Microporous Silicon Dioxide (TEOS) vs. Nitride Imaging
After a false start with non-microporous TEOS chips, I quickly did an image comparison with truly porous chips using polystyrene beads.
The TEOS and Silicon Nitride chips were both fabricated in the single-well CytoVu format (single 2 mm window, 5.4mm die size). The TEOS chips allows water and media with serum to much more easily spread under the chip. Zach will be providing more details later.
I suspended 10 micron polystyrene beads above the membranes. Water was also present below the membrane during imaging. Images were collected using phase contrast with a constant 20 ms exposure at varying focal planes. In all cases, the TEOS membrane was highly transparent with only minimal “pore background”. We will probably be collecting cell culture images next week. Images below were collected at 20x.
This looks promising and is exactly what should happen with the lower refractive index of this oxide material. I’m glad they also wetted better. Did you have any issues with breakage of the SiO2? Thus far, it appears that the TEOS oxide is much weaker than SiN, but this is hard to translate into a practical use metric. We may need to increase the thickness of the TEOS oxide to be comparable to the SiN in terms of robustness.
The 1000C seemed to be quite fragile. The 600C annealed chips were not that bad. I was careful and didn’t break any of the 600C chips. I allowed the bead solutions to dry up overnight and there was no breakage.
We also cultured cells on the non-porous TEOS 600C chips and there was no dissolution or breakage after 24 hours. I pipetted solution on and off top and bottom to try to induce transmembrane flow before discovering that they were not porous. None of the 600C TEOS membranes broke during that testing.
That’s good news. We’re currently optimizing the annealing conditions and trying to figure out what treatments will lead to the strongest and most chemically stable membranes, while maintaining flatness. Having 2mm membranes that are stable enough for experiments is a good start.
Thanks!