![Courtesy of Steven Engineering, Inc. - 230 Ryan Way, South ... · 2 x 0.5 mm2 403 PTDA series ... Rated surge voltage [kV] 2.5 2.5 2.5 2.5 2.5 2.5 Approval data (UL/CUL) Use Group](https://static.fdocuments.net/doc/165x107/5c78029809d3f21d538c775a/courtesy-of-steven-engineering-inc-230-ryan-way-south-2-x-05-mm2-403.jpg)
Advanced Anneal Solutions for Silicides · 0.5 1.0 1.5 2.0 2.5 Normalized Ion 0.5 1.0 1.5 2.0 2.5...
16
Advanced Anneal Solutions for Silicides Shankar Muthukrishnan RTP Anneals GPM July 15 th 2010
Transcript of Advanced Anneal Solutions for Silicides · 0.5 1.0 1.5 2.0 2.5 Normalized Ion 0.5 1.0 1.5 2.0 2.5...
Advanced Anneal Solutions for Silicides
Shankar Muthukrishnan
RTP Anneals GPM
July 15th 2010
2
Challenges Scaling Nickel Silicide Contacts
Improved silicide contact is key to scaling
45nm32nm
shrinking
increases risk
of leakage
!yield inhibiting
Diffusion control for thinner layer formation
Silicide-Silicon interface morphology
Avoiding yield limiting piping defects, especially with proximity to SDE, channel
Sensitivity to agglomeration
3
Integration Alternatives to Suppress NiSi Piping
Pre amorphous implant (PAI)
Pre clean using “flourine” dry etching
Co implant dopant in substrate (F, C)
Nickel alloy (Pt)
Modifying anneals
4
Nickel-Silicide Process Steps
NiPt Salicide Flow:
pre clean (native oxide removal)
metal deposition (NiPt/TiN)
RTP-1 anneal : Ni diffusion into Si creating Ni2Si
RTP-2 anneal : transformation into low resistance NiSi
non-reacted (NiPt) metal removal by selective etching
280-300ºC Soak
400-500ºC Soak
Traditional
anneals
Advanced
alternative
200-250ºC Soak
~850ºC millisecond
5
Soak to Millisecond Transition for Silicides
Driven by
Enabling thinner silicides without comprising conductivity or yield
Millisecond Provides
Improved Yield performance
Improved Junction Leakage performance
Fewer leakage-inducing defects through steep temperature profile
Logic (65nm) Logic (45nm) Logic (32nm)
Ni Silicidation / Diffusion Soak Low temp Soak Low temp Soak
NiSi Phase
TransformationSoak
Soak or
Millisecond Millisecond
6
Nickel Silicidation Transformation Curve
0.00
5.00
10.00
15.00
20.00
25.00
30.00
700 725 750 775 800 825 850 875 900 925 950
Temperature
Rs Rs Average
Large process window available for RTA 2
7
The Ideal Time-Temperature Profile for NiSi
Key parameters to consider
Preheat : Must be low enough to avoid uncontrolled diffusion
Jump temp : Must be high enough to enable optimal activation w/o wafer breakage or agglomeration
Dwell time : Must be short to avoid wafer bow and breakage
Above profile is key to enabling optimal nickel silicide
Preheat temperature
Time
Te
mp
era
ture
Dwell time
700
800
900
1,000
100
200
300
400
500
600 Jump
temperature
8
Yield Results and Scaling Impact for NiSi Process
imec confidential 2009 24
Why it seems to be less interesting compare to 1st trials
Piping
induces
leakage(Statistical)
Large spacer = less
sensitivity
0
10
20
30
40
50
60
70
80
90
100
1.E-11 1.E-10 1.E-09 1.E-08 1.E-07 1.E-06
Leakage from Source to Drain [A/µm]
Pro
ba
bilit
y [
%]
Silicidation by Soak
Silicidation by Laser
x5000 PMOS, L=60nm
Ioff Source
0V
0V
-1V
Statistical
silicide pipe
AL090036: small spacer
0
10
20
30
40
50
60
70
80
90
100
1.E-11 1.E-10 1.E-09 1.E-08 1.E-07 1.E-06
Ioff_Source [A/µm]
Pro
ba
bilit
y [
%]
Soak/Soak
Soak/850°C (Ch150°C)
700°C
AL090131: Large spacer
Larger spacer could reduce the
sensitivity against piping:
It is the reason why the potential yield gain seems
to be less pronounce in
AL090131RTA1/RTA2
Single laser step
Scaling spacer
affects RTA2
leakage, but not
DSA2
Data source : IMEC IEDM 2009
No difference in
leakage when using
large spacer
9
Results: NiSi Morphology Improvement
Process B (lower RTP-1 Soak) roughness reduced vs process A (baseline)
Process C (lower RTP-1 Soak + DSA RTP-2) achieves lowest roughness
T1 Soak
T2 Soak
T1-40ºC Soak
T2 Soak
T1-40ºC Soak
T2 + 400ºC DSA
Source: IEEE - RTP2009 Conference, Advances on 32nm NiPt Salicide Process,
Dr. Chen et al. Co-published by UMC / Applied Materials
RTP-1 anneal:
RTP-2 anneal:
RTP-1 anneal:
RTP-2 anneal:
RTP-1 anneal:
RTP-2 anneal:
10
Results: NiSi Piping Defects Reduction
Process B (lower RTP-1 Soak) achieves lower defect count than baseline
Process C (lower RTP-1 Soak + RTP-2 DSA) achieves lowest piping defect count
0
20
40
60
80
100
60%
B
A
30%N
orm
ali
ze
d D
efe
cts
(%
)
Based on e-beam Bright Voltage Contrast (BVC) measurement count on a pattern wafer, post CMP
C
T1 Soak
T2 Soak
T1-40ºC Soak
T2 Soak
T1-40ºC Soak
T2 + 400ºC DSA
Source: IEEE - RTP2009 Conference, Advances on 32nm NiPt Salicide Process,
Dr. Chen et al. Co-published by UMC / Applied Materials
TEM Top View Piping Defects Count
RTP-1 anneal:
RTP-2 anneal:
11
Results: Electrical Performance – RS/D
The series S/D Rs (Rsd) decreased using DSA
CBACBA
8%6%
Rsd
_N
+(a
.u.)
Rsd
_P
+(a
.u.)
Source: IEEE - RTP2009 Conference, Advances on 32nm NiPt Salicide Process,
Dr. Chen et al. Co-published by UMC / Applied Materials
T1 Soak
T2 Soak
T1-40ºC Soak
T2 Soak
T1-40ºC Soak
T2 + 400ºC DSA
RTP-1 anneal:
RTP-2 anneal:
T1 Soak
T2 Soak
T1-40ºC Soak
T2 Soak
T1-40ºC Soak
T2 + 400ºC DSA
NMOS PMOS
12
Results: Electrical Performance – Ion/Ioff
Drive current gain achieved by enhancing nickel silicide contact using
millisecond annealing with low temperature RTP-1 Soak
RTP-1: Soak @ T1-40ºC
RTP-2: Millisecond (T2+400ºC)Io
ff (
nA
/m
)
NMOS gain = 4% PMOS gain = 3%
0.5 1.0 1.5 2.0 2.5
Normalized Ion
0.5 1.0 1.5 2.0 2.5
Normalized Ion
100
100
0
10
1
0.1
Source: IEEE - RTP2009 Conference, Advances on 32nm NiPt Salicide Process,
Dr. Chen et al. Co-published by UMC / Applied Materials
13
The Ideal Time-Temperature Profile for NiSi
Key parameters to consider
Preheat : Must be low enough to avoid uncontrolled diffusion
Jump temp : Must be high enough to enable optimal activation w/o wafer breakage or agglomeration
Dwell time : Must be short to avoid wafer bow and breakage
Delivered by Applied Materials’ Vantage Astra system
Applied Vantage Astra
Flash Competitor
Laser Competitor
250ºC or less
350ºC
400ºC
0.25msec
0.8msec0.4msec
Time
Te
mp
era
ture
100
700
800
900
1,000
200
300
400
500
600
Ideal Profile
14
Applied Vantage® Astra™ DSA System
Uniquely powerful silicidation
– Up to 5% greater device speed
– Higher yields enabled by up to 15x lower leakage
– Less wafer stress
Versatile dynamic millisecond anneal
– Broad range of processing conditions
– Ambient control
– Extendible to high-k/metal gate applications
Compact, reliable, cost-effective
– Simple, compact and smart chamber design
– Solid-state laser with prolonged lifetime
– > 40 WPH per two- chamber system
– Compatible with an RTP chamber on same system as hybrid
FAST. SMART. RELIABLE.
Simply Better Anneal
15
Vantage Astra DSA System Photos
![[XLS] · Web view0 0.2 0.5 1 0 0.02 0.04 0.1 0.2 0.35 0.5 0.7 0.75 1 1.5 2.5 3.7 12.5 0 0.2 0.5 1 0 0.02 0.04 0.1 0.2 0.35 0.5 0.7 0.75 1 1.5 2.5 3.7 12.5 0 0.2 0.5 1 0 0.02 0.04](https://static.fdocuments.net/doc/165x107/5af0fdb97f8b9ac2468eca80/xls-view0-02-05-1-0-002-004-01-02-035-05-07-075-1-15-25-37-125-0.jpg)
