Sepcon vs Polished samples
In an effort to imitate Sepcon, I applied nail polish on two samples covering the edges and the back side (well side) and compared its discoloration to Sepcon format and w/o polish sample in 2D format. I used two polished samples in which I covered the edges in both but the back side was not covered fully in one of them to see if that brings any change in discoloration profile.
These samples were from W684 w/both intact slits and kept in DMEM(-FBS) @ 37C in oven.
As seen, polished samples discolor in similar fashion to that of 2D format. There is hardly any similarity between the polished and the Sepcon samples. The similar blue ring pattern near the outer edge can be seen in both polished samples.
It could be that the samples were not polished properly and there were places which were left uncovered though I tried my best.
Is discoloration associated to stress, as the membrane is clamped in a Sepcon? Can this be imitated somehow?
To list some of the hypotheses (however whimsical) concerning why the SepCons promote membrane stability:
Number 1 seems to be contradicted by this study and some other recent studies. Number 2 seems to have been contradicted too in various places (right?). Number 3 is non-intuitive for me, but it is a theory that fits. It would seem to explain the intentionally broken membrane experiments (assuming breaking membranes relaxes stresses).
Can we vary the degree of compression at the edge of the membrane to change the stress profile in the film? Can we place membranes in SepCons w/o sealing with O-rings (lower the compression). Can we use larger O-rings and force a tighter fit (more compression)?
Vlassak did say he thought chemical vulnerability and film stresses were highly related. Unfortunately, I really didn’t understand his reasoning.
Regarding Vlassak’s comment, do you want me to get in touch with him and ask him to clarify?
I think stress is a long shot, and I doubt it would generate the observed pattern. IT still looks like some geometric effect caused by the sepcon. It’s hard to follow all the experiments that have been done, so hopefully we will be able to discuss this in detail at some point.
Jim – These results definitely contradict #1. #2 was tested sort of – remember that I tried to stand up chips along the well walls. Didn’t seem to help, but maybe we could brainstorm a better experiment to test this out.
From my notes, Vlassak said that higher stressed membranes are less “tough” in water than unstressed. I didn’t get a reason for this either. Shouldn’t the membranes in the Sepcon format be more stressed (due to O-ring clamping) than flat chips, and thus discolor faster?
I think I observed this today – I was assembling Sepcons with w667, and the membranes looked kind of wrinkled by my naked eye. After I clamped them into the Sepcons, the membranes seemed normal (stretched out?).
There could be another possibility with the Sepcon material coming into play. Such as, components from media getting adsorbed/absorbed by its surface and thus slowing the discoloration. This can be mimicked using a similar geometry like – cloning ring or Sepcon sort of cup made from different material. I will recheck how the cloning rings effect this process in comparison.
Barrett – if the membranes are flattening out, that’s a tremendous amount of stress to the wafer, and I’m surprised that it would not break. Back when the Sepcons first arrived, I looked at this using Brian’s scope. The Nomarski setting at 5X allows nanoscale curvature to be imaged, and I saw nothing but flat samples in the SepCon. It might be worth looking at this again, to see if this has changed. Are the samples too big – maybe pressing too hard on the walls?
Anant – with the cloning rings, make sure you reproduce the exact geometry of the specon, as well as possible. I can’t remember – is the membrane-side facing down into the external solution or up into the SepCon? I still think the effect is related to dead volumes and diffusion, but we’ll see…
Chris- in all my recent experiments (unless mentioned) the membrane sides are facing down into the external solution.
I have figured out certain geometries to mimic Sepcon using cloning rings, except that the seal here would be using vacuum grease (silicone) instead of the o-ring clamp. This could give us an idea whether is it the dead volume or the clamping which is responsible.