Hi Michael,
The f(x) term is called "f of chi" and is the inverse of the absorption correction. Your value indicates that there is a significant absorption correction, so that simply means that you might want to look at the MAC value for C ka by the various absorbers in your system, i.e., Fe, Cr and Ni.
I would refer to the "Empirical MACs" dialog in PFE and see how those differ from the default values you are using from the default MAC tables.
But if the carbon is in trace amounts, the background corrections, carbon contamination and spectral interferences will be your main concerns. Here are some other topics to look at:
Trace carbon and contamination:
https://probesoftware.com/smf/index.php?topic=933.0Also consider the use of a blank standard for obtaining a zero carbon measurement:
https://probesoftware.com/smf/index.php?topic=1005.0Basically there are two traditional approaches to trace carbon: First the use of multiple trace carbon standards and a calibration curve (e.g., Nippon Steel and Caterpiller). Second the use of an MAN curve type background correction (e.g., Pinard and Donovan).
Both methods are available in Probe for EPMA in addition to a blank correction method.
Let me know if you would like me to expand on any of these topics.