Faster measurements are a very understandable desire. Especially in analytics applied in scanning electron microscopy mainly two opportunities exist: increase of the source intensity (beam current) and/or improvement of the detection sensitivity. Regarding the last one practically no constrains exist, whereas the first opportunity represents a limitation which needs to keep in view: the interaction with the sample should not change the object of desire itself. In this respect it is concerning that more and more SEMs reaches a beam current in the scale of 100nA and more. Focused on a very small beam size (FE-SEM) it enables to "melt samples", at least to increase the local temperature in an unverifiable manner.
How can one be sure that one does not affect the microstructure of a material (as occasionally reported during FIB preparation)?
What about beam deceleration? Apparently, recently developed electron optics provide high voltages inside the electron column (thus avoiding aberration effects, which would reduce the lateral resolution) by a so-called beam-booster, but decelerate the electrons close to the sample surface to low landing energies, in order to prevent sample alteration as well as to improve depth resolution.
The reason is as you allude to - the analytical detection limit. For materials that are stable under the beam (metals, refractory oxides), the ability to up the beam current is a real bonus High beam currents are especially useful if one has dropped the accelerating voltage to reduce penetration depth - this also potentially drops the over voltage and the probability of x-ray generation.
(Even on a FEG there is an increase in spot size with beam current, but as you say, they can end up being an expensive electron spot welder.)
Moin Gert,
how necessary? It depends on what you want to use the instrument for, but in my opinion, it is fantastic to have that option available. As Jon Wade writes, detection limit vs time spent per spot analysis improves, but accuracy or thermal stability of the phase you're zapping is primarily a material property thing. If your user base wants to determine the Ti content of tiny olivines, you want to be able to reach a stable high current. If you want to look at fluffy polymers the next day, you will reduce the current. That is a great luxury to have. If you know in advance that you exclusively do fragile materials, you may not need the high-current option at all.
Actually, [thread hijack] one aspect we worried about is what happens to the stable material after high-current analysis. A lot of the work we do is FEG-EPMA followed by SIMS on meteorites/volcanic rocks. For instance, water content or D/H of olivine. The tests we did showed that regions exposed to focused beam at high current for 10 minutes vs pristine were indistinguishable in terms of H2O content, 18/16O and D/H. Has anybody done similar tests on different materials/isotopes etc.?
I can imagine that the crystal structure can change during the analysis, but it may not be permanent. Nothing we could detect with Raman, but then we intentionally stayed away from really fragile minerals.
Cheers
Eric
Hi Eric,
it is clear that for EPMA or EDS the current usually cannot be high enough. I was mainly focusing on microstructure changes caused by high beam current (or better: energy). As you mentioned the chemical composition is perhaps not that much affected, but crystal structure as one aspect of the microstructure is. It is clear that if you see the change in your (surface) image you can reduce the electron energy until the visible effect disappears, but who tells you that you don't already change something which you cannot see? I am not sure how much we adulterate our samples with High-energy SEMs since we only look for short measurement times and not think about possible side effects.
During my studentship at Michigan State University I used FE-STEM, and I had similar concerns regarding the homogeneity of my multilayer (superlattice) and beam effects at the measured spot. Therefore I was routinely doing EDX (for elemental analysis) and microdffraction at the same spot (for structural analysis), in addition to EELS in the high energy range (> 100 eV) to lock for core edges of possible contaminants. This type of measurement was repeated at the same spot to check for reproducibility of the results, and effects of the interaction of the beam with the sample during previous measurements. That way I had a clear picture of how the beam affects my sample at the measured spot (~ 10 A).
Hi there,
we use the SEM to excite cathodoluminescence in semiconductor samples. Most of the investigations are done with liquid helium cooling, so melting does not take place. With samples having a higher defect concentration, we are partly unable to saturate nonradiative defects. An increase of the acceleration voltage is mostly not an option because the spatial resolution would suffer from this. So if I had the chance to increase the primary electron current by some ordners of magnitude, I would do so!
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Material Characterization