วีดีโอ

Hey, thanks for your support and for watching the video. Performing EDS on very small, low Z particles can be challenging because 1) the fluorescence yield is very low and 2) the volume of material available to generate X-rays is very small. if you have the capability, this would probably be a good scenario to try STEM-EELS instead of STEM-EDS. If you can't do this or you want to stick with EDS for some reason, you need to make sure there is optimal line of sight between your sample and the detectors. When you load the grid in the holder basket, make sure you load it with the film side down, as this side will ultimately face up in the S/TEM; this should eliminate (or at least minimize) any shadowing effects from the grid bars. Otherwise, if grid shadowing may be occurring, try to work on particles in the middle of the openings and not near the grid bars and lastly, you can play around with increasing the alpha tilt to see if this improve anything. If grid shadowing is the issue, then there should be a point when alpha is high enough to give clear line of sight. Of course, this is another advantage of STEM-EELS; the issue of "line of sight' is irrelevant for all intents and purposes.

Yes, it's amazing how often a reboot fixes most problems!

Hi Pkshah 420, At a fixed magnification, reducing the resolution of the image will increase the pixel size, the "detail" possibly lose due to that if pixel size if larger than the feature's size. However, with same dwell time, the frame time and total scanning time would be reduced by roughtly around 4 times and you would less influenced by the instability of the system. For 5.93Mx or higher, using 512*512 (0.33A pixel size of lower) would give better results than 4K*4K/2K*2K as the pixel size is not the limitation factor of the final image. It is hard to tell whether 1K or 512 resolution is better in your system for a specific sample as the sample drift, anti-vibration system, filed control, temerature stability are also the crucial factors.

Technically, you are indeed reducing the Nyquist limit by reducing the pixel dimensions to 512x512, but the field of view is unaffected by this. However, the reduction in Nyquist limit will only reduce the meaningful detail in the image if this falls below the actual instrument resolution. In the case of the Talos, the resolution is 0.16 nm, but the Nyquist limit for a STEM image at the magnification in question with 512x512 pixel resolution is still significantly smaller than this Thus, there really isn't any meaningful detriment to the image detail. If the pixel resolution was dropped further to 256x256, this may not be the case, but I'd have to bin the image to know for sure. If you check out the video on my channel about FFT fundamentals, this is all explained in greater detail (apologies if you were already aware of this): th-cam.com/video/8e3esi4DbHA/w-d-xo.html

@@NicholasRudawski got it buddy

@@xuzhaoliu2534 thanks for response

Thank you for your support and I hope you found it useful.

Hello! I believe you are correct about the model number. I will verify this sometime in the near future. If it turns out to be a different model, I will let you know.

Hello; my apologies for the late reply. My understanding is that the LHP trackball and "beam shift" in direct alignments are activating the same deflector coils and doing the exact same thing. The only difference is that when you select "beam shift", you can store a specific "location" for the beam. If you select "beam shift" from direct alignments, then center the beam with the MF knobs, then select "done", this beam position will be saved, so if you accidentally deflect the beam far away from the screen, you can easily find the beam again by selecting "beam shift" from direct alignments again. This is really more for convenience than anything else. I've never noticed if the trackball is disabled when "beam shift" is activated, but it would make sense if this is indeed the case to avoid redundancy.

Hello, and sorry for the late response. You are correct that the allowance of spots is related to the structure factor for each crystal structure. I believe in Williams and Carter there are calculated diffraction patterns for some common zone axes for some common crystal structures. I think Fultz and Howe may also have something like this, but I'm not sure. I really need to do a follow up video specifically discussing the structure factor (though it actually would've made more sense to do this before the video on DP indexing).

You're welcome; unfortunately, Velox does not offer an option to generate "corrected" maps, only "intensity" (no background subtraction, no peak de-convolution), net, at%, and wt% (which is derived from at%). Since Velox is generating at% maps, it must have tabulated and/or calculated parameters necessary to calculate the associated correction factors for each X-ray peak. Sadly, there is no way to see these correction factors, so the results Velox is producing are somewhat "black box" in nature. If you want to calculate the correction factors yourself, please check out the description of this video I did a while back on quantitative S/TEM-EDS (keeping in mind that the detector efficiency data is unique for every EDS detector): th-cam.com/video/H5HA9K3eRlo/w-d-xo.html

@@NicholasRudawski I watched the video that you suggested. I want to use theoretical correction, but I did not understand deriving transition probability values for high Z elements. Second one, for ionization cross section should we use m^2 unit to calculate k factor? last one, to get corrected maps for individual elements, simply I need to find (sigma*a*omega*epsilon) and divide each pixel with this number, right?

@@mehmetozdogan3023 transition probabilities are actually not derived or theoretically calculated, but measured experimentally. Please see this website: xraydb.xrayabsorption.org/ If you select any element you can see the relative intensity of each X-ray line within a given family, and this is used to calculate the transition probability. For the ionization cross-section, you may use any area units you wish as long as you are consistent; for calculating absolute atomic fractions, the answer should be independent of this. You are correct about generating the "corrected" maps: just divide each pixel value of a "net" map by the correction factor (sigma*a*omega*epsilon).

You're welcome.

Thanks, Adrian.

Yes, good for some non-hr imaging. STEM-BF dislocation image of some material could be way better than TEM-BF ones.

Of course it can! BF microprobe - the best thing for all metals, because of good diffraction contrast. Even relatively good high-res STEM in nanoProbe mode could be realized without probe corrector. Just make main alignment: condenser aperture, condenser astigmatism and comma (by seeing ronchigramm) - and you will be pleasantly surprised!

Hello! The people who already replied covered this quite well, but yes, STEM imaging can still be useful without a probe corrector as you can still perform all the different imaging modes I discuss without having a probe corrector (albeit at lower resolution). Plus, you still need STEM mode for performing any type of spatially-resolved analytical work (EDS or EELS). That being said, there is no question that probe correctors have dramatically extended the utility of STEM mode. Without probe correctors, I still believe the STEM revolution happens, but not to the same degree of significance and impact.

You're welcome; and yes, I tried to squeeze literally everything about STEM I could think of into the webinar.

Thank you!

You are absolutely correct; the Talos vacuum system is much simpler than this one!

Thank you!

You're welcome; thank you so much for your support.

What good timing! What system were you working on and what sample were you looking at?

You're welcome; thank you for your support of my channel!

Thank you!

Hi Alexander: thank you so much for the well wishes and for your support of my channel.

Hi Nataraj: thank you so much for the well wishes and your support of my channel.

Hi Stuart: thank you for your support. I've actually tweaked quite a few things with my lamella prep compared to what I did when I made this video. I don't tend to go quite as crazy with the E-beam Pt and I use a much lower current density when depositing I-beam Pt. I also no longer PFIB final thin below 8 kV, because I find that the curtaining at 5 kV (or less) is simply too severe (in spite of the supposedly thinner damage layer). A long time ago, I did use the rectangle pattern more when doing final thinning, but I now basically only use the cleaning cross-section. I find I simply have far more control of the thinning process and I end up with cleaner sidewalls using the CCS (which I imagine is how "cleaning" got in the name). If I want to be more aggressive with the thinning process, I still stick with the CCS, but I just increase the beam current. I have indeed heard good things about W as a protective cap, but have not had the chance to test this out yet. I'm curious as to what (if any) morphological differences exist between W and Pt deposited via both E-beam and I-beam.

Indexing a polycrystalline DP is actually much easier compared to a single crystal DP because there is no zone axis to consider. That being said, the process is very similar with measuring ring diameters and calculating ratios. Indexing polycrystalline DPs would definitely be a good addition to the channel and I hope to do this at some point soon.

@@NicholasRudawski thankyou

You're welcome.

Hi Ajay: there could be a lot of reasons why the Omniprobe is responding how you are describing, but the most obvious one is that the frame of reference is set incorrectly. You want this set on "Stage" instead of "Port". There could be other reasons, too, but it's a bit too lengthy to cover in a YT response. The jitters are most likely due to some kind of mechanical coupling/transfer. Again, a bit hard to cover in a YT response, but I would check your Omniprobe cabling to make sure it is properly isolated because that is the most obvious way to get vibration transfer.

Thank you for your support; I hope you found the video useful.

You're welcome

Thanks, Stan. About your question, I do not believe manual control of the energy range for a given element is possible with Velox; you are stuck with whatever "window" the software assigns. However, if you collect the map with TIA, I believe that you can do this. That being said, while TIA can give you more functionality in this regard, it is very limited in terms of collecting the actual EDS data because it only allows for "single pass" analysis whereas Velox allows for as many passes as you wish.

Sure thing; I hope you found it helpful.

Hey, so sorry for the very late reply. Yes, you can absolutely do this in STEM mode; the instrument mode doesn't really matter as long as you are getting sufficient EDS signal.

You're welcome; thanks for your support.

You're welcome.

Both de-ionization and distillation are processes to remove dissolved solids, so either will be a much better option compared to normal tap water. As far as one being "purer" than the other, I'm not entirely sure on that. I've never encountered a problem using our in house DI-water for sample prep, so there really isn't any reason for me to try distilled instead. That being said, it would be an interesting experiment to simply put a drop of DI (or distilled, or tap) water (no particles) on a grid, let it dry, and then image the grid to see what (if anything) still remains. Regarding the tweezers, these are actually correctly described as "self-closing" tweezers (rather than "self-locking"). Here is a link from Ted Pella: www.tedpella.com/twzr-sc_html/twzr-sc.aspx

Thank you, glad to hear you found my video helpful; do let me know how your samples turn out for you.

Thank you for the information about the pipettes, this was certainly something I was obviously not paying a lot of attention to compared to everything else. Also, I tend to avoid support films with formvar; I've had too many issues with instability under the electron beam (though the manufacturers claim over and over that formvar is fine and stable) and plus having to dip grids in chloroform to dissolve the formvar (if needed) can introduce a source of hydrocarbon contamination (though some plasma cleaning may sufficiently address this). Did you ever have hydrocarbon contamination issues when using chloroform to prep grids?

@@NicholasRudawski Hi..it would be great if you introduce us about great variety of TEM grids and their specific applications(I mean when to choose which grid).

@@NicholasRudawski That question brings back a lot of painful memories! Actually, we don’t do that anymore. A chloroform wash can be helpful, but it can also end in a mess. Let me explain the story a bit more. Patterned grids (like quantifoils) are made using a technique called soft lithography. It’s like creating a pattern on a wafer, then making a negative pattern using a soft silicone resin (e.g., Sylgard). This pattern is then used to stamp a polymer solution (like formvar) dissolved in a solvent (probably chloroform) onto a substrate like glass. Afterward, they cover the patterned film with another layer of carbon film on top to make it conductive and stable. For random holey carbon grids, the method is slightly different, but the idea is the same: first, make a patterned formvar film, and then deposit the carbon. However, I’m not sure how lacey grids are made, but I speculate that they’re produced in a similar way. It would be best to consult older EM books. Anyway, the carbon grids are supposed to be pre-washed and free of formvar or polymer backing film from the supplier. But in the past, we often received grids with formvar contamination. In some cases, a chloroform wash wasn’t enough or even made the situation worse. It was like trying to clean a huge mess with a wipe. Also, chloroform sometimes introduced its own contamination, which was really problematic. The contamination could come from mishandled tools or consumables, and in most cases, the source wasn’t clear. It could be from filter papers, mishandled parafilm, tweezers, pipettes, or even contamination in the chloroform bottle. So yes, I did experience contamination after using a chloroform wash. But nowadays, I see far fewer of these issues. Grids are usually clean from the supplier. In case I receive contaminated grids, I pass them on to other EM groups with less strict cleanliness standards. 😂 Finally, back to the original topic: heating the grids helps stabilize many of these microscopic contaminants. However, overdoing it may also make the carbon film fragile. It's also worth mentioning that in cryo-EM, we glow discharge all the grids, which effectively removes loose hydrocarbons. Although in glowdischaging cleanliness is not intended; It’s done to make the grids hydrophilic.

Thank you for your support and glad to hear you found the video helpful. I've been meaning to set up an open access Dropbox folder for all of the presentations I post on YT, but have not gotten around to doing this yet. In the interim, I will gladly send you the presentation via email, so please send me an email requesting the presentation and I will take care of this.

Yes indeed, doing the gun alignment (even with the monochromator) isn't too difficult (and the provided instructions are very good, too). As far as time to stabilize the mono after excitation, leaving it overnight is indeed probably overkill (though not a bad idea if you have the time to allow it); I think the apps person from ThermoFisher said that a couple hours was sufficient for this.

Yes, you are correct about the discs not overlapping (or at least overlapping less) in uP-STEM versus nP-STEM. I actually discussed this in a video about CBED I did several years ago using our Tecnai. You are also correct that true BF-STEM is not really possible in nP because of the discs overlapping, so uP also has an advantage here, too. It would be interesting to do an image comparison of BF-STEM in nP versus uP using a material with lots of dislocations. This would definitely be another good video to do at some point.

You're welcome; yes uP-STEM is definitely advantageous when you are trying to avoid disc overlaps in the DPs and form true BF or DF STEM images to improve contrast. I am planning to do another MSA webinar in mid November specifically about STEM imaging so that would be a great opportunity to discuss this (as well as the comparison with nP-STEM mode I did here).

You're welcome, glad to hear you found my video helpful. Your question is a very interesting one indeed. Should there be any difference between nP and uP if the convergence semi-angle and probe currents are the same in both cases? Well, in principle, there should be no difference because angles are angles and current is current, right? With the Themis being a 3 condenser lens system, it has the capability to arbitrarily set the value of alpha in either mode, but I don't know if nP has the range to obtain alpha < 3 mrad. It actually may be easier to make alpha larger in uP mode to be comparable to alpha in nP mode. Even with alpha being the same, I believe there would be a discrepancy with the currents in a 3 condenser system because of the "zoom" between C2 and C3 which causes the current to decrease as alpha is decreased.

Great question! While the OL effectively acts like two lenses, these cannot be controlled independently, so you are changing the strength (focal length) of both at the same time. This is why the MC coil is present in the upper pole piece, to reduce the effect of the OL pre-field when it (the MC) is (optically) on, which corresponds to microprobe mode. Otherwise, the system would only have nanoprobe mode.

You're welcome, and thanks so much for your well wishes. We were actually quite lucky and spared anything remotely close to serious and we never lost power. The good news of the university being closed was that it gave me an opportunity to record and upload this video!