This paper has some good information on bonding PDMS. They mention that if the
surface is left exposed in air for a long time, it may become unresponsive to
the plasma activation. But if you dip the PDMS in dilute HCl, then you can
reactivate the surface (I'm not sure exactly why this works with HCl, but it
does).

Three-dimensional micro-channel fabrication in polydimethylsiloxane (PDMS)
elastomer
Jo, B.-H.; Van Lerberghe, L.M.; Motsegood, K.M.; Beebe, D.J.;
Microelectromechanical Systems, Journal of
Volume 9,  Issue 1,  March 2000 Page(s):76 - 81


Also, I've found the time to be an incredibly sensitive parameter for bonding
PDMS. I typically expose the PDMS to plasma at 35W for 10-15 seconds (if I go
for 40-60 sec the bond may not work at all). Also, after putting the PDMS
pieces together, you should leave them alone for an hour or so because the
bond seems to improve from its initial condition.

good luck,
Joe Grogan



> -----Original Message-----
> From: Vishwanath Somashekar
> Subject: [mems-talk] PDMS to PDMS bonding
>
> Hi,
> I have been having problems with bonding PDMS to PDMS.
>
> One of the layers have the microchannel (3 mm X 100 microns). I cast this on
> a SI wafer on which contains the replica of the channel manufactured using
> SU-8. I use 10:1 ratio of sylgard 184 from Dow Corning. I mix the two
> solutions for about 3 mins and then degas for about 30 minutes till there
> are no air bubbles present. When I pour them, the over all height of the
> layer will be around 3 to 4 mm.( 20 ml + 2ml)
>
> The other layer is just a covering layer to close the top of the channel. I
> just cast in on a cleaned SI wafer. Again, I use the same
> 10:1 ratio of sylgard 184 here. The over all thickness here is about 2 mm.
> (15ml + 1.5ml)
>
> I then sonicate both of these layers in Ethanol for about 15 mins. I then
> rinse them with ethanol and let them sit in a 60 degree oven till they are
> dry.
>
> I then plasma treat them. I use a plasma cleaner manufactured I think by
> Yield Engineering Systems. I use O2 plasma at 1 to 1.2 torr, for 1min at
> about 22 watts. As soon the plasma treatment is done, I then bond them only
> to realize that the bonding is no good. I have been reading some discussions
> on the forum about using different proportions for different layers. I would
> appreciate if you could guys could suggest something to improve my dilemma.
>