Friction%20Updates

1
Friction Updates

• Last FtF
• Past issues solutions
• Latest data
• CoF vs. slurry dilution
• CoF vs. rotation rate
• Fz vs. slurry dilution
• Laser displacement sensor tests
• Feasibility
• Specifications
• Thesis Roadmap
• Microscale friction and polishing model
• Macroscale measurements
• Questions/Requests
• Future work

Jim Vlahakis PhD. Candidate Tufts University
2
Friction Updates Last FtF

• External noise issues
• New wafer motor power supply eliminated noise
• Issues concerning slurry dilution
• Moved from 91 to 32 dilution, more in line with
  industry practice

• Data runs at 30/60rpm
• Now using 60/120rpm (.5 1m/s), more in line
  with industry practice
• Irregularities due to changes in wafer shape
• Controls in place to monitor wafer shape

3
Friction Updates Latest Data

• CoF vs. slurry dilution
• 60rpm 1.7psi
• Note large CoF (and large s) for pure slurry. A
  result of shear thickening?
• CoF remains fairly constant over a wide range of
  slurry dilutions
• CoF for pure H2O seems strongly dependent on pH
• Next steps
• study slurry compositions in the range of 1 part
  H2O to 2/3/5/7 parts slurry
• Investigate pH dependence for H2O
• Perform experiment backwards

4
Friction Updates Latest Data

• CoF vs. rotation rate, note development of CoF
  plots and spectra

5
Friction Updates Latest Data

• CoF vs. rotation rate, note development of CoF
  plots and spectra

6
Friction Updates Latest Data

• Fx vs. slurry dilution
• 60rpm - 1.7psi
• Decrease in Fz tracks decrease in density of
  slurry as it is mixed with H20
• Increasing s corresponds to increasing chatter
  (except for pure slurry case)
• Next steps
• study slurry compositions in the range of 1 part
  H2O to 2/3/5/7 parts slurry
• Investigate pH dependence for H2O

7
Friction Updates Latest Data

• CoF vs. rotation rate
• 60rpm - 1.7psi
• CoF remains fairly constant over a wide range of
  velocities
• Below 30rpm data is unreliable motion of wafer
  drive is noticeably non-uniform
• Next step
• Investigating deeper into the Stribeck curve will
  require lower pressures and much higher
  velocities. Likely not possible with our current
  setup

8
Friction Updates Laser Sensor

• Feasibility of using multiple laser sensors to
  monitor wafer displacement
• How much does support frame displace during worst
  case polishing
• 60rpm2.5psipure H20 (strong chatter)
• Frame displacements are on the order of .1mm
• Spectrum of displacement signal is very familiar
• Safe to say that wafer motion wont be lost in
  frame motion

9
Friction Updates Laser Sensor

• Generate rough idea of wafer displacement
• 60rpm-1.7psi-32 (minimal chatter)
• range /- 1mm
• Wafer regularly moving out of range
• To be safe, we should require a range of at least
  /- 2mm or even /-5mm
• Must design a system that allows for quick
  adjustment of sensor position
• Given the accuracy of the laser (1µ), is this a
  good way to measure fluid film thickness?
• Sensor size is an important consideration

10
Friction Updates Thesis Roadmap

• Microscale Changes and Their Effect on Macroscale
  Process Variables in Chemical Mechanical
  Planarization
• Submitted by
• James Vlahakis
• In Partial Fulfillment of the Requirements for a
  PhD.
• School of Engineering
• Tufts University
• May 2008

• Develop a micro-friction model utilizing the
  expertise of the BostonCMP group
• Identify the processes that contribute to CoF
  and/or MRR
• Examples
• CoF µfluidµpadµparticles
• Polishing is more a plucking (chemical)
  process, rather than a plowing (mechanical)
  process
• Make changes on the microscale and measure their
  effect on the macroscale, for example
• Vary roughness and moduli of pads
• Various particle loads
• Maintain constant particle loading, while varying
  the load of particles that participate in the
  polishing
• Measure effect on both CoF and MRR
• Which changes effect CoF or MRR only
• Which effect both
• Does this confirm our model or do we need to make
  adjustments?

11
Friction Updates - Questions/Requests

• Questions
• How can we measure MRR?
• Can we etch a radial channel in our wafers and
  measure the change in step heights?
• Can we replace slurry silica particles with
  particles that do not participate in the
  chemistry?

• Requests
• More slurry please!
• Pads of various roughness and modulus

12
Friction Updates Future Work

• High Level View
• Continue to develop models with Boston CMP group
• Identify the experiments that will show the
  accuracy (or inaccuracy) of our models
• Perform the experiments and, if necessary,
  iterate until our models can explain our data
• Write it up!

• Low Level View
• Complete CoF vs. slurry dilution experiments
• Determine µ and shear thickening effects for
  various slurry dilutions
• Purchase and begin installation of laser
  displacement package
• Prepare for various conferences and papers