+Model ARTICLE IN PRESS JECS-9176; No. of Pages 9 Available online at www.sciencedirect.com Journal of the European Ceramic Society xxx (2013) xxx–xxx The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars Elisa Adorni ∗ University of Parma, Department of Civil-Environmental Engineering and Architecture, via G. Usberti 181/A, 43124 Parma, Italy Received 18 February 2013; received in revised form 9 April 2013; accepted 18 April 2013 Abstract In October 2009, a terrible lightning struck the steeple spire of the Parma Cathedral, causing a fire. The fire-fighting operation made possible the discovery of the original spire ceiling made up by dichromatic glazed bricks, white and black, dating from the 14th century. Original materials presented a relevant decay, both for the high temperatures reached during the fire and for lack of maintenance. The research presents the first study of glazed bricks of the 14th century in Po Valley (Italy) with the purpose of collect chemical, mineralogical and petrographic data on the dichromatic glazed bricks. Brick samples with different kind of glazes and mortars exposed at different condition of fire were analyzed. The following techniques were used in the study: X-ray powder diffraction, optical microscopy, scanning electron microscopy analysis, inductively coupled plasma atomic emission spectroscopy and Raman spectroscopy. Glazes, applied on to Ca-rich paste, have a high lead content (41–57 wt%), with an high amount of tin (19–24 wt%) for the white opacified glazes and manganese (about 4.0 wt%) for the black ones. Typological and historical analysis allowed us to define the production technique of bricks and glazes. Mortars are mainly composed of lime binder and carbonate aggregate. © 2013 Elsevier Ltd. All rights reserved. Keywords: Glaze; Brick; Mortar; SEM-EDS; ICP-AES 1. Introduction In October 2009, a lightning struck the spire of the Parma Cathedral (Fig. 1). Following the fire-fighting operation was possible to discover the original roof of the conical spire, built in the 14th century with rounded profile brick elements with white and black glazes (Fig. 2). The precious dichromate roof was covered at the end of 16th century. In literature, the glazing technique of the Asian, Syrian and Egyptian areas was studied extensively: early Islamic,1,2 preIslamic3 and both.4 Several studies have been performed on glazed ceramics from Spain and Portugal.5–8 A detailed description of the history and production techniques of glazing can be found in literature.2,10–12 In particular, some studies analyze high lead glazes and alkali-lead glazes,2 tin-opacifier glazes,4,9 coloring,8,12,13 methods of glaze application, directly over the air-dried ceramic body or after a first ∗ Tel.: +39 0521 905961. E-mail address: elisa.adorni@unipr.it firing (biscuit-fired bodies)2 and the interaction between glazes and the ceramic body.2,5 In Italy most of studies have been performed on glazed pottery of the central regions,14 e.g. Gubbio and Orvieto,15 Siena,16 no studies have been done on glazed bricks in the Po valley. The literature also reports a series of studies to define the firing temperature of ceramics, especially in archeology: mineralogical, textural and chemical transformation of clay-rich materials during firing.17–19 The study presents the first original study of 14th century glazed bricks in Po Valley, North Italy. From the original masonry a number of samples were collected, glazed bricks and mortars, to characterize the original materials of the cover of the ancient spire of the Parma Cathedral by determining the production technique through a mineralogical, petrographic and chemical analysis of bricks and glazes. In particular, the firing process was defined by means of the study of the shape of the bricks and the discovery of some defects of glazing. 14th century and 16th century mortars of the spire were also characterized to define the provenance of the materials and the original recipes. 0955-2219/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019 Please cite this article in press as: Adorni E. The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars. J Eur Ceram Soc (2013), http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019 +Model JECS-9176; No. of Pages 9 2 ARTICLE IN PRESS E. Adorni / Journal of the European Ceramic Society xxx (2013) xxx–xxx Fig. 3. (A) Dimensions and shape of the bricks. (B) and (C) Overlapping model of the bricks. Fig. 1. Panoramic view of the Cathedral of Parma before the fire. The spire still had the copper covering, now removed. 2. Historical note The Cathedral of Parma was edified starting from 1050 on the site of an early Christian cathedral, which dates back to the fifth century. The edifice was built following a unique compositional scheme with three naves completed by apses, a protruding transept and a crypt, in agreement with the tradition of the Romanesque churches in the Po area. Fig. 2. Particular of the conic spire, realized with dichromatic glazed bricks: white and black. The masonry shows deterioration, cracks and erosion of the bricks. The building of the bell tower began in 1284, when the tower of the ancient cathedral was demolished. In 1291 the edifice was progressed to the height of the castle of the bells. Documents attest that the whole mural-architectonic parts of the tower were already ended in 1294.20 According to the Chronicle by Salimbene De Adam,20 the conical spire was built after 1336 and before the 15th century. The model of the conic spire, realized with curved radial bricks (Fig. 2), was common in the Po Valley since the 13th century: see the bell tower of the Cathedral of Piacenza and the San Donnino Basilica in Fidenza, Parma. Other more recent examples are the majolica spires in Sicily, documented since 1579 for the construction of the spire of S. Giacomo in Collesano (Palermo), now collapsed.21 In Parma the primitive glazed brick masonry of the spire endures to the end of the 16th century when, because of the materials deterioration, spire was covered with a copper roofing. The project, by Giorgio Edoari da Erba, was executed in 1596. Since the end of the 16th century to the recent fire the original covering was not visible. The dichromatic brick decoration remained occulted for centuries.22 The conic spire, 15 m high, was realized in masonry with trapeze-shaped quoins, which have a curved external profile disposed in alternated courses (Fig. 3). The external part of bricks is covered by a dichromatic glaze, white and black (Fig. 2). The bicolor courses consist of 4 rows of white bricks and 4 of blacks. Some cuneiform bricks are more tapered to close the circumference of the spire without an excessive amount of mortar and to correct the radial position of the bricks. The structure leans on a polygonal basement with 16 sides, composed by regular courses of bricks which have smooth surface and reduced mortar joints. The masonry is about 33–35 cm thick, i.e. the depth of the bricks which constitute the structure. At the top of the spire a cone of stones (arenaceous stone and Vicenza stone) supports a golden anemograph angel. Please cite this article in press as: Adorni E. The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars. J Eur Ceram Soc (2013), http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019 +Model JECS-9176; No. of Pages 9 ARTICLE IN PRESS E. Adorni / Journal of the European Ceramic Society xxx (2013) xxx–xxx 2.1. Conservation state After the fire, the whole copper cover was removed; except a wooden beam of the basement, nothing of the ancient frame was left. Iron nails remained linked to masonry; they were inserted in the 16th century as anchorage for the cover to the masonry. The ancient masonry structure was in poor state of preservation and, thus, the fire worsened an already advanced deterioration (Fig. 2). The masonry showed deep fractures and some areas presented deep erosions and lacunas. Glazes showed superficial cracks and alterations which were probably due to the high temperature reached during the fire and by the following thermal shock produced by the water used to extinguish it. Glazes have portions of copper (from the cover) recast on the surface, pink shades are present on the surface, probably due to a reducing environment during the fire (the fire burn up in the space between the copper covering and the masonry), the glazes show small surface cracks with detached parts and swellings. Mortar was affected by decohesion (powder consistency) and blackening on the surface caused by the thermal shock and the smoke. 3. Analytical methods Nine samples of glazed bricks of different colors, white and black, six samples of mortars and two sample of plaster were collected at different level from the steeple spire (Fig. 4). 3 Sampling was planned in order to represent the two construction phases of the spire: the building (14th century) and the later intervention of covering (16th century). Glazed bricks and mortars were investigated by petrographic, mineralogical and geochemical methods with the aim of defining glaze, brick and mortar characteristics. The methods of investigation were: - In situ macroscopic observations; in particular, color, structure, glaze-brick interaction, surface aspect, manufacturing techniques, and pathology. - Optical mineralogical microscopy, transmitted and reflected light (Nikon Eclipse LV100 POL), three thin sections for each sample. - Scanning electron microscopy (JEOL 6400, Jeol Ltd., Tokyo, Japan) and energy-dispersive system microanalysis (Oxford Isis 300, Oxford Instruments, Abingdon, UK) on polished thin section, gold or carbon coated. - X-ray powder diffraction (Philips PW-1710 diffractometer, Cu K␣ radiation, [Philips Analytical Inc., Natick, MA]; Step Size: 0.05◦ .; Scan Rate: 0.750000; Scan Mode: Step; λ: 1.540562) with XPowder Ver. 2004.04.47 Pro program (PDF2 data base).23 - Raman spectroscopy (Jobin-Yvon Horiba LabRam Raman micro-spectrometer). - Chemical analysis of major and trace elements by Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES – FUS-ICP analysis method) of the bricks and mortar powder (8 g each sample, binder and aggregate together). Fig. 4. Survey of the spire with sampling location. Please cite this article in press as: Adorni E. The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars. J Eur Ceram Soc (2013), http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019 +Model JECS-9176; No. of Pages 9 4 ARTICLE IN PRESS E. Adorni / Journal of the European Ceramic Society xxx (2013) xxx–xxx Fig. 5. (A) Black drops of black glaze over some white glazed bricks. (B) Reconstruction of the stacked bricks in the kiln during firing. Mortars characteristics, porosity, grain morphology and size, have been determined using optical analysis, according to the Italian Standard Normal 12/83.24 To estimate the aggregate/binder ratio, the Volume % estimation diagrams was used25,26 ; image analysis was used to evaluate the porosity.27 4. Experimental results and discussion 4.1. Production technique The bricks and masonry analyses allowed defining the ancient production techniques. Through the survey of the bricks (Fig. 3A) and removing part of the masonry during the restoration works, the assembly system of the masonry has been defined (Figs. 3B and C). The analysis of the characteristic shape of the bricks and the discovery of horizontal glaze casting over some bricks (Fig. 5A), helped to define the production and firing processes: the bricks, still uncooked, were immersed manually in the glaze, stacked as shown in Fig. 6B and fired in the kiln. Is presumable that, to optimize the available space in the kiln, the bricks were stacked close to each other. These suppositions have been confirmed by mineralogical analysis of the bricks that show an uneven single-firing. Furthermore, the presence of horizontal sagging (Fig. 5A) indicate that the white and black bricks were fired together, stacked on each other. 4.2. Bricks The ceramic paste is a highly calcareous clay, 19–21 wt% CaO and 47–52 wt% SiO2 (Table 1), with an iron oxide contents lower than 5 wt%. The chemical results show a substantially homogeneous composition. In the calcareous clay bodies the iron oxides are incorporated into calcium iron silicates, for this reason the brick paste has a pale pink color, allowing a minor quantity of opacifier into the glaze to cover the brick.2,5,14 Together with the chemical analysis, a mineralogical characterization of the paste was carried out. Powder diffractions reveal significant amounts of quartz, residual calcite, neo-formed diopside and gehlenite, anorthite and sanidine in trace. Trace of neo-formed diopside, gehlenite and anorthite were detected in all the samples. The decomposition of calcite in oxidizing atmosphere occurs at a low temperature, above 800 ◦ C the free CaO can react with clay and SiO2 , forming new phases: gehlenite, diopside and plagioclase (anorthite). Experimental studies showed that gehlenite crystals begins to form at 800 ◦ C through a reaction that occurs at the grain boundaries of CaO, Al2 O3 and SiO2 and from 900 ◦ C it begins to disappear quickly.13,17,28,29 The presence of diopside, anorthite and gehlenite of neoformation is indicative of a firing temperature of about 900–950 ◦ C. The high content of calcite in the samples CD 1, CD 2 and CD 14, together with high temperature phases, may be indicative of an uneven firing, caused by a low circulation of the air, a stacking of bricks (Fig. 5B) or a different position in the kiln with respect to the fire. The clayey matrix of the samples has a mainly homogeneous texture characterized by open and closed pores (pores size: 0.02–4 mm), in a few cases they are of primary origin due to the preparation of the dough, while for the most part are of secondary origin, generated after the release of fluids during decomposition of the same phase which produces them. In the brick body are present quartz crystals, lithic fragments, altered mica minerals, apatite minerals, Fe-oxides and-Ti-oxides. Minerals show zoned compositions, as a result of the firing process, and coronal reaction, indices of a slow kinetic.30 4.3. Glazes Brick samples consist of a portion of ceramic body with a layer of high lead glaze (PbO 41–57 wt%) of two different colors: white and black. Glazes have been investigated by scanning electron microscopy with EDS microanalysis to have information about their chemical (Table 2) and mineralogical features. Please cite this article in press as: Adorni E. The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars. J Eur Ceram Soc (2013), http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019 +Model ARTICLE IN PRESS JECS-9176; No. of Pages 9 5 E. Adorni / Journal of the European Ceramic Society xxx (2013) xxx–xxx Fig. 6. Sample CD 2. Throughout the black glaze are present thin skeletal crystals formed during the firing; the blacks grains are constituted by Si, Ca and Mn (ED spectrum). Thus, the black pigment is attributable to a MnO–FeO compound. Table 1 Major (oxide wt%) and trace elements (wt ppm) in the investigated bricks. Samples Brick CD 2 CD 3 CD 4 CD 8 Mortar CD 13 CD 18 CD 9 CD 12 CD 10 CD 16 SiO2 Al2 O3 Fe2 O3 (T) MnO MgO CaO Na2 O K2 O TiO2 P2 O5 LOI Total Ba 52.59 47.01 51.36 48.36 12.33 10.55 11.65 11.33 4.86 4.31 4.79 4.49 0.118 0.122 0.148 0.113 2.36 2.16 2.43 2.23 19.32 21.69 21.38 19.76 0.91 0.91 1.01 0.77 1.66 1.68 2.38 1.82 0.608 0.516 0.581 0.555 0.19 0.28 0.41 0.22 5.59 8.98 2.25 10.65 100.54 98.21 98.39 100.30 368 352 435 391 26.11 20.67 27.01 24.00 5.30 5.97 3.61 3.05 3.64 3.08 1.29 1.45 2.37 1.76 2.23 1.91 0.57 0.61 0.151 0.123 0.168 0.144 0.021 0.021 1.28 1.00 1.12 1.11 0.98 1.02 36.54 38.61 35.08 36.04 34.01 35.71 0.32 0.36 0.22 0.28 0.23 0.18 0.70 0.58 0.42 0.56 0.36 0.46 0.150 0.133 0.153 0.130 0.060 0.067 0.16 0.18 0.14 0.19 0.26 0.05 29.03 32.5 29.54 30.87 11.49 9.42 100.42 98.97 99.72 98.31 54.57 54.96 198 263 131 205 58 65 Sr Y Sc Zr V 588 672 750 597 25 23 24 24 12 10 11 11 151 139 144 136 82 79 76 88 1043 1054 655 780 1034 1076 12 12 13 12 2 3 5 4 5 4 2 2 49 36 45 41 22 19 44 42 53 58 17 17 wt% = per cent weight; ppm wt = part per million weight; LOI includes H2 O and CO2 ; T = total Fe as Fe2 O3 . The analyses reveal a similar composition of white and black glazes: samples have the same proportion of SiO2 –PbO and differ only for the presence of FeO–MnO in black glazes and tin as opacifier in the whites. The total alkaline content is less than about 1 wt% (Na2 O + K2 O) and Al2 O3 about 1% (except CD 2), suggesting the use in all the sample of a quite pure quartz sand for glazes production5 . White glazes belong to the class of ‘tin-opacified high lead glazes’: in addition to lead, they have a high SnO2 content. Cassiterite crystals were added to improve glaze flux during firing and to confer an opaque shade to the glaze layer.31 Table 2 Energy-dispersive X-ray spectroscopy microanalysis results of glazes (wt%).a Samples CD CD CD CD CD CD CD 1 2 3 6 8 11 13 B B W W W W W Na2 O MgO AlO3 SiO2 K2 O CaO MnO SnO2 FeO PbO 0 0.34 0.61 0.71 0.62 0.71 0.95 0.40 0.90 0.34 0.36 0 0.38 0 1.00 5.17 0.81 1.06 1.27 1.05 2.07 34.60 33.01 29.40 27.64 28.16 30.11 33.75 0 1.41 0 0 0 0 0 2.52 10.02 2.40 2.68 1.99 2.55 2.57 3.80 4.20 0 0 0 0 0 0 0 24.30 21.08 21.94 21.35 19.38 0 2.80 0 0 0 0 0 57.48 42.19 42.14 44.34 46.02 43.85 41.28 a Normalized to 100%. B = black color glaze; W = white color glaze. Please cite this article in press as: Adorni E. The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars. J Eur Ceram Soc (2013), http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019 +Model JECS-9176; No. of Pages 9 6 ARTICLE IN PRESS E. Adorni / Journal of the European Ceramic Society xxx (2013) xxx–xxx Fig. 7. Sample CD 6. (A) Interface: illite grains from the clay body show the growth of lead–potassium feldspar crystallites around. Within the glaze may be noted bright inclusions of tin oxide. (B) Heterogeneous distribution of tin oxide crystals in the glaze. The analysis of the samples showed that the composition and microstructure are similar for all the white glazed, which could suggest a common production technique. The tin content are in the range 19–24 wt% SnO2 , with SnO2 /PbO ratios in the range 0.5–0.6 wt%.14 The content of SnO2 is very high compared to the contemporary productions from Eastern Spain (13–14th centuries), with SnO2 in the range between 6 and 7 wt%,5 to the previous productions from Egypt, with tin oxide rates of 5–9 wt%4 and to the Italian Renaissance glazed pottery form Gubbio and Deruta (about 5.8 wt% SnO2 ).15 SEM-EDS analysis evidence the uniform distribution of cassiterite crystals (SnO2 ), confirmed by Raman Spectroscopy.32 Tin oxide was first dissolved in silica and lead vitreous matrix; during the firing process, tin oxide recrystallized to well develop cassiterite crystals. Recrystallization of cassiterite was dependent on time and temperature. In some samples, CD 3 and CD 13, the punctual SEM microanalysis revealed heterogeneity in the distribution of SnO2 . The study of geological features and the identification of the quarries already existing at the time of the tower construction, allowed to hypothesize that tin used could be sourced from the mines of Campiglia Marittima (Tuscany), not far from Parma.33 Black glazes have an high lead content. From the analyses, the samples have different composition (Table 2): CD 2 has an high content of iron (FeO 2.8 wt%) and alkaline components, instead absent in CD 1. Manganese content (MnO about 4.0 wt%) is similar in the two samples and gives the black color, tinged with brown, to the glaze.34,35 SEM-EDS analysis evidenced the presence, especially near the external edge, of very thin skeletal crystals containing lead, Mn, Fe, and Si, formed during the firing and black grains constituted by Si, Ca and Mn (Fig. 6). Microanalysis and microscopic examination indicated that an opacifier was not present in black glazes. 4.4. Body-glaze interface During firing, interaction between glaze and clay body produces an interface layer, that is strongly related to the clay body typology, the temperature and time of firing.1,2,11 In particular, in dry unfired bodies, elements such as aluminum, iron, potassium, calcium and magnesium diffuse from the body into the glaze and, as the concentration of these elements in the glaze increases, crystals of potassium–lead–aluminum-silicate are formed at the body–glaze interface, this has been detected Fig. 8. The thickness of glaze and interface layer. (A) Sample CD 6, unreacted quartz crystals and elongated potassium–lead–aluminum-silicate crystals are present at the interface. (B) Sample CD 2, extended interface with coal particles, manganese dioxide and trace of Fe-oxide. Please cite this article in press as: Adorni E. The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars. J Eur Ceram Soc (2013), http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019 +Model JECS-9176; No. of Pages 9 ARTICLE IN PRESS 7 E. Adorni / Journal of the European Ceramic Society xxx (2013) xxx–xxx Fig. 9. Glaze damages: (A) fracture parallel to the surface in the sample CD 6; (B) sample CD 4, glaze damages during the fire, with large irregular bubbles and cracks. in all analyzed samples (Figs. 6 and 7A). The development at the interface is significantly greater when the glaze suspension is applied to the unfired body.2 In the samples, tin-oxide crystals are heterogeneously distributed, suggesting a non-fritting process for the glaze production or a not accurate preparation technique5 (Fig. 7B). Morphological analysis of these crystals and their “dragging” into the glaze, allowed to hypothesize that, in the production process, the cooling of the ceramics has been slow, probably inside the kiln.36 Many sub-rectangular particles corresponding to quartz in form of ‘unreacted quartz’ are detected at the interface glazebrick, these depend on the glaze composition, in this case the SiO2 content is 12.9–16.9 wt%. In all samples, calcium incorporated into the glaze formed Ca-rich pyroxene. The noticeable amount of bubbles in the glaze, produced by release of carbon dioxide generated by decomposition of carbonate of the clay, and the large amount of crystals developed at the interface indicate a single firing process (Fig. 8). In the samples, the thickness of glaze and interface layer is variable and irregular (Fig. 8): (1) glaze thickness: 30–700 ␮m and (2) interface layer: 20–150 ␮m. White glazes have a higher thickness than the black ones, with a more developed interface layer. A particular case concerns the sample CD 4, profoundly altered during the blaze. The sample shows signs of re-firing, with swellings and large irregular bubbles in the glaze and an accentuated vitrification of the brick (Fig. 9B). Notwithstanding the high lead glazes have a low coefficient of thermal expansion than the alkali lead glazes, the few cracks present in the glaze are probably due to the shrinkage of the ceramic body during firing or to thermal shock suffered during the recent fire (Fig. 9A). 4.5. Mortars Joint mortars and repair mortars have been investigated by optical microscopy, electron microscopy (SEM-EDS), ICP-AES analysis and X-ray diffraction (Table 3) to have information on their chemical (Table 1), mineralogical and petrographic features.37,38 Plaster samples were took in correspondence of one of the ancient scaffolding hole; mortar samples were taken at different heights of the spire, sampling points are reported in Fig. 5. Samples, despite having different compositions and mixtures, are mostly composed of carbonates both in aggregate and binder. On the basis of the sampling point, two main typologies of mortars could be identified: mortars from the original masonry, 14th century, collected inside the masonry; samples taken in correspondence of the covering structure installed in the 16th century. The ancientest mortars (CD 13, CD 15, CD 18; 14th century) are composed of a lime-based binder with carbonate and silicate aggregate, in a 1:1 ratio. The interaction between binder and aggregate is poor and reaction rims are not developed around grains boundaries. Small amounts of phyllosilicates, talc and Table 3 Principal mineralogical composition of the mortars (XRD). Sample CD CD CD CD CD CD 13 15 9 12 10 16 Phases Anhydrite Calcite Chlorite Gypsum Phyllosilicates Quartz – t – t +++ + +++ +++ +++ +++ + ++ – – t t – – t + t – +++ +++ ++ t + + – – +++ +++ ++ ++ – – +++ = major; ++ = small; + = minor; t = trace. Please cite this article in press as: Adorni E. The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars. J Eur Ceram Soc (2013), http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019 +Model JECS-9176; No. of Pages 9 8 ARTICLE IN PRESS E. Adorni / Journal of the European Ceramic Society xxx (2013) xxx–xxx Fig. 10. (A) Sample CD 13. Microfossils (Pliocene gastropoda) (optical microscope, 10×, transmitted light, parallel polarizers). (B) Sample CD 9. Ophiolite grain found in the mortar of the end of the XVI century (optical microscope, 10×, transmitted light, crossed polarizers). microfossils are found in all three samples (Fig. 10A). By optical analysis and visual estimation,25,26,38 the size of the grains (aggregates size: 0.2–3 mm) and the pores (macropores size: 0.1–2 mm) are homogeneous. Mortars, CD 9 and CD 12, employed during the installation of copper roofing in the 16th century, are very similar in composition and grain size distribution (aggregates size: 0.08–6 mm), with a good interaction between binder and aggregate, as observed at the carbonate–phyllosilicates interface; high porosity (macropores size: 0.1–0.8 mm) and fracture are present. The ratio between binder and aggregate is visually estimated to 1:2. Ophiolite grains are present in both the samples, reveling the use of aggregate maybe from Parma Apennine area (Fig. 10B). As shown by chemical analysis, for samples CD 9, CD 12, CD 13, CD 18, CaO is mostly related to carbonates suggested by the high value of LOI. Plaster samples CD 10 and CD 16 have predominant content of gypsum and anhydrite (Table 3), as observed by XRD and SEM-EDS analysis and also showed in the chemical analysis by the high content of CaO as well as low LOI and high Sr values (Table 1). In the samples, the dimension of the aggregates varies from 0.1 to 7 mm, with a macroporosity (size > 1/16 mm) of 10–15%. The binder and aggregate ratio is 1:2. The presence of anhydrite in CD 15, CD 12, CD 16 and especially in CD 10 (Table 3) could be related to the gypsum dehydration as effect of the recent fire. 5. Conclusions The results of this study allow characterizing the dichromatic glazed masonry of the conical spire of the steeple of the Parma Cathedral and defining the production techniques of the 14th century glazed bricks. Samples consist of a portion of calcareous clay body with a layer of high lead glaze of two different colors: white and black. White glazes belong to the class of ‘tin-opacified high lead glazes’, analysis reveal an high tin content (19–24 wt% SnO2 ), probably extracted from Campiglia Marittima tin mines in Tuscany (Italy), not far from Parma. The high content of SnO2 distinguishes this production from the contemporary from Eastern Spain, from the Italian Renaissance glazed pottery and from the previous Egyptian production, which have lower values (SnO2 about 5–9 wt%). For black glazes, the color tinged with brown is conferred by the addition of manganese, without trace of opacifier. The samples have different composition: compared to CD 1, CD 2 contains iron (2.8 wt%) and alkali. The calcareous paste has a low content of iron oxides and a pale pink color. Traces of neo-formed gehlenite, diopside and anorthite, detected by XRD, together with calcite suggest a uneven firing temperature, about 900–950 ◦ C. Little difference of firing temperature could be probably due to poor air circulation in the kiln and closeness between the bricks. The interface between brick and glaze has a variable thickness of 20–150 ␮m, in general higher in the white glazes. The thickness of the interface and the presence of crystals of potassium–lead–aluminum silicate are indicative of a single cooking, with glaze applied on an un-fired body. Unreacted quartz are also present at the interface and into the glaze, together with bubbles generated by carbonate decomposition of the clay. Mortars of the 14th and 16th centuries have been analyzed. All the samples are composed of a lime-based binder with different aggregate, in particular: the mortar of 14th century (CD 13, CD 15, CD 18) has a low content of silicate, small amounts of phyllosilicates, talc and microfossils and poor interaction between binder and aggregate; the mortars of the 16th century (CD 9 and CD 12) are very similar in composition and grain size distribution, with a good interaction between binder and aggregate, aggregates contain ophiolite grains, probably deriving from Parma Apennines. Plaster samples, CD 10 and CD 16, have predominant content of gypsum and anhydrite, the latter could be related to the gypsum dehydration as effect of the recent fire. Acknowledgments Thanks to Silvia Simeti and Stefano Volta of Archè Restauri society for the essential and kind assistance in the building site. A special thanks to professor Judit Molera of the University of Vic for her essential advices and corrections and to professor Please cite this article in press as: Adorni E. The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars. J Eur Ceram Soc (2013), http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019 +Model JECS-9176; No. of Pages 9 ARTICLE IN PRESS E. Adorni / Journal of the European Ceramic Society xxx (2013) xxx–xxx Giampiero Venturelli of the University of Parma for his suggestions during the research. References 1. Kleinmann B. History and development of early Islamic pottery glazes. In: Olin JS, Blackman MJ, editors. Proceedings of the 24th international archaeometry simposium. Washington, DC: Smithsonian Institution Press; 1986. p. 73–84. 2. Tite MS, Freestone I, Mason R, Molera J, Vendrell-Saz M, Wood N. Lead glazes in antiquity – methods of production and reasons for use. Archaeometry 1998;40(2):241–60. 3. Moorey PRS. Materials and manufacture in ancient Mesopotamia: the evidence of archaeology and art. Oxford: British Archaeological Reports, International Series; 1985. p. 237. 4. Mason RB, Tite MS. The beginnings of tin-opacification of pottery glazes. Archaeometry 1997;39(I):41L 58. 5. Molera J, Pérez-Arantegui J, Vendrell-Saz M. Chemical and textural characterisation of tin glazes in Islamic ceramics from Eastern Spain. J Archaeol Sci 2001;28:331–40. 6. Molera J, Vendrell-Saz V, Garcìa-Vallés M, Pradell T. Technology and colour development of hispano-moresque lead-glazed pottery. Archeometry 1997;39(1):23–39. 7. Coentro S, Mimoso JM, Lima AM, Silva AS, Pais AN, Muralha VSF. Multi-analytical identification of pigments and pigment mixtures used in 17th century Portuguese azulejos. J Eur Ceram Soc 2012;32: 37–48. 8. Guilherme A, Coroado J, Dos Santos JMF, Lühl L, Wolff T, Kanngießer B, et al. X-ray fluorescence (conventional and 3D) and scanning electron microscopy for the investigation of Portuguese polychrome glazed ceramics: advances in the knowledge of the manufacturing techniques. Spectrochim Acta B: Atomi Spectrosc 2011;66(5):297–307. 9. Cuomo Di Caprio N. Ceramica in archeologia 2: antiche tecniche di lavorazione e moderni metodi di indagine. Roma: L’Erma di Bretschneider; 2007. 10. Kingery WD, Vandivier PB. Ceramic masterpiece: art, structure technology. New York: Macmillan; 1986. 11. Molera J, Pradell T, Vendrell-Saz M. The colours of Ca-rich ceramic pastes: origin and characterization. Appl Clay Sci 1998;13:187–202. 12. Bevilacqua N, Borgioli L, Adrover Gracia I. I pigmenti nell’arte dalla preistoria alla rivoluzione industriale. Il Prato ed. Padova; 2010. 13. Ricci C, Borgia I, Brunetti BG, Sgamellotti A, Fabbri B, Burla MC, et al. A study on late medieval transparent-glazed pottery and archaic majolica from Orvieto (central Italy). Archaeometry 2005;47(3): 557–70. 14. Tite MS. The production technology of Italian maiolica: a reassessment. J Archaeol Sci 2009;36:2065–80. 15. Viti C, Borgia I, Brunetti B, Sgamellotti A, Mellini M. Microtexture and microchemistry of glaze and pigments in Italian Renaissance pottery from Gubbio and Deruta. J Cult Herit 2003;4(3):199–210. 16. Fortina C, Santagostino Barbone A, Turbanti Memmi I. Sienese ‘archaic’ majolica: a technological study of ceramic bodies and coatings. Archaeometry 2005;47(3):535–55. 9 17. Cultrone G, Rodriguez-Navarro C, Sebastian E, Cazalla O, De La Torre MJ. Carbonate and silicate phase reactions during ceramic firing. Eur J Miner 2001;13(3):621–34. 18. Trindade MJ, Dias MI, Coroado J, Rocha F. Mineralogical transformations of calcareous rich clays with firing: a comparative study between calcite and dolomite rich clays from Algarve, Portugal. Appl Clay Sci 2009;42:345–55. 19. Rasmussen KL, De La Fuente GA, Bond AD, Korsholm Mathiesen K, Vera SD. Pottery firing temperatures: a new method for determining the firing temperature of ceramics and burnt clay. J Archaeol Sci 2012;39(6):1705–16. 20. De Adams S. Chronica fr. Salimbene parmensis ordinis minorum ex codice bibliothecae Vaticanae nunc primum edita. Parma: Ex offina Petri Fiaccadorii; 1857. 21. Fatta G, Vinci C. Cuspidi maiolicate nell’architettura religiosa minore in area siciliana in Aymerich C. Firenze: Architettura di Base; 2007. 22. Testi L. La Cattedrale di Parma. Bergamo: Istituto Italiano Arti Grafiche; 1934. 23. Martin JD. Using XPowder: a software package for powder X-ray diffraction analysis; 2004 www.xpowder.com 24. Normal 12/83. Aggregati artificiali di clasti a matrice legante non argillosa: schema di descrizione. Roma: CNR – ICR; 1983. 25. Ricci Lucchi F. Sedimentologia parte I: Materiali e tessiture dei sedimenti. Bologna: Clueb; 1980. 26. Myron Best G. Igneous and metamorphic petrology. Oxford: Blackwell Publishing; 2003. 27. Miriello D, Barca D, Bloise A, Ciarallo A, Crisci GM, De Rose T, et al. Characterisation of archaeological mortars from Pompeii (Campania, Italy) and identification of construction phases by compositional data analysis. J Archaeol Sci 2010;37:2207–23. 28. Riccardi MP, Messiga B, Duminuco P. An approach to the dynamics of clay firing. Appl Clay Sci 1999;15:393–409. 29. Tarvornpanich T, Souza GP, Lee WE. Microstructural evolution in claybased ceramic I: single components and binary mixtures of clay, flux, and quartz filler. J Am Ceram Soc 2008;91(7.). 30. Messiga B. Reaction pathway in conventional ceramics. In: Engineering mineralogy of ceramic materials, Proceedings of the international school earth and planetary sciences. 2001. 31. Molera J, Pradell T, Salvado N, Vendrell-Saz V. Evidence of tin oxide recrystallization in opacified lead glazes. J Am Ceram Soc 1999;82(10):2871–5. 32. Bouchard M, Smith DC. Catalogue of 45 reference Raman spectra of minerals concerning research in art history or archaeology especially on corroded metals and coloured glass. Spectrochim Acta A 2003;59:2247–66. 33. Strobel G. Antiche miniere di stagno. In: Bollettino di paletnologia italiana, vol. 4.; 1879. p. 28–9. 34. Caruso N. Ceramica viva. Milano: Hoepli; 2007. 35. Mangone A, Giannossa LC, Laganara C, Laviano R, Traini A. Manufacturing expedients in medieval ceramics in Apulia. J Cult Herit 2009;10:134–43. 36. Molera J, Pradell T, Salvado N, Vendrell-Saz V. Interactions between clay bodies and lead glazes. J Am Ceram Soc 2001;84(5):1120–8. 37. Carò F, Riccardi MP, Mazzilli Savini MT. Characterization of plasters and mortars as a tool in archaeological studies: the case of Lardirago Castle in Pavia, Northern Italy. Archaeometry 2008;50:85–100. 38. Miriello D, Crisci GM. Image analysis and flatbed scanners. A visual procedure in order to study the macro-porosity of the archaeological and historical mortars. J Cult Herit 2006;7:186–92. Please cite this article in press as: Adorni E. The steeple spire of the Parma Cathedral: An analysis of the glazed bricks and mortars. J Eur Ceram Soc (2013), http://dx.doi.org/10.1016/j.jeurceramsoc.2013.04.019