 31.5 32.0 32.5 33.0 33.5
Experimental V CD (nm)
RMS error 0.3nm, maximum delta between experiment and simulation 1nm
Gian Francesco Lorusso
© imec 2009 9

Outline
• Full Resist Model Calibration and Verification
– Introduction
– Calibration Data and Input Parameters
– Model Fit and Output Parameters
– Flare, CD and Pitch Validation
– CDU Analysis and Simulation
• OPC Model Calibration and Verification
– Data
• Description
• Proximity through Flare
• MEEF
• Precision and Stability
– Flare
– Shadowing
– Modeling
• Flare Map Accuracy
• Flow and RMS Error
• 2D Model Verification
• Conclusions
2009 International Symposium of Extreme Ultraviolet Lithography
Gian Francesco Lorusso
© imec 2009 10

OPC Model Calibration and Verification:
Introduction
TM08 EUV OPC1
• Model Calibration and Verification
• 3 flare flavors, H and V features
Similar
flare range
Goals:
• Corrected and uncorrected fields
• 2 flare flavors
• 2 corrected sub-dies
1. Calibrate and verify EUV OPC models on TM08
2. Produce and verify corrected designs on EUV OPC1
2009 International Symposium of Extreme Ultraviolet Lithography
Gian Francesco Lorusso
© imec 2009 11

OPC Model Calibration and Verification Data:
Description
• Reticle
– Die: 1 slit position, 3 scan positions
• Features
– L-bar CD 36 ÷ 46nm, Pitch 1:1 ÷ 1:5
– MEEF S CD 35 ÷ 40nm, Pitch 1:6
– MEEF L CD 32 ÷ 50nm, Pitch 1:1 ÷ 1:5
– EOL, T-EOL, Brick Wall CD 32 ÷ 40nm
• Vertical and Horizontal orientation
• Mask measurements
• > 700 CD measurements/die at wafer
• Layout
– 23 fields exposed in nominal focus and dose
• Resist
– ShinEtsu SEVR-59 (65nm thickness)
• Metrology / Analysis
– Hitachi CG4000
2009 International Symposium of Extreme Ultraviolet Lithography
Die Layout
MEEF S MEEF L
L-bar
EOL
Gian Francesco Lorusso
© imec 2009 12

OPC Model Calibration and Verification Data:
Proximity Through Flare
shadowing
CD (nm)
CD (nm)
CD (nm)
Low Flare
2009 International Symposium of Extreme Ultraviolet Lithography
High Flare
42
42
42
40
40
40
38
36
38
CD 36 V (nm)
CD 36 36 H (nm)
38
36
CD 36 V (nm)
CD 36 H (nm)
34
34
34
32
32
32
30
30
30
46
44
0 100 200
Pitch (nm)
300 0
46
44
100 200
Pitch (nm)
300
46
44
0 100 200
Pitch (nm)
300
42
40
42
CD 40 V (nm)
40
CD 40 H (nm)
42
40
flare
CD 40 V (nm)
CD 40 H (nm)
38
38
38
36
36
36
34
34
34
0 100 200 300 0 100 200 300 0 100 200 300
50
Pitch (nm)
50
Pitch (nm)
50
Pitch (nm)
48
48
48
46
44
42
46
CD 44 V (nm)
44
CD 44 H (nm)
42
46
44
42
CD 44 V (nm)
CD 44 H (nm)
40
40
40
38
38
38
0 100 200 300 0 100 200 300 0 100 200 300
Pitch (nm)
Pitch (nm)
Pitch (nm)
( )
Medium Flare
• Clear impact of flare on CD
• Clear Impact of shadowing on HV bias
Gian Francesco Lorusso
© imec 2009 13

OPC Model Calibration and Verification Data:
MEEF
MEEF (a.u.)
MEEF (a.u.)
3
2
1
0
3
2
1
0
Dense Lines
32nm 35nm 40nm
CD (nm)
45nm 50nm
Isolated Lines
32nm 35nm 40nm
CD (nm)
45nm 50nm
DL V
DL H
ISO V
ISO H
2009 International Symposium of Extreme Ultraviolet Lithography
MEEF (a.u.)
MEEF (a.u.)
3
2
1
0
3
2
1
0
Semi-dense Lines
32nm 35nm 40nm
CD (nm)
45nm 50nm
Spaces
32nm 35nm 40nm
CD (nm)
45nm 50nm
SEMI V
SEMI H
SPACE V
SPACE H
• MEEF data were used for calibration/validation and MEEF analysis
• Larger MEEF for Small DL and for Spaces; MEEF ~1 for Semidense and Iso
• Marginal differences between MEEF H and V or through flare
Gian Francesco Lorusso
© imec 2009 14

3 (nm)
3 (nm)
3 (nm)
1.2
1.0
0.8
0.6
0.4
0.2
0.0
1.2
1.0
0.8
0.6
0.4
0.2
0.0
1.2
1.0
0.8
0.6
0.4
0.2
0.0
OPC Model Calibration and Verification Data:
Precision and Stability
Low Flare
36 38 40 42 44 46
CD (nm)
Medium Flare
36 38 40 42 44 46
CD (nm)
High Flare
36 38 40 42 44 46
CD (nm)
Precision calculated across uniformly
exposed wafer on 23 fields
V
H
V
H
V
H
CD (nm)
42
40
38
36
34
32
30
First Run (Mar 19, 2009)
17mJ/cm
0 100 200 300
Pitch (nm)
2
• Good precision across wafer (~ 1nm)
• Good CD and Flare stability
2009 International Symposium of Extreme Ultraviolet Lithography
CD (nm)
42
40
38
Second Run (Aug 9, 2009)
25mJ/cm
CD 36nm V
36
CD 36nm H
34
32
30
0 100 200 300
Pitch (nm)
2
• Flare has been monitored at IMEC
for over 1 year and resulted stable
CD 36nm V
CD 36nm H
Gian Francesco Lorusso
© imec 2009 15

OPC Model Calibration and Verification:
Flare
Low Flare
Medium Flare
High Flare
• Flare sensitivity ~ 1.3nm/%
9mJ/cm 2
Measured CD (nm)
y = -1.2847x + 49.52
2009 International Symposium of Extreme Ultraviolet Lithography
44
43
42
41
40
39
38
37
36
35
20mJ/cm 2 28mJ/cm 2
Flare measured using contrast-based flare metrology *
0 5
Experimental Flare (%)
10
* GF Lorusso et al, “Flare in EUVL: Metrology, OoB,
Fractal PSF, and Flare Map Calibration” JM3, in press
Gian Francesco Lorusso
© imec 2009 16

OPC Model Calibration and Verification:
Shadowing
RMS (nm)
6
5
4
3
2
1
0
Variable
Measured Measured
-6 -4 -2 0
HV Bias (nm)
ALL
SPACE
ISO
DENSE
SEMI
RMS (nm)
2009 International Symposium of Extreme Ultraviolet Lithography
5
4
3
2
1
0
NO BIAS SINGLE BIAS MULTIPLE
BIAS
• Single Bias = one bias for all feature type
• Analysis
• Based on experimentally measured MEEF
• Through Flare
• Single bias correction effective to
compensate shadowing for 1D features
• Experimental analysis in agreement with
simulations
Gian Francesco Lorusso
© imec 2009 17

OPC Model Calibration and Verification:
Flare Map Accuracy
REMA Open (26.0mm) REMA Closed (1.1mm)
6
5
4
3
2
1
• 6 flare sites were measured with REMA
blades closed and open on 2µm targets
• The 12 flare measurements (4-12%
range) were compared to flare simulations
to check accuracy
6
5
4
3
2
1
2µm
2009 International Symposium of Extreme Ultraviolet Lithography
Flare (%)
18
16
14
12
10
8
6
4
2
0
Simulation
Open
Closed
0 10 20 30 40
Position (mm)
0.15
0.10
0.05
0.00
2µm
y = 0.9665x + 0.0217
R 2 = 0.9847
0.00 0.05 0.10 0.15
Simulated Flare (%)
• The flare simulation agrees well with experimental flare data
• A ~2% offset is observed, attributed to DUV
Experimental Flare (%)
Flare (%)
18
16
14
12
10
8
6
4
2
0
Experiment
Open
Closed
0 1 2 3 4 5 6
Site (#)
Gian Francesco Lorusso
© imec 2009 18

OPC Model Calibration and Verification:
Model Flow and RMS
• Optical Model
• NA = 0.25
• σ = 0.50
• Resist Model
• Simple Gaussian model
• Shadowing
• Single bias rule-based correction
• HV bias = 1.5nm/edge
•Flare
• Flare map-based correction
• Flare sensitivity = 1.3nm/%
experimental
RMS 1D RMS all RMS
L_Bar
RMS
MEEF
RMS ~ 1nm
RMS
Space
RMS
EOL
0.64nm 0.74nm 0.71nm 0.63nm 0.41nm 1.34nm
2009 International Symposium of Extreme Ultraviolet Lithography
Gian Francesco Lorusso
© imec 2009 19

OPC Model Calibration and Verification:
2D Model Verification – Brick Wall
Accurate modeling of Brick Walls
2009 International Symposium of Extreme Ultraviolet Lithography
Brick Wall 32nm
Gian Francesco Lorusso
© imec 2009 20

OPC Model Calibration and Verification:
2D Model Verification – T End-of-Line
Accurate modeling of T-End-of-Lines
2009 International Symposium of Extreme Ultraviolet Lithography
Gian Francesco Lorusso
© imec 2009 21

Conclusions
• Full resist model
– The model was calibrated for ShinEtsu SEVR 59
– The model accurately predicts Bossung tilt and focus shift
– The model validation through flare, CD and pitch demonstrated an RMS ~ 1nm.
– The error drops to RMS ~ 0.3nm by using dose, focus, shadowing and mask CD
for 32nm DL.
• OPC
– Our goal is to test/build the OPC infrastructure that need to be put in place
before the introduction of EUVL in production
– High quality calibration and verification data sets for CD and flare were
generated
– We demonstrated the ability to compute an accurate flare map require for EUV
OPC
– We demonstrated an RMS ~ 1nm in terms of CD for various 1D and 2D structures
– Verification of 2D features was successful
2009 International Symposium of Extreme Ultraviolet Lithography
Gian Francesco Lorusso
© imec 2009 22

Acknowledgements
• This work would not have been possible without the following
contributions:
– ASML ADT team at IMEC and Veldhoven
– Philips for providing source support
– TEL for providing track support
– Hitachi for providing CD-SEM support
– Resist vendors for providing resists
– IMEC colleagues and IIAP partners for their contributions
2009 International Symposium of Extreme Ultraviolet Lithography
Gian Francesco Lorusso
© imec 2009 23

2009 International Symposium of Extreme Ultraviolet Lithography
Gian Francesco Lorusso
© imec 2009 24