Characterization of ChemicalMechanical Polishing Slurries using Ultrasound.

1
Characterization of Chemical-Mechanical Polishing
Slurries using Ultrasound.
Ultrasound based Techniques. Acoustics and
Electroacoustics.

• Andrei Dukhin and Philip Goetz

• DTI was established in 1996 for manufacturing
  ultrasound based instruments for characterizing
  concentrates..
• About 300 instruments in 15 countries for sizing,
  zeta potential, rheology.
• More than 500 instruments for zeta potential
  world wide.

Dispersion Technology Inc. Bedford Hills, NY
10507, USA
2
ACOUSTICS
Particle sizing in concentrates, mixtures,
structures
Rheology longitudinal, bulk viscosity,
compressibility
Aggregation, flocculation, stability
ELECTRO- ACOUSTICS
Zeta potential, surface conductivity
Titrations, iso-electric point, surfactant
dozation
No Sample preparation
Continuous on-line monitoring
3
History of Ultrasound in Colloids
4
transmitter
receiver
L
Iin
Iout
Attenuation dB/cm/MHz
Sound speed cm/sec
5
transmitter
receiver
L
Iin
Iout
Iscattered
Light measures only scattered energy
at certain angle Iscattered
Ultrasound measures total energy loss, scattered
adsorbed Iin-Iout
6
Cabot CMP silica
Starting and final attenuation spectra of the 200
continuous measurements
7
Cabot CMP silica
Median and mean size for 200 continuous
measurements.
Precision of the median and mean particle size
characterization of CMP slurry is in the range
1-3 nanometer if precision of the attenuation
measurement is 0.01 dB/cm/MHz.
8
Piezo crystal Electrodes
A
Zeta Potential Probe
ElectroAcoustics General.

• - volume fraction
• rm,p s,- densities
• Km,s - conductivities

9
Cabot CMP silica
Colloid Vibration Current and corresponding
z-potential for silica CMP 200 continuous
measurements.
Precision of the z-potential characterization is
about 0.3 mV with no special requirements for
temperature control.
10
Sensitivity to the large particles.
Dukhin, A.S. and Goetz, P.J Characterization of
Chemical Polishing Materials (monomodal and
bimodal) by means of Acoustic Spectroscopy ,
Colloids and Surfaces, 158, 343-354 (1999)
Characterization of CERIA slurry. Zeta potential.
Flocculation.
Newsletter 13 on DTI web site. 2003.
Fractal model for interpreting Acoustics of CMP
slurry.
Dukhin, A.S., Fluck, D., Goetz, P.J., Shilov,
V.N. and Dukhin, S.S. Characterization of
fractal particles using Acoustics,
Electroacoustics, Light Scattering, Image
analysis and Conductivity, Langmuir, 23, 10,
pp.5338-5351 (2007)
11
Attenuation spectra of Cabot SS25 with various
additions of Geltech 05. Total solid content
12wt.
12
Attenuation spectra of Cabot SS25 with various
additions of Geltech 05. Total solid content
12wt.
The detection limit is a single 1 micron
particle per 100,000 particles of 100 nm size.
This is equivalent to a large particle content
of 1 wt. relative to the total solid content of
the chemical mechanical polishing (CMP) slurry.
13
Block scheme of DT-1200 sample chamber
14
Ceria slurries. Stirring On.
15
Stirring Off.
16
Attenuation at 12.8 MHz for both slurries.
Stirring Off.
z-42.9 mV
z-40.9 mV
Flocculation might not be related to the
z-potential value. Attenuation time dependence is
alternative way to characterize stability.
17
Semisperse SS25 silica
18
(No Transcript)
19
Aggregate size is independent on fractal number.
Fractal number can be determined when primary
size is known.
20
Volume fraction and aggregate density are
independent on fractal number.
21

• Precision of the median particle size
  characterization of CMP slurry is 1-3 nanometer.
• Precision of the z-potential characterization is
  about 0.3 mV.
• The detection limit of the large particles is a
  single 1 micron particle per 100,000 particles of
  100 nm size.
• Absence of the time dependence of the measured
  attenuation and/or CVI indicates that CMP slurry
  is stable and does not contain large particulate
  entities.
• Time dependence of attenuation spectra with no
  stirring indicates either flocculation or
  aggregation. It is possible to distinguish
  between them.
• Decay of attenuation spectra with time at the
  same rate for all frequencies (parallel shift)
  indicates flocculation.
• Flocculation might not be related to the
  z-potential value. Attenuation time dependence is
  the alternative way to characterize stability.
• Decay of attenuation spectra with time at the
  different frequency dependent rate (assymetric
  shift) indicates aggregation.
• Fractal model yields accurate particle size,
  volume fraction and aggregates density.

22
Potential problems for on-line monitoring.
Volume fraction is required for calculating
particle size and z-potential. Variation of
volume fraction would cause variation of size and
z.
Bubbles. Large bubbles are not a problem. Small
ones with sizes under 100 micron are.
Mixed composition. Software can d mixtures, but
this requires special studies.
Maintenance. Contamination of sensor services is
not problem for Acoustics. It is problem for
Electroacoustics.
Calibration. Sizing does not need calibration, z
does.
Troubleshooting. Measurement of water is
sufficient for testing sensors and electronics.
Design. Customer designs and maintains pluming
and other aspects of conecting sensors to the
stream.