Hi Felix

1. I don’t know how to assess its smoothness.  The military
specifications it adheres to don’t seem to quantify this either.  I
agree the compliance should help; I can see the two surfaces come into
contact after pressure.

2. We use a Harrick Plasma Cleaner (PDC-001) at 30W for 20-30 seconds
for bonding PDMS to glass.  We activate both surfaces at the same
time.  I replicated this protocol when I attempted bonding to MgF
instead of glass.  It is air plasma; we also have access to oxygen
plasma and could try this at higher powers.

3. With PDMS to glass the surfaces wick together after a point
pressure.  With the PDMS to MgF I needed to apply some pressure by
hand for the surfaces to meet.  After this they appeared to remain in
contact.  Only after testing the bond did I find that there was none.
Typically I apply no pressure after bringing the surfaces into contact
and bake at 65C for at least 5 minutes.

4. I think I’ve read the plasma treatment creates silanols on both
glass and PDMS and these condense to form the bond.  The plasma
cleaner in our nanolab also offers carbon tetrafluoride.  Perhaps I
could try this.

Daniel

>
> ---------- Forwarded message ----------
> From: Felix Lu 
> To: General MEMS discussion 
> Date: Tue, 29 Jun 2010 09:28:54 -0400
> Subject: Re: [mems-talk] Plasma activation of Magnesium Fluoride coated
surfaces
> Hi Daniel,
>
>   That's an interesting question I never got to explore in my plasma bonding
investigations for Si and oxidized Si.
>
> A few questions:
>
> 1. How flat and smooth is your magnesium fluoride surface? I'm not sure what
the requirements would be since your PDMS is probably somewhat complaint.
Presumably, the compliance would make things easier...
>
> 2. What plasma (and plasma chamber type?) are you using to activate the MgF
surface? how long are you activating the surface for? (for me, 30-60 s @ 150 W
in an RIE parallel plate reactor at "low" pressures seemed close to optimal);
did you try activating both bonding surfaces? What kind of plasma environment
are you using? It has been shown that with an oxygen plasma for Si-Si or Si-
SiOx, an RCA-1 dip after bonding would significantly enhance the bond strength.
For argon or nitrogen plasmas, this was not the case. Why is this is not clear
to me (maybe there's been some literature published in the last 6-7 years or so
that I've missed), but you can probably make educated guesses.
>
> 3. How much force are you applying to the bonded pair? what temperature are
you baking at while pushing them together?
>
> 4. I've always wondered if the activation by plasma was purely physical,
chemical or a bit of both...there's a line of thinking speculating that water
combined with a "damaged" surface from the plasma enhances local, low
temperature oxidation to fill in small crevices and gaps - increasing the
contact surface - and thereby increasing the bond strength. With MgF, I am not
sure what the model would be? Perhaps you need a plasma with some fluorine in
it?
>
> Felix