Glass by Design? Raw Materials, Recipes and Compositional Data*

Abstract

Compositional analyses of ancient and historic glasses have often been interpreted in terms of the use of specific raw materials in glass manufacture. However, the known inhomogeneity of many glass-making raw materials and the insolubilities of some compounds make any explanation of compositional data problematic. This paper looks at three glass-making alkalis with a view to understanding how the compositions of these raw materials are carried through to the final glass. The chemistry and variability of the raw materials are discussed, as is their contribution to the final glass composition. In addition, the choices and decisions made by glassmakers are acknowledged in the final glass compositions. This combination of factors addresses the complexity of predicting the use of specific raw materials from the finished glass composition.

Interactions between silicate and salt melts in LBA glassmaking

“…Vegetable ash compositions are rather heterogeneous in their composition, particularly as regards their ratios of soda (Na 2 O) to potash (K 2 O) and lime (CaO) to magnesia (MgO). Another important factor is the amount of non-reactive compounds such as sulphates and chlorides present; the latter can reach 50 wt% of the total ash content (Turner 1956b;Brill 1999b;Busz and Sengle 1999;Jackson and Smedley 2004;Jackson et al 2005;Tite et al 2006). The presence of these non-reactive components in the ancient glass batch has been known for centuries and much industrial research and development has focussed on making these salts either useable for glassmaking or to reduce their presence in the batch.…”

Prehistoric copper production and technological reproduction in the Khao Wong Prachan Valley of Central Thailand

“…Though fuel ashes have a variable composition (Jackson et al 2005), they tend to be rich in alkali oxides, particularly calcia. and are known to contribute signifi cantly towards slag formation (e.g.…”

“…This result possibly indicates two separate gangue additions, but also the probably signifi cant contribution of fuel ash, especially considering the elevated levels of magnesia and potash in the samples (Table 6.15 and cf. Jackson et al 2005, Merkel 1990). This situation could result from a higher fuel to mineral ratio in the smelting charge, and/or that the smelt was hotter, and/or was of a longer duration.…”

“…calcia, magnesia, potash, and titania, see Jackson et al 2005, Merkel 1990. This suite of components would typically be regarded as fl uxes, producing a more fl uid melt with less impedance to the agglomeration of copper prills -i.e.…”

Pastoralist iron production on the Laikipia Plateau, Kenya: wider implications for archaeometallurgical studies

“…An accurate mass balance calculation is impossible, due to our lack of reliable compositional data for the ores and fuels employed. However, simply to illustrate the plausibility of this argument at a general level, we can model our data using the following variables: an ore mixture with a nominal FeO content of 75 wt% (to account for the high titania and other gangue), hardwood charcoal with a nominal lime content in the ash of 50 wt% (even higher CaO values are reported in ashes by Jackson et al (2005), with correlated magnesia, strontium and barium, as documented in the Laikipia slag), and a 5 wt% ash yield for the charcoal (as estimated by Crew, 2000). Using these parameters, and a fuel to ore ratio of 2.5:1, we could feasibly produce slag containing 55 wt% FeO and 6.25 wt% CaO, while leaving 20 wt% of the FeO in the ore mixture to be reduced to metal (see Table 4).…”