Hi Colin,
Up until the day he passed, Joe Goldstein used a natural Schreibersite included in a nickel iron meteorite as a P standard for meteoritic phosphides. He maintained that the P was stoichiometric in Schreibersite, so just knowledge of the Fe,Ni ratio was the only thing that needed to be determined (not a very big effect anyway). I don't know which meteorite he used, but it was (and still is) in his own mount along with Fe, Ni, and Co pure metals. As PKa originates from a transition involving the valence band, the peak position changes substantially compared to phosphates. Probably a peak shape effect too, so using an actual (FeNi)phosphide is undoubtedly a little better than just using a phosphate standard and then updating the peak position on the unknown phosphide. I know the community is throwing a lot of shade on these sorts of 'internal' laboratory standards, but one might say that Joe did okay.
Mike J.