In order that the measured intensities get properly corrected for matrix effects we need to know the composition of the sample (and the standard).
For standards this is easy because of course we already know the composition of the standard (or it wouldn't be a standard!). So based on the "published" standard composition from the Standard.mdb database we calculate the physics relative to the theoretical pure element. This is called the standard k-factor and is calculated as seen here for each element.
If the standard is a pure element, then the concentration is 1.0 and the matrix correction is 1.0 and therefore the std k-factor is 1.0. In any case the std k-factor is included in the calculation of the unknown concentration by iteratively determining the unknown composition starting with the "1st approximation" where the ratio of intensities (I
unk/I
std) is assumed to be equal to the ratio of the concentrations (C
unk/C
std) and calculated as seen here:
What does all this mean? It means that we need to know the correct composition of the unknown in order to correctly calculate the concentration of each measured element. Therefore, if some elements are *not* measured, for example, geologists often (but not always) assumed formula stoichiometry for including oxygen in the matrix correction. Obviously in the case of variable oxidation states, e.g., FeO vs. Fe2O3, replacement of stoichiometric oxygen by halogens, and glasses containing significant H2O, there is some ambiguity in the matrix chemistry which can significantly affect the matrix correction for measured elements as seen in these posts:
http://probesoftware.com/smf/index.php?topic=62.msg235#msg235http://probesoftware.com/smf/index.php?topic=81.msg292#msg292In practice we can sometimes "get away" with a simplification of the matrix assumption, for a quite amazing example we might try measuring only U, Th, Pb (and Y and La for the interference correction on Pb Ma) and assuming CePO4 by difference in the matrix correction of the mineral monazite. Surprisingly this assumption works to an extent that is quite impressive, though obviously one cannot assume that this simplification of the matrix composition is accurate enough without careful testing both ways.
However, it is not uncommon that if the element in question does not cause an interference and is not especially critical in the matrix correction (run a model in CalcZAF and check the absorption correction!),
we can simply specify the element as an "unanalyzed" element. In CalcZAF and Probe for EPMA this means that the element does not have an x-ray line specified as shown here in the Elements/Cations dialog from the Analyze! or Acquire! windows:
Once the element is entered as an "unanalyzed" element, we can use the Specified Concentrations (or Calculation Options) dialog from the Analyze! window to enter element concentrations by "specification" as seen here:
This dialog, seen below, has many options for entering elements that are unanalyzed, or even measured by another technique:
The simplest method is to click on a row in the element grid containing a specified (or unanalyzed) element and enter the specified concentration in elemental or oxide weight percent. One can also specify by formula, by standard composition (useful when analyzing a standard as an unknown), by text file input (see User's Reference manual for details), or by the (one time) analysis of another sample or by the analysis of a sample just prior to the analysis of the currently selected sample. Whew!
Other options for matrix specification are available in the Calculation Options dialog from the Analyze! window which we turn to next.
If you are a member, please feel free to add your website URL to your forum profile