Hello MEMStalk,

I'm currently using a Materials Research Corporation parallel-plate RIE tool
to etch 100nm of thermally-grown SiO2 with a nanoporous block copolymer film
as the etch mask.  The polymer film is ~40nm thick and is composed of
UV-crosslinked polystyrene (no idea how crosslinked - I can't dissolve it
completely to run it through a GPC; it was exposed to 25J/cm^2 of 254nm
light) with vertical cylindrical pores acting as the etch windows.  The
pores are ~20nm in diameter.  The etch profile needs to be as vertical as
possible (desired aspect ratio = 5:1).  I'm using 20sccm CHF3 @ 2mT and
500V/75W RIE power to etch the oxide.  The unmasked oxide etch rate (i.e.
the etch rate of clean, uncovered thermally grown SiO2 - same stuff as
what's under the polymer film) is 16nm/min under these conditions.  My
problem is that the polymer film is not robust enough and is removed before
the etch is complete.

Does anyone have any suggestions as to how I might improve the etch
resistance of the polymer film or improve selectivity?  Maybe some treatment
for the film itself (like further crosslinking), or modifications to the
etching process?  I have the following gases available on the RIE tool: O2,
Ar, H2, CF4, CHF3 and Ar.  I also have access to an ICP RIE tool with the
following gases: Cl2, BCl3, CF4, CHF3, SF6, Ar, N2, He and O2.  I had
thought about depositing ~10-20nm of Cr on the oxide before depositing the
polymer film, transferring the pattern to the Cr, using the patterned Cr
film as the mask for SiO2 etching, but it would be nice if I could avoid the
additional processing steps.

Thanks,

Brian

--
Brian C. Stahl
Graduate Student Researcher
UCSB Materials Research Laboratory
[email protected] / [email protected]
Cell: (805) 748-5839
Office: MRL 3117A