Acceptability of Probe Current Interpolation

[Jacob]

I've been developing protocols for doing standards based EDX analysis in my lab using DTSA-II.  We have a Tungsten filament SEM with a 30mm Bruker SDD.  The problem, my SEM not being a probe, is that the Faraday cup is a bit of a hike from my sample.

I've set up the system to monitor probe current over long timescales a couple of times to get a feel for what I'm dealing with (plot attached).  On an hour long interval, it looks more or less linear.

So, the question is this: Is it acceptable to read the probe current, collect for 10 minutes, then read the probe current again, and assign probe current start and end values to the collected spectra using a linear interpolation between the points tied to the timestamp?  Assume a sanity check for accepted probe current rise or fall is in place to catch transient spikes or dips as well as an appropriate warm up time of 30 minutes for the gun to stabilize.

[John Donovan]

So, the question is this: Is it acceptable to read the probe current, collect for 10 minutes, then read the probe current again, and assign probe current start and end values to the collected spectra using a linear interpolation between the points tied to the timestamp?  Assume a sanity check for accepted probe current rise or fall is in place to catch transient spikes or dips as well as an appropriate warm up time of 30 minutes for the gun to stabilize.

Hi Jacob,
I think it's entirely appropriate what you suggest.

In EPMA we generally measure the beam current for every acquisition since we have a faraday cup but, for example, for our standards I perform a similar drift correction for any drift in the standard intensities if they were acquired both before and after the unknown(s), based on the unknown time stamp as you mentioned above.

So we perform a beam drift correction for each analysis, but also a standard drift correction for each standard on an element by element basis. Why? Because with an EPMA instrument different spectrometers (and analyzing crystals) can drift independently of each other (unlike most EDS spectrometers).
john

[Jacob]

Thanks John!  I'm writing up some utilities in python that should make should help anyone with a similar problem.  My hope is to make a utility to read the probe current with a timestamp (complete), then batch process my EMSA files to assign doses according to timestamp (incomplete).