Using autofocus during automated quant acquisitions

[wrigke]

John,

I run some very long (e.g. 24hr +) automated quant acquisitions and due to the extreme temperature fluctuations of the lab it is desirable to use both the drift correction and the autofocus feature.  The problem that I have run into is that for some reason, my autofocus does not correctly focus on my iron standard, possible due to its proximity to the edge of the sample mount.  It is typically off by at least 15um.

Is is possible to specify autofocus use on some standard measurements but not others?  The error generated by the improper autofocus result is much larger than the error resulting from the temperature-induced focus change.

I realize that this is not a great way to approach analysis but my lab is terrible and there is nothing I can do to fix that.

Karen

[Mike Jercinovic]

Hi Karen,
Is the iron standard metallic?  Is it carbon coated?  I think carbon coated iron metal gives a dark blue color depending on the thickness.  This light frequency is trouble for the camera I think as you just don't get the brightness that's needed for the autofocus to find it's maximum intensity.  I vaguely seem to remember running into this with either iron meteorites or steel years ago.  There may be a way around this by increasing the light intensity used for autofocus in the configuration, or maybe just analyzing uncoated (or with a thinner coat).  Could be something else, and obviously if you are not using C-coated metal then my little theory here does not apply, but if the autofocus is only failing on one material (and you can get it in focus with stage Z control manually), there must be a good reason.

Mike J.

[wrigke]

Hi Mike,

The iron is metallic but is aluminum-coated, not carbon coated.

Karen

[Mike Jercinovic]

OK, I have never done Al coating on Fe before.  Does it just end up looking bright and reflective or do you get some color in reflected light with the aluminum thickness you use?

Mike

[wrigke]

Hi Mike,

The coating is about 15nm thick and tends to be translucent to reflective depending on the angle of observation and the angle of light.  I have not noticed any particular pigment or apparent color change upon coating like one might observe with carbon.

Karen

[Probeman]

Hi Karen,
The first thing that comes to mind is are you already using the smallest field of view for the optical microscope, say 250 um?  Maybe that would help the auto-focus mechanism.

As for software there's no option (yet) in Probe for EPMA to specify to use autofocus on std vs. unk, vs wav samples (though it is a good idea), but you can explicitly digitize which samples/points you want to perform an auto-focus on from the Automate! windows as seen here:



So just select the Digitized option for the auto-focus and then right click the digitized points in the Automate position grid that you want to perform an auto-focus on. For standards, there is usually only one point explicitly digitized because the software auto-generates additional standard positions as necessary so that is obvious, but for unknowns you can just select the first point of each unknown or multiple points.

And/or you can select the Use Digitized Auto-Focus option and it will automatically "turn on" auto-focus for all positions as you digitize them.