Analysis of the elemental composition of glass fragments by LA-ICP-MS method. Part II. subsurface layers in container glass

Laser ablation inductively coupled plasma with mass spectrometry (LA-ICP-MS) is versatile method for direct analysis of elemental composition of glass fragments. Good spatial and depth resolutions of the method allow for mapping and depth profiling of layers within the sample. Container glass is not homogenous material and subsurface layers can be enriched in elements introduced during manufacture (i.a. hot-end coatings) and leached due to corrosion processes in moisture conditions. The research was focused on within- and between-object variability of container glass, especially in relation to elemental composition of subsurface layers. Durability of subsurface layers in container glass was checked during exposition to different environment. The research shows that information about elemental composition of subsurface layers can be additional and useful feature in comparative analysis of glass fragments. Between-object variability of subsurface layers is connected with content of several oxides (SnO2, ZrO2, HfO2, Fe2O3 and CuO). Whereas, within-object variability of glass objects is connected especially with differences in composition of external and internal subsurface layers in glass containers. Moisture, water, and acidic environment lead to changes in elemental composition of subsurface layers. Corrosion changes in glass can generate type I errors. It has been found that corrosion changes primarily affect the content of alkaline components. Nevertheless, it is possible to detect the characteristic components in subsurface layers using the LA-ICP-MS method even after prolonged exposure to corrosive agents. SnO2 film in subsurface layers can be found in the majority of colorless container glasses, while it is a rare feature in green and brown bottles. The increased content of ZrO2 in subsurface layers is a rare feature and it is more important feature during interpretation of physiochemical data.