Urgent Sealing Problem

[JMOwen]

I have a particulary difficult sealing problem that I hope someone may be able to help me with. I'm trying to create a gasket seal between an thick aluminium plate and the side walls of a 24 deep well polycarbonate plate. Within the plate are different strains of bacteria isolated in separate wells which I am trying to contain. ANY leakage across the seal quickly results in contamination of adjacent wells as the bacteria doubles its numbers every 20mins. At present I have a 3mm 40 shore silicone sheet sandwiched between the two components. The sides of the rectangular wells are 1mm thick and flat and free from visible defects and act as a knife edge seal into the rubber. I am applying 150kgf which equates to ~200psi accross the seal (imagine a grid of sealing edges) and I am still seeing cross contamination. Any ideas?

 

[IRstuff]

how are you verifying that the contamination occurs while the seals are engaged? e.g., can you rule out contamination during the contact process or release process?

How are the bacteria climbing the walls in the first place? Unless there is some media or pathway, they shouldn't be able to climb the walls, which makes me think that the contamination occurs elsewhen.

TTFN

 

[IRstuff]

Can you mark the bacteria with fluorescence to see how and when they are breaching the seals.

TTFN

 

[JMOwen]

IRstuff,

Thanks for you replies. The bacteria are in media which is agitated and splashes against the side of the wells and wets the seal. I'm autoclaving all components before assembly and innoculation and have performed enough controls to be fairly confident its coming across the seal. Contamination does not occur during contact but does on release so we measure bacterial levels immediately after disassembly to prevent a false positive (bacteria does not have time to grow to detectable levels)

It's such a challenge because the cells are ~1um in diameter and they must be contained for ~18 hours. With a 20mins doubling time cell can multiply to 9 E15 under optimum conditions! In practice they run out of food but

Flourescence is an interesting idea, I was considering doing something similar with tartrazine but think that carry over on dissassembly will yeild a false positive.

Thanks again - keep those ideas coming!

 

[monkeydog]

Two things come to mind for me.
Could the gasket material be too thin, and result in being cut through to the aluminim plate? This would result in a leak path.
Can there be a nick on the knife edge providing a leak path?

What if you changed the polycarbonite knife edge to a flat edge with ridges? I will try to illustrate below:

---^---^-- <<< Sealing surface
| | <<< Poly wall

Can you helium leak check the bacteria's home before they move in?

Monkeydog

 

[JMOwen]

Monkeydog,

The seal is 3mm thick, even at that thickness with the force I'm applying I was worried about cutting the seal but it appears to be very robust with no sign of damage. The nick in the knife edge thoery may hold water (pun intended) but there is no pattern the the xcontamination. It's probably down to probabilities.. if there is a nick where the infected media lands, if the media makes it through, if the cells grow through the media to the other well and if they manage to contact the media in the other well. The profiled lip is a nice idea but will be difficult to implement due to the plate tool design.

Thanks for your help!

 

[IRstuff]

It would seem that the overall concept may not be workable given the scenario you describe.

It would seem that you need a seal that can NEVER be opened in conjunction with any other well, since the splash is not controlled and there is no obvious way to coax the little buggers back into their wells.

>> Something like a rubbery egg carton construction with the high points opening above the well? The low points permanently seal around the wells. The metal blade contraption closes the openings during processing. Release the blades and the wells open up with no hopping from one cell to another.

>> Deeper tapered wells so that the metal taper actually enters the cell and seals down in the cavity. Release requires a slight lift, followed by some vibration to dislodge drops and then full removal. A pointy end could mitigate the drop retention.

TTFN

 

[JMOwen]

Thanks IR, both nice concepts but the challenge is only to prevent cross contamination while the assembly is sealed. A small amount of contamination on opening the wells does not matter as the contents are immediately processed and any competing bacterial strains in the wrong place do not have time to establish themselves. Sounds so simple doesn't it? Short of gluing the two components together,which is not possible as the aluminium component is reused, it is proving very difficult to acheive consistant results. I'm wondering whether a conformal coating on the top of the wells would help? Another more wacky possible solution is to impregnate the sealing faces with an antibiotic....

 

[monkeydog]

JMOwen,
Mechanics will sometimes cheat, and put a grease on a sealing surface to help seal a troublesome gasket or o-ring. So maybe your thought of putting an antibiotic on the sealing surface is not so far off.

 

[EdDanzer]

My guess is that the cover and or container are deflecting when clamping the seal so some areas are not actually sealed air tight. Too much clamping force in thin parts could be part if the problem. Pressure testing the seal with a dye to mark the leak path should show the problem.

 

[tothepoint]

I tend to agree with the last comment- that your top cover is rotating (bending) and allowing either lift off of some sections of the gasket, or further reducing what is already a low seating stress for an elastomer seal. How is the compressive load applied? With bolts or jacks around the perimeter of the cover? Or uniformly and centrally loaded?