John,
What happens with respect to calibration drift when analysing an unknown that has been created using the "combined selected samples into a new sample" option, specifically in the situation of combining unknowns with different elements (and potentially different conditions) into a combined sample? e.g. If unknowns x(el1, el2, el3) @t1, y(el4, el5, el6) @t2 and z(el7, el8, el9) @t3, are combined into sample xyz(el1 ..el9), with standards collected at t0 and t5. The time stamp for the 'combined sample' is assigned t1 (I think). Will elements acquired at t1, t2 and t3 be assigned a drift corrected intensity associated with t1 (the 'combined sample' timestamp)? or does each element maintain a timestamp independent of the combined sample timestamp?
perhaps the answer is to use a different data collection strategy ![Smiley :)](https://probesoftware.com/smf/Smileys/default/smiley.gif)
Cheers,
Gareth
Hi Gareth,
The short answer is that the real time stamp for the standard intensity drift correction occurs on a data "point" basis. That is all the elements for a single data line/point acquisition. I could have made the acquisition date/time stamp on an element basis but I didn't.
That said, Cameca and JEOL don't even have a standard intensity drift correction at all, so just be happy you have any standard intensity drift correction! Seriously, assuming your various separately acquired, but subsequently combined for quant samples, don't have severe instrument drift issues, it shouldn't be a problem (the beam drift correction *is* on an element by element basis so you are OK there!).
If you are having severe instrumental drift issues affecting the standard intensities, you might, as you suggested above, acquire the samples to be combined later for quant, in groups, to avoid any really nasty standard intensity drift issues.
Can you give me an example of why you are acquiring samples separately for subsequent combining for quant? Are there crystal flips involved?
john