Forums » Instruments » Microscopes »
Leopard skin ice "contamination" on new F20
Added by Gabriel Lander over 10 years ago
We've been having an ice problem with our brand new F20, and we, along with FEI engineers, are baffled. After a few hours of cryo data collection, the ice on our grid begins to become blotchy, and after 5 to 6 hours the grid is unusable. I have attached 2 images that exemplify what we see appearing after a couple hours in the scope, along with their FFTs.
If anyone has any ideas on how to remedy this issue that has been plaguing us since the microscope's installation, we would be happy to hear it. We have made the following observations:
1) This problem is not holder related - we have tested 4 different cryo holders in this scope and other F20s, and we always see this problem arise in this particular scope, but not others.
2) The holders maintain the same temperature in this scope as in all the others
3) This ice appears over the entirety of the grid, and is NOT localized specifically where we have been imaging.
4) This problem seems to be worse if the sample has detergent in the buffer.
5) There are no obvious vacuum leaks, according to the FEI engineers.
6) The cold trap aperture is definitely in, and is observable at low mag. The liquid nitrogen level of the dewar cooling the copper braid decreases at the same rate as all the other scopes.
7) Repeated short cryo cycles, long cryo cycles, or short cryo cycles followed by extended pumping of the column by the IGPs does not affect this problem.
8) This ice appears whether a short (20s) or long (90s) airlock cycle is used.
9) The oil diffusion pump at the projection chamber has been replaced with a turbomolecular pump in these new generation F20s
| iceBefore.jpg (410 KB) iceBefore.jpg | ice at start of data collection | ||
| iceAfter.jpg (410 KB) iceAfter.jpg | ice after a couple hours of collection |
Replies (4)
RE: Leopard skin ice "contamination" on new F20 - Added by Arne Moeller over 10 years ago
Hey Gabe,
I had the same effect on T1 - thus it is not scope related.
As in your case there were no obvious problems during data collection (no vacuum break or low N2 etc).
I only experienced this problem when I was imaging samples that contain detergent - like you stated in 4.
To me the structure of the ice looks detergent influenced.
Many detergents form filamentous structures that look a lot like the ice that you have seen.
DDM for example is very prone to form filaments - what detergent are you using?
Arne
RE: Leopard skin ice "contamination" on new F20 - Added by Gabriel Lander over 10 years ago
Hi Arne,
I was under the impression that Dmitry collected all the HIV Trimer data on T1, which had detergent in the sample. Do you recall how long it took for the contamination to set in on T1? We should check the same detergent grids out on T1, but we definitely don't have this problem on the Talos or the Krios. FEI recently confirmed that the rate of contamination is significantly worse when the sample has detergent on T3. We have this problem when using either DDM or NP40 (these are the only detergents anyone has tried out).
RE: Leopard skin ice "contamination" on new F20 - Added by Dmitry Lyumkis over 10 years ago
Hi Gabe, Arne,
Yes, I did collect the trimer data with detergent, at 2X below CMC for DDM, and I didn't get contamination on T1 for those 3 datasets (1-4 days of collection), But as I mentioned before, I have seen the same exact contamination (to greater or lesser extents) both on T1 and on T2 with, nominally, the same concentration of detergent. This is why I think it is important to test different detergents, and also different concentrations of detergents. It is possible that there is something on T3 that exacerbates the issue, which is why you may not see the same effect on the Talos or the Krios, but I do believe that the root cause of this is detergent. I would suggest trying shorter chain detergents (e.g. decyl maltoside (DM) or octyl-maltoside (OM) or OG) and/or milder detergents. I already have plans to do this when I get up and running here, but if there are more poeple that are concerned with this issue, I believe we could speed up finding some solution.
RE: Leopard skin ice "contamination" on new F20 - Added by Arne Moeller over 10 years ago
Hi Gabe,
in the most prominent case the filaments appeared after 8 hours - however interestingly after I switched to a different square.
As Dmitry already stated I also believe that this is detergent dependent (mostly). In my case I used a very high concentration of DDM (way above cmc)- at this concentration you also see filaments in stain - and they look very similar to those in ice.
It definitively is a big difference whether you are using the detergent above or below CMC - this has to be kept in mind when preparing the sample!
I also think that DDM is not the best detergent if one wants to achieve a homogenous particle orientation - while it certainly works (as Dmitry showed - HIV)- milder detergents should be able to do this job too, hopefully not creating the background.
Arne