Respected Madam,

I am currently using CAMECA SX 100 for the analysis of various ceramic, metallic/alloy materials and found that the instrument measures B fairly well with the available pseudocrystals as mentioned by Prof. Essen. I have gathered this experience by analysis boride samples as well as borosilicate glass samples. I have no idea about Be analysis. Li off course can not be determined by EPMA.

It is to be borne in mind that accuracy will depend on the associated elements also. For example if you have B in a matrix of high Z elements then you have to use higher acceleration voltage which will produce large amount of B Ka X-rays which may cross the detection limit for the detectors. Some body told me that it is always advisable to set the acceleration voltage in such a way that we get signal (cps) in the order of 10000.

For O, I observed that 'stoichometric adjustment' yeilds best result for oxides and silicate. Determination of O is difficult in metal-oxide systems where one can have an oxide phase as well as a metallic phase with some oxygen (depending upon the solubility) at the interface. For such complicated situation I follow 'stoichiometric' route for the oxide and 'balance is O' route for the metallic phase. I have rechecked the results with XRD and found that the phases which were suggested on the basis of EPMA analysis matched with the XRD results. So far I have carried out very few analyses in metal-oxide systems and would request all the list members if they can share their experience in this regard.

Hope this information will help you.

With best of regards,

Pranesh Sengupta

Materials Science Division,

Bhabha Atomic Research Centre,

Mumbai, India