Damage of tin-glaze tiles due to salt crystallization occurs in museum environments as well as in architectural contexts. While the problem of salt damage is well documented there has been little or no research into the factors that affect the susceptibility of tiles to glaze loss. Glaze detachment is primarily attributed to a poor ‘fit’ between the glaze and ceramic body. A good glaze fit is obtained when there is a good match in the contraction of the glaze and ceramic body on cooling at the end of the firing process (thermal coefficient of expansion), which largely depends on the composition of the glaze and clay.1
Also of importance is good glaze-ceramic bond, which is related to composition in combination with the firing process as well.2 It is known that 17–25% well-dispersed calcium carbonate in the clay will ensure a good glaze fit when using a tin glaze. Many primary and secondary sources mention the importance of the calcium content in obtaining a successful product. As there were no known calcium-rich clay sources in the Northern Netherlands at that time, calcium-rich clays (marl) were imported from England and Flanders from the late 16th century.
While much has been published on the stylistic development of Dutch tin-glaze tiles, there has been very little research into the history of their production, specifically the raw materials used to produce the tiles and the composition of the glazes and clays.
The starting point for this research was a survey of glaze damage on tiles in collections in order to characterise the problem. The present study covers two main fields. On the one hand research is being undertaken into the historic context of production, looking at primary written sources on clay use and import, glaze recipes and firing practices. On the other hand tiles with and without ‘shivering’ are being analysed in order to characterize the damage pattern. The technique being used at present is scanning electron microscopy with energy dispersive x-ray spectroscopy (SEM-EDS). This analytical technique not only provides relatively accurate quantitative information about the main elements in the glaze and ceramic body, but also enables morphological and elemental mapping of the glaze-body attachment. With these results a start has been made in assessing the relationship between damage patterns and historical material use and production. Questions emerging from the initial results are being further assessed through the production of replicas of traditional glaze and clay recipes.
Supervisors and research partners
Prof. dr. Maarten van Bommel, UvA
Em. prof. dr Norman Tennent, UvA
Dr. Luc Megens, Senior Researcher Cultural Heritage Agency of the Netherlands
Peter Oltheten, European Ceramic Work Centre, Den Bosch
1. A greater contraction of the glaze leads to ‘crazing’ or ‘craquelé,’ while a greater contraction of the body leads to glaze detachment or ‘shivering’. In this case the glaze is reasonably well attached to the ceramic surface so that the glaze removes a thin layer of the ceramic surface.
2. A good bond at the ceramic-glaze interface is a result of the diffusion of material from the body to the glaze and the formation of crystals at the interface during firing.