I'm thinking of trying my hand at fabricating microfluidic devices. I'll be using a filament based on PETG (Zortrax's Z-glass filament, which is translucent and resistant to acids and bases). For my plan to work, I'll need to find a good way to bond the PETG model onto a glass microscope slide.

What is the best technique for doing this? Should I use epoxy resin or some other adhesive, or can I solvent bond it? I've heard stories about PETG permanently bonding to a glass print bed, so it seems plausible that solvent bonding might work. Will it work, and if so, what solvent should I use?

  • 2
    $\begingroup$ Nathaniel, were you able to solve this problem? We have similar issues here. Trying to bond PET-G to a glass microscopy slide... $\endgroup$ Nov 19, 2019 at 16:24
  • $\begingroup$ @SébastjenSchoenaers sorry, we ended up changing plans and not building this device. If you have any success it would be interesting to hear about your results, though! $\endgroup$
    – N. Virgo
    Nov 20, 2019 at 9:12
  • $\begingroup$ @SébastjenSchoenaers I just remembered why we changed plans: it turns out that you can make microfluidic devices very easily by cutting shapes out of Parafilm and sandwiching it between two slides. A craft knife works ok, but the Silhouette Portrait paper cutting machine works even better and is inexpensive. It won't work for every purpose of course, but it's what we ended up using instead. $\endgroup$
    – N. Virgo
    Nov 24, 2019 at 12:52
  • $\begingroup$ Any chance you can set up the glass you want to bond to as your build plate. Maybe heat the glass to 100C to build on. People have reported getting PETG to bond to glass build plates so strong that it pulls out chunks of the glass build plate. $\endgroup$
    – Perry Webb
    Sep 24, 2020 at 18:58
  • 1
    $\begingroup$ @PerryWebb at the time of asking the question I was using the Zortrax M200, which would have made that impossible, because it was a very closed-source machine and there was no option to print without a raft! If I was trying to do the same thing now, I'd probably try that. $\endgroup$
    – N. Virgo
    Sep 25, 2020 at 3:19

1 Answer 1


The solvents that can dissolve PET are pretty nasty -- I wouldn't personally handle any of them outside a lab fume hood. If you have that, a 50/50 mix of MEK and methylene chloride should work. (Increase MEK ratio if you want faster adhesion / less working time, and vice versa.)

First thing I would try is printing directly onto the (super clean) glass. Print the first layer at high temp to try to get a good bond. Then over-extrude to get watertight perimeters. It MIGHT be better to print on a cold bed to keep the PET from popping off when the glass cools, but you would need to do some experimenting.

You could also try heating the slide to the melting temp of PET on a hot plate and then attaching the printed part.

Failing that, a transparent superglue could be a good approach.

  • $\begingroup$ Thanks, this is very useful advice. Just heating the glass and melting it on seems like an easy first thing to try. The others sound a bit trickier, though we do have fume hoods. (I don't think I can try printing straight onto the glass unfortunately, because the Zortrax is a pretty closed system and doesn't give me control over those parameters.) I'll let you know if and when I try any of these! $\endgroup$
    – N. Virgo
    May 28, 2016 at 5:28
  • $\begingroup$ Oh yeah, mandatory rafts will keep you from printing onto the glass. I hadn't thought of that. $\endgroup$ May 28, 2016 at 16:40
  • $\begingroup$ Mandatory rafts and also mandatory autocalibration. If I put a glass slide on top of the build platform I'm pretty sure it will just smash the glass, because the first thing it always does is raise the platform until the nozzle makes electrical contact with it. It's an excellent machine, just not really suited to experimenting with technique. $\endgroup$
    – N. Virgo
    May 28, 2016 at 23:54

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .