Caracterización química de la vidriera del rosetón del Duomo de Siena (Italia, 1288-1289)

The chemical composition of nine medieval coloured stained glasses from Duomo (Catedral) of Siena, Italy, has been characterized. They come from the rose window elaborated under the drawing of Duccio di Buoninsegna masterwork (1288- 89 AD). This note explains the results obtained by EMPA, representative of bulk chemistry of several coloured glasses (deep green, olive green, yellow, purple, pink, deep blue, light blue, red plaqué and also uncoloured), as well as the associated trace elements (obtained by Induced Coupled Plasma Mass Spectrometry) that represent the chemical fingerprint of these glasses. The studied samples are sodium-calcium glass (chemical compositions in the range 13-14 wt% Na 2 O, 56-64 wt% SiO 2 , 4 wt% MgO, 9-10 wt% CaO, 2,5-4 wt% K 2 O); thus of Mediterranean tradition.This fact has been found by our team in previous studies (stained glasses from the church of Monestir de Pedralbes, Barcelona) for XIV century glass made at least 40 years later. As a general rule, the conservation state of these sodium glasses is good, except for the cohesion of grisaille to glass mesostase. This study allowed to identify three glass groups developed from different raw material formulations: a first group is constituted by deep green, olive green, light blue and yellow glass; a second one comprises by uncoloured, deep purple, deep blue and pink glass; and finally, the plaqué red glass that shows deeply different composition. Pink and yellow glass was produced following the traditional recipe compiled by Theophilus, a monk who lived at the beginning of XII century. This implies a separate process of raw material purification and a careful control of the redox kiln conditions; thus, these glasses can be considered as traditional or technologically not evolved. The deep blue, light blue and deep purple were obtained after the artisan dosed addition of a cobalt salt. Olive green and deep green glasses were produced with addition of copper (together with manganese and iron), previously prepared as a pigment that has as excipient a potassium glass. The use of potassium glass instead of the sodium glass locally produced strongly suggests that the pigment could be bought directly from Middle or Northern Europe markets and/or could be a sub-product of copper ore melting. On the other hand, the production of sodium red ruby plaqué glass, technologically more evolved and comparable to the coeval potassium glass coming from the Central Europe glass factories, would