Accuracy improvement by means of porosity assessment and standards optimization in SEM-EDS and XRF elemental analyses on archaeological and historical pottery and porcelain

SEM-EDS and XRF analyses of pellets produced with powdery clay and ceramic standards and fired at increasing temperatures showed a systematic overestimation of the abundance of heavier detected elements (K, Ca, Ti and Fe) using the conventional procedure of calibration with massive mineral/glass certified materials followed by normalization of the detected values. Errors were particularly noticeable for samples fired in the typical range of temperatures of archaeological and historical pottery (600-900 °C) and for unfired samples, and were attributed to material porosity. An extremely simple method based on the SEM-BSE image analysis is proposed for the semi-quantitative evaluation of porosity. A remarkable increase of accuracy, especially for SEM-EDS, was evidenced when the calibration is performed using a standard with porosity comparable to the samples, with regard to the pottery temperature range. Conversely, for the analysis of high-fired samples simulating porcelain (1200 °C) no substantial difference was observed with respect to the usual massive minerals/glass calibration. Finally, results showed the unsuitability of calibration performed by means of unfired pellets, for both pottery and porcelains.