1
FLCC Seminar
Title: Effects of CMP Slurry
Chemistry on Agglomeration of
Alumina Particles and
Copper Surface Hardness
Faculty: Jan B. Talbot
Student: Robin Ihnfeldt
Department: Chemical Engineering
University: University of California,
San Diego
March 19, 2007
CMP
FLCC
2
Introduction
Integrated Circuit manufacturing requires material removal and
global planarity of wafer surface – Chemical Mechanical
Planarization (CMP)
–CMP slurries provide material removal by:
• Mechanical abrasion
–Nanometer sized abrasive particles
(alumina)
• Chemical reaction
–Chemical additives (glycine, H2O2,
etc.)
–Material Removal Rate (MRR) is affected
by:
•Abrasive size and size distribution
•Wafer surface hardness
–Cu is the interconnect of choice- our
research focus
March 19, 2007
CMP
FLCC
3
CMP Schematic
P = 1.5-13 psi
V= 20-90 rpm
wafer carrier
slurry
(100-300 ml/min)
polishing pad
(polyurethane)
wafer
platen
Cu MRR= 50 - 600 nm/min
Planarization time = 1- 3 min
RMS roughness = < 1 nm
wafer
Particle concentration = 1 - 30 wt%
Particle size = 50 - 1000 nm dia
slurry
polishing pad
March 19, 2007
CMP
FLCC
4
Motivation
– Better process control
• Understand role of slurry chemistry (additives, pH, etc.)
• Develop slurries to provide adequate removal rates and global
planarity
– Prediction of material removal rates (MRR)
• Predictive CMP models - optimize process consumables
• Improve understanding of effects of CMP variables
• Reduce cost of CMP
– Reduce defects
• Control of abrasive particle size
• Control of interactions between the wafer surface and the slurry
March 19, 2007
CMP
FLCC
5
Research Approach
• Experimental study of colloidal behavior of CMP
slurries
– Zeta potential and particle size distribution measurements
• Function of pH, ionic strength, additives
– Alumina particles in presence of common Cu CMP additives
– Alumina particles in presence of copper nanoparticles
• Measurement of surface hardness as function of slurry
chemistry
• Develop comprehensive model (Lou & Dornfeld, IEEE, 2003)
– Mechanical effects (Dornfeld et al., UCB)
– Electrochemical effects (Doyle et al., UCB)
– Colloidal effects (Talbot et al., UCSD)
March 19, 2007
CMP
FLCC
6
Common Cu Slurry Additives
Additives
Name
Concentration
Buffering agent
NH4OH, KOH, HNO3
bulk pH 3-8
Complexing agent - Glycine,
bind with partial or fully Ethylene-diamine-tetra-acetate
charged species in solution
(EDTA), citric acid
0.01-0.1M
Corrosion inhibitor - Benzotriazole (BTA)
protect the wafer surface 3-amino-triazole (ATA) KI
0.01-1wt%
Oxidizer
0-2 wt%
by
controlling
passive
etching or corrosion
-
cause
growth of oxide film
Surfactant -
increase
the solubility of surface
and compounds
H2O2, KIO3, K3Fe(CN)
citric acid
Sodium-dodecyl-sulfate (SDS), 1-20 mM
cetyltrimethyl-ammoniumbromide (CTAB)
Robin Ihnfeldt and J.B. Talbot. J. Electrochem. Soc., 153, G948 (2006).
Tanuja Gopal and J.B. Talbot. J. Electrochem. Soc., 153, G622 (2006).
March 19, 2007
CMP
FLCC
7
Cu CMP Chemical Reactions
Dissolution:
Cu(s) + HL  CuL+(aq) + H+ + e
CuL+, Cu2+, Cu+
Oxidation:
2Cu + H2O  Cu2O + 2H+ + 2e
CuO, Cu2O, CuL2
Cu
Oxide dissolution:
Cu2O + 3H2O  2CuO22- + 6H+ + 2e
Complexation (to enhance solubility)
Cu2+ + HL  CuL+ + H+
March 19, 2007
CMP
FLCC
8
Chemical Phenomena
Chemistry of Glycine-Water System
copper-water system
copper-water-glycine system
[CuT]=10-5M
[LT]=10-1M, [CuT]=10-5M
Ref.: Pourbaix (1957); (Aksu and Doyle (2002)
March 19, 2007
CMP
FLCC
9
Colloidal Aspects of CMP
1) Particle – particle
2) Particle – surface
3) Particle – dissolution product
4) Surface – dissolution product
Abrasive particle
Dissolution product
Surface
March 19, 2007
CMP
FLCC
10
Experimental Procedure
Slurry Abrasives
•40 wt% a-alumina slurry (from Cabot Corp.)
•150nm average aggregate diameter – 20nm primary particle diameter
Common Copper CMP Slurry Additives
•Glycine, EDTA, H2O2, BTA, SDS
Copper nano-particles
•Added 0.12 mM to simulate removal of copper surface during CMP
•<100 nm in diameter (from Aldrich)
Zeta Potential and Agglomerate Size Distribution
•Brookhaven ZetaPlus
– Zeta Potential – Electrophoretic light scattering technique (±2%)
– Agglomerate Size – Quasi-elastic light scattering (QELS) technique (±1%)
•All samples diluted to 0.05 wt% in a 1 mM KNO3 solution
•Solution pH adjusted using KOH and HNO3 and ultrasonicated for 5
min prior to measuring
March 19, 2007
CMP
FLCC
11
Electrical Double Layer
Diffuse Layer
Shear Plane
+
 2000 F 2 I 

  
  o r RT 
+
+
1
2
I   ci zi = ionic strength
2 i
Zeta Potential
  u / 
+
++
+
•Potential at surface usually
stems from adsorption of
lattice ions, H+ or OH-
+
+ ++
a +
+
+
+
+ ++
+
1/ 2
+
Potential
Particle Surface
+
+
+
+
•Slurries are stable when all
particles carry same charge;
electrical repulsion
overcomes van de Waals
attractive forces
•If potentials are near zero,
abrasive particles may
agglomerate

1/
March 19, 2007
•Potential is highly sensitive
to chemistry of slurry
CMP
Distance
FLCC
12
Zeta Potential
Zeta Potential - Potential at the Stern Layer
Electrophoresis – Zeta potential estimated by applying electric field and
measuring particle velocity
M-OH + OH- → M-O- + H2O
M-OH + H+ → M-OH2+
Surface charge on metal oxides is
pH dependant:
IEP at  = 0
Slurries are stable
when | | > 25 mV
4000
60
3000
40
20
2000
0
-20
2
4
6
8
10
12
1000
-40
-60
g Agglomerate Size (nm)
•
•
t Zeta Potential (mV)
80
0
pH
Cabot alumina without additives in 10-3M KNO3 solution (bars
indicate standard deviation of agglomerate size
distribution)
March 19, 2007
CMP
FLCC
13
Zeta Potential
Cabot alumina in 10-3M KNO3 solution with and without 0.12mM copper
Zeta Potential (mV)
80
Without Cu
With Cu
40
0
0
2
4
6
10
12
-40
pH
-80
•
•
•
•
8
IEP ~6.5 with and without copper
IEP~9.2 for a-alumina from literature*
Impurities (NO3-, SO42-, etc.) may lower IEP**
At high pH values magnitude of zeta potential lower with copper than without
*M.R. Oliver, Chemical-Mechanical Planarization of Semiconductor Material, Springer-Verlag, Berlin (2004).
**G.A. Parks, Chem. Tevs., 65, 177 (1965).
March 19, 2007
CMP
FLCC
14
Agglomerate Size Distribution
Cabot alumina dispersion in 1mM KNO3 solution with (red) and without (blue)
0.12 mM copper and without chemical additives
40
pH 2
% in solution
% in solution
30
20
10
30
20
10
0
0
0
2
0
4
5
10
Agglomerate size (mm)
Agglomerate size (mm)
•
•
pH 7
pH 2 – presence of copper causes decrease in agglomeration
pH 7 – presence of copper causes increase in agglomeration
March 19, 2007
CMP
FLCC
15
Copper-Alumina-Water System
Potential-pH for
Copper-water
System
[Cu]=10-4M at 250C and
1atm (M. Pourbaix 1957)
IEP of CuO ~ 9.5*
■ Agglomeration behavior is consistent
with the Pourbaix diagram
Average agglomerate size of bimodal
distributions in a 1 mM KNO3 solution
pH
Without Copper
With Copper
Small
Large
Small
Large
Possible State of Copper
Average (nm) Average (nm) Average (nm) Average (nm)
2
7
Cu, Cu+
Cu, Cu2O, CuO, Cu(OH)2
170
580
5000
3300
160
1700
810
9400
10
Cu, Cu2O, CuO, Cu(OH)2
150
720
300
1600
*G.A. Parks, Chem. Tevs., 65, 177 (1965).
Robin Ihnfeldt and J.B. Talbot. J. Electrochem. Soc., 153, G948 (2006).
March 19, 2007
CMP
FLCC
16
Zeta Potential
Cabot alumina in 0.1M glycine and 10-3M KNO3 solution with and without
0.12mM copper
Zeta Potential (mV)
80
Without Cu
With Cu
40
0
0
2
4
8
10
12
-40
-80
•
•
6
pH
IEP ~6.5 without copper
IEP~9.2 increased with copper
*M.R. Oliver, Chemical-Mechanical Planarization of Semiconductor Material, Springer-Verlag, Berlin (2004).
**G.A. Parks, Chem. Tevs., 65, 177 (1965).
March 19, 2007
CMP
FLCC
17
Copper-Glycine-Water System
Potential-pH for CopperGlycine-Water System*
1
CuL+
0.6
[Cu]=10-4M, [Glycine]=10-1M at 250C
and 1atm
CuHL2+
0.4
E, V vs. SHE
•
Cu2+
0.8
Agglomeration behavior is
consistent with Pourbaix diagram
CuO22CuL2
CuO/
Cu(OH) 2
0.2
CuL2-
0
-0.2
Cu2O
Cu
-0.4
-0.6
Average agglomerate size of bimodal
distributions in a 1 mM KNO3 solution with
various additives
-0.8
-1
0
2
4
6 pH 8
10
12
14
Without Copper
With Copper
Small
Large
Small Average
Large
Average (nm) Average (nm)
(nm)
Average (nm)
Solution
pH
Possible State of
Copper
Cu, CuHL2+
Cu, CuL2
310
2000
8100
220
1900
700
0.1M Glycine
2
7
10
Cu, CuL2-, CuL2
1030
6300
350
2100
2
7
Cu, CuHL2+
Cu, CuL2
130
1800
300
163
1400
10
Cu, CuL2-, CuL2
1200
0.1M
Glycine+2.0wt%
H2 O 2
*S. Aksu and F. M. Doyle, J. Electrochemical Soc., 148, 1, B51 (2006).
March 19, 2007
16
CMP
1500
FLCC
18
Measuring Wafer Hardness
TriboScope Nanomechanical Testing
system from Hysitron Inc.
•1 cm2 silicon wafer pieces
sputter deposited with 30 nm
Ta + 1000 nm Cu
•10 min exposure in 100 ml of
slurry solution (without
abrasives), then removed and
dried with air and measured
■ Considerations
– Large applied load will increase indentation depth –
• more likely for underlying layer to affect nanohardness measurements
– Slurry solutions with high etch rates will decrease copper thickness –
• thinner copper layer more likely for underlying layer to affect
measurements
Robin Ihnfeldt and J.B. Talbot. 210th Meeting Electrochem. Soc., Cancun, Mexico, Oct. 29-Nov. 3, 602, 1147 (2006).
March 19, 2007
CMP
FLCC
19
Copper Surface in Solution
Bulk metallic Cu H~ 2.3 GPa*
Ta2O5 H~9 GPa
Surface nanohardness of Cu on Ta/Si (100uN applied load) after exposure
to 1mM KNO3 solution
pH
Possible State of Copper
Contact
Depth (nm)
Etch Rate
(nm/min)
Copper
Nanohardness
Thickness (nm)
(GPa)
2
7
Cu, Cu+, Cu2+
Cu, Cu2O, CuO, Cu(OH)2
30
50
16
9
810
870
2.9
1.2
12
Cu, Cu2O, CuO, Cu(OH)2
43
0
960
1.2
■ pH 2 – appears that state of surface is Cu metal with increase in nanohardness from
underlying layer
■ pH 7 and 12 – hardness less than that of bulk metallic Cu
–
Cupric hydroxide, Cu(OH)2, is most likely forming
*S. Chang, T. Chang, and Y. Lee, J. Electrochemical Soc., 152, (10), C657 (2005).
March 19, 2007
CMP
FLCC
20
Copper Surface in Solution
Surface nanohardness of Cu on Ta/Si (100uN applied load) after exposure
to 1mM KNO3 solution and other additives
Chemistry
pH
Possible State of
Copper
0.1M Glycine
2
7
Cu, CuHL
Cu, CuL2
41
62
0
6
960
870
1.3
2.5
12
Cu, Cu2O, CuL2
81
14
780
0.5
2
7
Cu, CuHL
Cu, CuL2
135
122
22
55
640
330
0.8
1.1
12
Cu, Cu2O, CuL2
48
-7
1020
3.6
0.1M Glycine +
2.0wt% H2O2
2+
2+
Contact
Depth (nm)
Etch Rate
Copper
Nanohardness
(nm/min) Thickness (nm)
(GPa)
Film Growth
Glycine
• Surface hardness is less than that of bulk Cu at pH 2 and 12 –
Increased
Hardness
– Glycine may interact with surface layer to decrease compactness
• pH 7 appears to be Cu metal with increase due to underlying layer
Glycine + H2O2
• H2O2 increases solubility of Cu-glycinate complex or increases Cu oxidation
• Surface is less than bulk Cu at pH 2 and 7 – decrease in compactness due to glycine
• pH 12 appears to be cuprous oxide, Cu2O
March 19, 2007
CMP
FLCC
21
CMP Experiments
Toyoda Polishing apparatus
(UC Berkeley)
– IC1000 polishing pad preconditioned for 20 minutes
with diamond conditioner
– Polished 2 min with Cabot
alumina
Silicon wafers (100 mm dia.) with 1 mm copper on 30 nm tantalum
– Total of 18 wafers polished with various slurry chemistries and at
various pH values
March 19, 2007
CMP
FLCC
22
Experimental Copper CMP MRR
No additives
20
0.1M Glycine
MRR (nm/min)
15
0.01M EDTA, 0.01wt% BTA,
1mM SDS, 0.1wt% H2O2
MRR is <20 nm/min for all pH values
without additives, with 0.1M glycine
10
5
0
4
6
pH
8
10
12
0.1M Glycine, 0.1wt% H2O2
0.1M Glycine, 2wt% H2O2
400
MRR is >100 nm/min for
several pH values where both
glycine and H202 are present
MRR (nm/min)
2
0.1M Glycine, 0.01wt% BTA,
1mM SDS, 0.1wt% H2O2
300
200
100
0
2
March 19, 2007
CMP
4
6
pH
8
FLCC
10
12
23
Lou and Dornfeld CMP Model
Basic Eqn. of Material Removal: MRR = N x Vol
Vol
N
Force F & Velocity
Slurry
Concentration C
Active Abrasive
Size Xact
Average Abrasive
Size Xavg
Wafer hardness Hw/
Slurry Chemicals &
Wafer Materials
Proportion of
Active Abrasives
March 19, 2007
CMP
FLCC
24
Conclusions
Colloidal Behavior
• pH has greatest effect on colloidal behavior
• Glycine acts as a stabilizing agent for alumina
• Presence of Cu nanoparticles can increase or decrease
agglomeration depending on the state of copper in solution
• Agglomeration behavior with copper is consistent with potentialpH diagrams
Nanohardness of Copper Surface
• pH of the slurry affects copper surface hardness
• Addition of chemical additives has large effect on the surface
hardness
• State of copper on surface is consistent with potential-pH
diagrams
• Under certain conditions glycine may cause decrease in copper
surface hardness
March 19, 2007
CMP
FLCC
25
Future Work
• Continue to investigate effect of copper on zeta potential and particle
size
– Determine state of Cu in solution
– Study agglomeration as a function of time
• Initial hardness measurements show large differences in copper
surface with pH and chemical addition
– Determine reproducibility of hardness measurements
– Determine state of Cu on surface
• Modeling – Luo and Dornfeld Model*
– Incorporate experimental measurements (hardness and
agglomerate size distribution) into model and compare with
experimental CMP data
*J. Luo and D. Dornfeld, IEEE Trans. Semi. Manuf., 14, 112 (2001).
March 19, 2007
CMP
FLCC
26
Acknowledgments
• Funded by FLCC Consortium through a UC Discovery grant. We
gratefully acknowledge the companies involved in the UC
Discovery grant: Advanced Micro Devices, Applied Materials,
Atmel, Cadence, Canon, Cymer, DuPont, Ebara, Intel, KLATencor, Mentor Graphics, Nikon Research, Novellus Systems,
Panoramic Technologies, Photronics, Synopsis, Tokyo Electron
• Prof. Dornfeld and his research group at UC Berkeley for use of
the CMP apparatus and model program
• Prof. Talke and his research group at UCSD for the use of the
Hysitron Instrument.
March 19, 2007
CMP
FLCC