Ebara Corporation - Slide 1
Manabu Tsujimura, David Watts1
Ebara Corporation, 1Ebara Technologies Inc.
Analysis and Prediction of Polish ProfilingAnalysis and Prediction of Polish Profiling
Ebara Corporation - Slide 2
Year of First Shipment 1999 2002 2005 2008 2011Technology Generation(nm) 180 130 100 70 50Planarity(nm) 18 14 11 9 7Aspec t r a t i o (Cu) 1. 4 1. 5 1 . 7 1. 9 2 . 1Metal effective resistivity(µ O -cm) 2. 2 2. 2 2 . 2 1. 8 <1. 8I LD effecteve dielectr ic constant(k) 3. 5- 4. 0 2. 7- 3. 5 1 . 6 - 2 . 2 0. 5 <0. 5Barr ier th ickness (nm) 17 13 10 0 0Wafer size (mm) 200 300 300 300 450
The “super planarity” requirement by ITRS demands that ALLALL CMP tools must have
““ProPro--ActiveActive”” Profile ControlProfile Control
What is the Message From ITRS ?What is the Message From ITRS ?
Brick WallBrick Wall
Ebara Corporation - Slide 3
What is What is ““ProPro--Active Profile ControlActive Profile Control””??
Wafer
Uniformdeposition
Wafer
Non-uniformdeposition
Wafer
Uniformpolish
Wafer
Non-uniformpolish
ProPro--Active profile control is the ability to produce Active profile control is the ability to produce uniform polish profiles regardless of the incoming uniform polish profiles regardless of the incoming wafer deposition profile.wafer deposition profile.
Proactive (pro ak’tiv), adj. Serving to prepare for, intervene in,or control an expected occurrence or situation
Ebara Corporation - Slide 4
Examples of Six Different Cu Deposition ProfilesExamples of Six Different Cu Deposition Profiles
+Anode
Chemical
Contact
Wafer
-
0.2-µ m dia. AR=4.50
200
400600
800
1000
12001400
0 5 10 15 20Distance from edge (mm)
Cu
thic
knes
s (n
m)
33%66%100%33%66%100%
PV= 200nm
5mm ==>? =5x4=20mm
Edge ProEdge Pro--profile control is required.profile control is required.
Ebara Corporation - Slide 5
Wafer carriercenter profile
Several Methods to Improve CMP PerformanceSeveral Methods to Improve CMP Performance
Back side pressureRetainerpressure
Uniform pressure distributionUniform relative speed
Uniform conditioning
Uniform slurry supply
Uniform movement
Uniform temperature
Wafer carrier edge profile
Ebara Corporation - Slide 6
ProPro--Active Profile Control for CLC & APCActive Profile Control for CLC & APC
Planarity requirement- Step height of ILD- Dishing & Erosion- Pro-active control
of polish profile
Best profileInitial polishprofile
Change of consumables
BSP press.
RR press.
Head center profile
Head edge profile
Center profile
Edge profile
Parameters to control profile
Slurry supplyDressing
Linear speedOthers
Consumables-Pad-Slurry-Backing film
Ebara Corporation - Slide 7
A Polish profile is predicted by the following equation, based on Preston’s equation;
q (r)= k0 (r) k1 (r)k2 (r) k3 (r) k4 (r) p (r) v (r)where,
q (r): Polish quantityk0 (r): Basic proportional co-efficientk1 (r): Back-side pressure co-efficientk2 (r): Retainer-ring pressure co-efficientk3 (r) : Head center-profile co-efficientk4 (r): Head edge profile co-efficient
p (r):Pressure on polished waferv (r):Linear velocity
Basic Calculation EquationsBasic Calculation Equations
Ebara Corporation - Slide 8
The analysis is performed under the following consumable set and experimental conditions:
Slurry :SS25Pad :IC1000/Suba 400Backing film :NF200Speed Carrier/Table :30/30 rpmCarrier force :500 g/cm2
Retainer-ring force :0-700 g/cm2
ILD :TEOS
V (r)=V (0)=0.5 m/secv(r)=V (r)/V (0)=1q (r)=k 0(r) Basicqq ’’((r)=r)=kk 00((r)r)??kk 11 ((r)r)??kk 22 ((r)r)??kk 33 ((r)r) ImprovedImproved
Conditions for AnalysisConditions for Analysis
Ebara Corporation - Slide 9
FEM Mesh drawing
X· Y direction restriction
X· Y direction restriction
Wafer load
Retainer ring load
Effect of Retainer Ring ForceEffect of Retainer Ring Force
Ebara Corporation - Slide 10
Wafer Position (mm)
Pol
ish
pro
file
Experimental
0 kPa
50 kPa
0.5
1.0
1.5
90 100
Wafer Position (mm)R
R f
orce
co-
eff k
2
FEM analysis
0 kPa
50 kPa
0.5
1.0
1.5
90 100
0.360.4
4%
FEM analysis follows the same trend as experimentFEM analysis follows the same trend as experiment
Effect of Retainer Ring ForceEffect of Retainer Ring Force
Ebara Corporation - Slide 11
X· Y direction restriction
X· Y direction restriction
Uniform load
FEM Mesh drawing
Convex Concave
Effect of Head Center ProfilesEffect of Head Center Profiles
Ebara Corporation - Slide 12
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
-100 -75 -50 -25 0 25 50 75 100
Pol
ish
pro
file
Wafer Position (mm)
Wafer Position (mm)
Head center effect
co-eff
k3
0.8
1.0
1.2
0 50 100
? F E M ?
? F E M ?
0.04
Effect of Head Center ProfilesEffect of Head Center Profiles
Convex
Concave
Ebara Corporation - Slide 13
Polish profile
1.0
1.2
0.8
Wafer position ( m m )-100 0 100
BSP= 0 kPaBSP= 20 kPaBSP= 30 kPa
0.8
1.0
1.2
0 50 100Wafer position ( m m )
BSP= 0 kPaBSP= 20 kPaBSP= 30 kPa
BSP Co-eff k
1
Effect of Back Side PressureEffect of Back Side Pressure
Wafer
Back side pressureWafer carrier
Pad
Table
Back side holes
P
N 2
Ebara Corporation - Slide 14
Analysis
0.190.22
80 90
0.190.25
Wafer position ( m m )
0 50100
Base profile
Improved profile
Experimental
0.5
1.0
1.5
Polish rate profile
0.5
1.0
1.5
Polish rate profile
Base profile
Improved profile
0 50 100
3 %
0 %
0.5
1.0
1.5
Polish rate profile
Analysis
Base profile
Improved profile
Wafer position ( m m )
90 95100
Experimental
Base profile
Improved profile
0.5
1.0
1.5
Polish rate profile
0.34
0.28
1mm
90 95 100
6 %
Comparison of Analysis and Experimental DataComparison of Analysis and Experimental Data
Ebara Corporation - Slide 15
1 Back-side effect co-efficient k1 has been obtained by experiment.
2 Retainer-ring force effect co-efficient k2 has been obtained by FEM analysis and has been confirmedto be highly predictable.
3 Carrier center-profile effect co-efficient k3 has been obtained by FEM analysis and confirmed to be highly predictable.
4 Carrier-edge profile effect co-efficient k4 has been obtained by experiment.
5 The analysis and presented example shows that the modelis acceptable for predicting polish profiles.
6 This type of analysis will be useful for automated CMPclosed-loop control.
CONCLUSIONCONCLUSION