Hello everyone,
The workshop in my lab printed a mold with a Formlab resin printer (black resin) to make a simple microfluidic chip.
I'm aware that photoinitiators in 3D printer resins can inhibit the cure of PDMS.
I tried several cleaning/curing of the resin and everytime my PDMS (1:10 sylgard 184) is overall ok, except for the parts in contact with the mold which is milky and slightly sticky.
This surface results in no correct bonding with glass after plasma treatment...
So my strategy was to cast an intermediate polymer on the PDMS chip and then recast PDMS on top to get my final crystal clear PDMS working chip.
I tried with what we had in the lab > NOA73 that I pushed to 10min UV exposure for curing but surprise I have the exact same: a milky side (in contact with the milky side of the PDMS) and a clear side (air side).
When I cast PDMS on that it's curing everywhere except where the NOA is (fully unpolymerised).
I read that PDMS casting was ok on NOA, I even found a method paper doing this.
So I wanted to know if someone has struggle with this kind of materials before and if they found a solution. I precise that I need the flexibility of the PDMS for my final chip (embeded electronic valves and tubing).
I will certainly try epoxy and polyurethane as intermediates.
Thank you,
Hi Nelson,
I worked quite a lot on this question of PDMS curing inhibition by 3D printed molds, and most of the time, a combination a UV post-curing and thermal treatment is working fine. We published a paper on that point (Article PDMS Curing Inhibition on 3D-Printed Molds: Why? Also, How t...
) and the Formlab Black was one of the resin we tested. Try simply to bake the mold 1h at 120°C or 24h at 60°C, and PDMS curing should be fine.If it does not work, I'm really interested in checking why. It seems that Formlabs changes from time to time the resin formulation while keeping the names...
Since we wrote that paper, I also have the feeling that PDMS replicated from 3D printed molds is harder to bond with plasma... Not quantitatively tested, definitively not impossible (we did a lot of projects using plasma bonding of PDMS from 3D printed molds), but it still seems weaker. So I'm not surprised that not fully polymerized PDMS does not bond to glass in your case.
For intermediate material... NOA 81 definitively inhibits A LOT the PDMS curing (I do not know for NOA73), and PDMS inhibits the curing of NOA because of the diffusion of oxygen through PDMS. PDMS curing on epoxy is normally totally fine, but I'm not sure about your strategy of trying first to mold an imperfect PDMS replica into an intermediate material then back to PDMS...
Hi Bastien and thanks,
I actually tried your curing method and it worked the best but the PDMS was still milky (not sticky though, so cured fully?). And yes bonding to glass was not very strong, better than nothing for sure.
I managed to reduced the opacity a little by putting mineral oil in my mold for 1h at 60C°, something I saw on a comment once (and it was easy to implement). I can't find this comment again but it was on researchgate. I was not as clear as on a wafer but good enough.
Anyway, I realized that this chip does not need to be actually closed by glass. I tried to bond my cured milky chip to a nicely flat 4mm PDMS and it worked pretty well (for 10:1 and 5:1).
I will try epoxy for further use.
Thank you for the info on NOA. I tried to replicate the method NOA/PDMS in 10.1007/s10404-017-1943-2 but left me with a messy cure/incured sticky PDMS mix instead. They used the NOA81 not 73 though. Might be that, according to NOA manufacturer, the 73 is more sensitive to temperature than the 81 just after UV curing. "NOA 73 bonds will initially withstand -15° C to 60°C and -60° C to 90° C after aging". I cure my PDMS at 60C° so maybe.
Thanks!
What do you mean by "milky"? Is it that the PDMS is really less transparent, or could it be that the high surface roughness of the 3D printed mold got transfered to the PDMS replica, leading to a surface which diffract/diffuse light a bit more?
I have to admit that I have never observed any decrease of the transparency of PDMS replicates...
Happy that this solution still gives you useful chips!
Hi Nelson,
I had the same problem with Anycubic Photon printer's resins. In our lab, we use Mann Ease Release 200 nonadhesive spray. After spraying (just a thin layer), place your molds on the hot plate (2 hours, 80 degrees) to create a thin and uniform covering layer.
It works fine, but still be ready, that 1 or 2 PDMS platforms will be just for the cleaning step.
The PDMS-glass, PDMS-PDMS bonding works well.
Cheers,
Michael.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Coating