Hi everyone, hope you're all okay. So I have some questions/doubts about eds mapping. Just some notes first: I always use the correct accelerating voltage to excite the elements I need, always use correct probe current to have a good dead time and counts and not forgetting to check the working distance, and last but not least I always try to perform eds with flat surface and polished (but sometime this is not possible).
My doubt is: sometimes when I do eds mapping the distribution of the elements appears dislocated. The pixels dislocates not corresponding to the SED image. Why this dislocations happens?
Besides the deslocated pixels, most maps is very "noisy", I've changed the resolution already on the software to 1024 instead of 4080.
Peform map in 4080 resolution takes too much time. So I would be very grateful if anyone give me some tips on how to improve eds maps.
Thank you for the attention
Please give some more detail. Which instrument (SEM+EDS) are you using, which software? Best yould be a screenshot with map and details.
How long do you map? Seconds, minutes, hours? Which elements, which kind of sample, what concentration, which resolution/magnification? Do you use drift correction?
One of the most common causes of mismatches being images and maps is sample movement. This may be caused by loss of volatiles from the sample (which will cause it to shrink) or, most often, the effects of temperature changes. These may be induced by the electon beam or through environmental changes - try letting the sample equilibriate in the beam for a while.
What do you mean the map is dislocated?
The map is fuzzy in one direction or this is more like each time the map is few micron shifted from the SE image but not distorted?
What is the nature of samples you use (conductive, insulating)?
How do you mount specimen for polishing and for SEM investigation?
What is the beam current you use?
What is the surface area of the EDS detector?
What is the scan area?
Do you really need 1024 resolution (consider excitation volume for energy and material you use and features you want to analyse)?
Can you change amplification time (amp time) in EDS software?
Can you change the sample-detector distance?
After capturing a map can you "click" one point and see how many counts at highest peak you get?
@AladinUllrich the SEM is a JEOL IT 300, EDS is a xMax Oxford the software that Oxford provides is called Aztec.
The elements was Nb, Cr and Fe (alloy steel) and I let it running for a hour. We have some problems where the SEM was installed, there's vibration, there is an AC but it doesn't handle well the temperature sometimes...I know it is not suitable to work with those conditions, but that's what the college offers to us (sadly).
I'm a bit knew working on a SEM, I've already heard about drifting, but nothing related on how to correct. May you explain me? Thanks!
@Grzegorz not distortion, but sometimes when doing the maps the elements distributions not coincide with the secondary electrons images, there is some dislocations. The sample is a alloy steel.
For polishing here in the lab is all manual, there's no automatic machine to polish, but the sample is always built-in in pmma resin to polish.
The sample is fixed on a sample holder.
Beam current/probe current to reach 30% of dead time is about 60/70.
The surface area of the detector is 20mm^2
The scan area varies depending on what I need to see.
Yes I need 1024 resolution otherwise the pixels on the image gets too big. The options are (316, 512, 1024, 2048, 4080) if my memory is not failing those are the resolution aviable.
The amp time here is called processing time, and it's set in 6 (as they recommend).
No, I can't change the sample detector distance.
No, I can't see the counts if I click in one point.
In the Aztec software, there should be drift correction included. I don't know how it works with the JEOL SEM, but at our Zeiss instruments the drift correction of Aztec works quite well in case of trackable sample structures (interval of 1min is fine in most cases). Try out the drift correction, but be sure that it does not reach or exceed the limits.
How do you fix your sample? Do you use these black carbon stick pads? These often lead to sample movement. Try to fix your sample with conductive silver or colloidal graphite (my favourite material for fixing samples in HV/UHV).
The structures on your material are not on nanoscale, right? So you don't use very high magnifications?
Is there also drift in SEM mode (low beam current)? As mentioned above, try to let your sample settle down for some time.
An AC that does not keep T=const is fatal for SEM&TEM...
Beam current/probe current to reach 30% of dead time is about 60/70.
What means 60/70 nano amps?
They recommend processing time of 6 because of peak resolution? Do you need best peak resolution for mapping?
Can you increase the beam current and adapt processing time to get dead time below 30%?
What is bad about big pixels? They simply look bad or are larger than features you are investigating? If so consider doing maps at higher magnifications but with smaller sizes in pixels. Remember that there is no point to scan with step size smaller than excitation volume for Xrays.
@grzegorzcios here we never do high resolution imaging, totally impossible because of the condition the SEM is installed as I said before.
Sorry I read wrong, the value of beam current it varies depending on the accelarating voltage you're using. Actually when I read the manual the SEM there's some standards values, for example: if you're using 20KV the beam current should be 86 uA.
To reach 30% of dead time, we increase de Std PC, in others SEM the Std PC is called Spotsize.
They recommend processing time 6 to get better count.
Since the SEM is installed in a college, the users use the equipments to publish their research, some articles and etc...that's why we set the configurations on the software for higher resolution, the users always want something more beautiful aesthetically.
@aladin I always a sample holder when I can to fix it, and not only the stub with carbon tape.
Usually samples fixed on carbon tape are powder sample.
And yes, here we never do high resolution imaging because of the condition the SEM is installed.
I've already read the software manual but never found something related to drift correction, so I will search more about that.
Thank you!
Try to use the optimum working distance provided by the manufacturer for EDS analysis . In most of the time the EDS-detector size is also matters in mapping quality.
- Badges
- Science method