SpecMat Response to PDMS Module Devlopment

[Ed Myers]

Prof. Nishi,

The question about running AFM Associates PDMS process in the facility was 
discussed at yesterday's SpecMat meeting.  SpecMat came to the following 
conclusions.

If the PDMS starting components are electronic grade, the process is 
approved from a chemical contamination point.  However, we are very 
reluctant to approve the process due to the concerns over mechanical 
contamination of the equipment set and from concern over process capability 
of the Nikon to resolve the 0.7um feature size.

The request process flow is:
Deposit and cure 500um thick layer of PDMS
Deposit 0.5um of Aluminum using Gryphon
Apply photoresist
Expose using Nikon stepper (CD = 0.7 um)
Develop
Timed etch of aluminum in P5000, down to PDMS layer
Resist strip

SpecMat's specific process concerns are as follows:
1) The PDMS must be electronic grade and must be mixed and out gassed 
outside of the fab.
2) The ability to dispense and cure the PDMS layer with sufficient flatness 
to resolve 0.7um features.
3) Complete backside PDMS edge bead removal, compatible with resolving 
0.7um features.
4) Deposition of the Al without any peeling.  There is concern the 
deposition could provide sufficient energy to the substrate which will 
result in flaking of the Al in the Gryphon deposition chamber.  If this 
occurs we will be looking at significant downtime (days) to clean and 
qualify the deposition chamber.
5) The ability of the Nikon to focus on the thick, transparent PDMS 
layer.  Someone skilled with the Nikon will most likely need to run the 
samples to adjust the focus depth outside of standard 
tolerances.  Subsequent to running the PDMS samples, the Nikon's focus will 
need to be calibrated to standard silicon substrate thickness.
6) The P5000 has a high density plasma which can significantly heat the 
substrate.  There is concern the heating will cause a delamination problem 
between the Al and PDMS.  If the wafers flake in the etch chamber we will 
lose a day to clean the chamber.  Hopefully, any particulation will not be 
carried in to the transfer module or load locks.

There has been a number of emails exchanged recommending alternative 
approaches to process development such as metal deposition in the Innotec 
to minimize substrate heating during metal deposition (and easier cleaning 
if the wafer peels and test the exposure) and starting with thinner 
PDMS.  Continued discussion on an approach to the process development is 
necessary to reach a comfort zone where we can minimize our equipment exposure.

Regards,
Ed