Microcrystals of antimony compounds in lead–potassium and lead glass and their effect on glass corrosion: a study of historical glass beads using electron microscopy

Abstract

Crystalline inclusions of antimony compounds in lead glass of the nineteenth century have been investigated by means of transmission electron microscopy, scanning electron microscopy, X-ray microanalysis, electron backscatter diffraction and microcathodoluminescence. Microcrystallites of orthorhombic $$\hbox {KSbOSiO}_4$$ (KSS) with the sizes ranging from about 200 nm to several micrometers have been detected in lead–potassium glass of turquoise seed beads prone to a glass disease causing the irrecoverable deterioration of beaded articles kept in museums. The KSS crystals have high number density and tend to form large colonies. Crystallites of cubic $$\hbox {Pb}_2\hbox {Fe}_{0.5}\hbox {Sb}_{1.5}\hbox {O}_{6.5}$$ have been detected in stable yellow lead glass beads. Their number density and sizes are much less than those of the KSS particles observed in turquoise glass; they do not form large clusters. We have come to conclusion that KSS precipitates are responsible for the internal strain-induced corrosion of turquoise lead–potassium glass eventually resulting in crumbling of beads to sand particles. The following scenario explains this phenomenon: $$\hbox {K}^+$$ and $$\hbox {Sb}^{5+}$$ used for glass doping form KSS crystallites during glass melting; tensile strain arising in the glass matrix during cooling because of difference in temperature coefficients of linear expansion of glass and KSS crystals gives rise to crack formation and in course of time results in glass falling apart to heterogeneous pieces. Small crystallites of $$\hbox {Pb}_2\hbox {Fe}_{0.5}\hbox {Sb}_{1.5}\hbox {O}_{6.5}$$ cannot induce a sufficient strain to break yellow lead glass, and internal cracks do not arise in this glass during its cooling. This may explain the stability of yellow lead glass.