Abstract: Electron Probe Micro Analyzer (EPMA) is a quick, non-destructive and in-situ technique with high precision, accuracy and sensitivity that is very powerful in micro domain chemical analyses. It is a fundamental research method in modern sciences.
EPMA quantitative analysis acquires element abundance in “quantity” instead of any “weight” or “mass” reflection, thus, it is
more appropriate to use atom proportion and/or atom ratio to examine the result’s correctness, but not simply depend on the mass total in the range of 100±2 wt%. The trace element in material (mineral) is of great importance as a typomorphic indicator for particular (geologic) genetic environment, which serves a key research subject in material science. EPMA is the best approach in (in-situ) analyzing trace elements considering its performance advantages. However, in practice, the trace element measurement often faces various technical difficulties with several unavoidable flaws, particularly with wavelength spectrometer measurement. The measurement’s precision, accuracy, reliability and reproducibility require professional, scrupulous and all-around attentions on analytical condition setting, which should be tested by multiple conditional analyses. Usually, to acquire optimal detection limit and low standard deviation, a simple approach is to set high initial energy (accelerate voltage, beam current) and long measurement time. In the meantime, at least other four major factors should also be paid attention to such as (1) analytical crystal selection in spectrometer, (2) characteristic X-ray spectrum interference and overlap identification and background value adjustment, (3) PHA function initiation and (4) standard material selection and standardization. In the last, we also need to answer the question about how many analyses data are required to ensure the validity and objectivity, i.e. the minimum measurement. It is suggested to apply iterative calculus from statistic mathematics in each measurement task, especially for studies with average content of trace elements in homogeneous samples.

Key words: EPMA, micro elements, WDS, standard, iterative calculus