Fractional emission distances from the probe in DTSA-II?

[pvonlant]

Hello,

I have been using the DTSA-II software for a while to simulate the non-absorbed X-rays emitted by geological materials under various acceleration voltages. The "Fractional Emission Depths and Volumes" table generated during this process is extremely useful to evaluate the spatial resolution of X-ray analyses, with depth values (in [µm]) corresponding to the 50.0%, 90.0% and 99.9% fractions of non-absorbed X-rays. I was wondering if the generation of a similar table, but displaying the "Fractional Emission Distances from the Probe" is planned for the future. Such a table would provide extremely useful complementary pieces of information to evaluate the radial distribution of non-absorbed X-rays, e.g. when selecting the acquisition parameteres to be used for X-ray maps.

Thanks in advance for your reply.

Cheers,

Pierre Vonlanthen

[Nicholas Ritchie]

Help me to understand better.  Do you mean radial distances in cylindrical coordinate system where the initial probe direction defines the cylinder's central axis?  (A measure of beam spread?)  Or do you mean a L2-distance from the point at which the probe enters the sample? (How far as the crow-flies does the electron travel from the point of first interaction?)  I tend to think that the former would be quite informative and relatively easy to interpret.  The later is a little harder to interpret as it convolves depth and radial distances.

[pvonlant]

Dear Nicholas,

Thanks for your reply. I meant the radial distances (in µm or nm) measured normal to the primary electron beam axis containing 50.0%, 90.0%, and 99.9% (called D50.0%, D90.0% and D99.9% hereafter, respectively) of the non-absorbed (i.e. escaping from the sample surface) X-ray intensities. The two figures attached to this message may help making things clearer. Figure 1 is a diagram showing the intensity of non-absorbed X-rays (Y axis) with respect to the radial distance (X axis) from the electron beam axis. Using integral calculations, I assume that D50.0% equals the X value for which the area below the curve (i.e. delimited by the curve, the XY axes and the vertical straight line passing by D50.0%) corresponds to 50.0% of the total area located below the curve (i.e. delimited by the curve, the XY axes and the vertical straight line passing by DMAX). Figure 2 is a sketch generated by another software performing Monte Carlo simulations (Casino 2.5). Unfortunately, a lot of post-processing is necessary extract D50.0%, D90.0%, and D99.9%, from this software, and I have to admit to have been unsuccessful yet. Obtaining those values from the DSTA-II software (in the same way we get the "Fractional Emission Depths") would be a great complementary piece of information I think.

Cheers,

Pierre

[Nicholas Ritchie]

I think this is a really good idea and I'd be happy to implement it.  If you don't hear back in a week or so, give me a kick.

[Nicholas Ritchie]

I've got an implementation for you to test. It can be downloaded from the standard pre-release link: https://drive.google.com/file/d/17YAXRYkQ9vAQ509AwP5eZagdGfShX-dE/view?usp=sharing.

[pvonlant]

Great! Thanks for your quick reply! It is exactly what I was thinking about. From a visual point of view, it corresponds very well with the feeling I got from qualitative maps in which phase boundaries are supposedly vertical. When compared to the curves generated by Casino 2.5, I was (visually) thinking the 50.0% radial distances to be even a bit shorter (i.e. closer to the beam axis), but science has spoken! My background in maths is very limited, but I guess that it highly depends on what we consider DMax to be (99.999% or 99.99999999999% makes a sizable difference here I think). Thanks again for your help. If you get positive feedbacks from other users, I would definitively integrate the "Radial Distances" to the next official release of DTSA-II.