Ornament “S-116” from Batalha Monastery
μ-X-ray diffraction patterns were acquired on six points of the ornament “S-116” and a peeled off fragment, as shown in Fig. 2. Point 1, 2, 4 and 6 were acquired on the orange surface patinas whereas Points 3 and 5 were acquired on the substrate where the orange surface had already peeled off with the limestone substrate showing extensive decay. The XRD pattern and peaks are presented in Fig. 4, and a semi-quantitative evaluation is given in Table 1.
In both samples, the stone substrate is an oolitic limestone consisting mainly of calcite and quartz (Ding et al, 2019): accordingly, peaks of both minerals are seen at all points. A considerable amount of gypsum can be detected at point 1, 2, 4, 6 and in the detached fragment, in areas corresponding to the orange surface, while at point 3 and 5 where the coating is completely removed, gypsum is absent. Halite was present in all analyzed spots, with its content being significantly higher at point 3 and 5, indicating that halite is concentrated on the substrate surface under the orange layer. Feldspar peaks (Al-K-silica) are present particularly at point 1, confirming the presence of soil dust deposition in this area. A minor amount of hematite was detected in all orange patinas suggesting the presence of red ochre pigment.
The sporadic presence of ettringite peaks in XRD results is somewhat intriguing. The presence of ettringite could be due to the cementitious materials exposed to sulphate: calcium aluminate hydrates and calcium silicate hydrates mixed with lime, can react with water and gypsum or other sulphate salts, to produce ettringite [Ca6Al2(SO4)3(OH)12·26H2O] and thaumasite [Ca3Si(OH)6(CO3)(SO4) 12H2O] Collepardi (1999). The application of cement-based mortar in the Batalha monastery during past restoration interventions (Soares 2001) could be the reason for the presence of ettringite on the ornament surface.
The peeled-off fragment was observed under the VP-SEM+EDS: results are shown in Fig. 5. There is a correlation between the distribution of sodium and chloride, demonstrating the presence of halite (NaCl) grains. Potassium (K), aluminum (Al) and silicon (Si) peaks imply the presence of quartz, feldspar and clays minerals probably as soil dust. The peak of calcium (Ca) is weak where the NaCl and feldspar / soil dust / clay grain are present, which means halite and aluminosilicates were formed subsequently on the top of the limestone substrate. This could also be seen from the SEM picture, the halite (up left) and feldspar (down central) show a well-formed crystalline habit with relatively large dimension grains (>50 μm), in contrast with the poorly crystallized gypsum grains.
The distribution of iron (Fe) is homogenous over the whole area, supporting the interpretation of a pigment surface application. On the other hand, hematite is also present as accessory mineral grains in the original limestone. The presence of sulfur (S) is mainly due to gypsum (CaSO4•2H2O); its distribution is basically uniform and some collective grains can be seen. By overlapping the distribution of Fe and S (Fig. 6) it shows the hematite was likely to be mixed and applied together with gypsum on the uppermost layer. This hypothesis can be supported by the recorded use of red and yellow ochre: in ancient Macedonian paintings, one of the major means of using ochre is to mix them with lime to produce tint plasters then apply on the monuments (Perdikatsis and Brecoulaki 2008). Therefore, the color on the ornament “S-116” surface could be by reason of the same procedure. It is noteworthy the presence of phosphorous (P), which was also detected by Lazzalini and Salvatori (1989) on the “scialbatura” of Cathedral and church of S.Zeno, Verona, they inferred that the presence of phosphorus suggests an artificial protective coating made from calcium caseinate.
Balustrade from Batalha Monastery Royal Cloister
The balustrade sample (Fig. 5) was analyzed only by XRD as a bulk to prevent causing any damage. Fig. 6 shows the XRD pattern, and in Table 2 the main minerals are listed together with their semi-quantitative evaluation. It is shown that, at point 1, where an orange surface is still visible, there is a high content of calcium oxalate including weddellite and whewellite, together with feldspar and hematite. Gypsum was detected where the orange surface is worn-out and the white substrate appears, the main composition is calcite, and the contents of gypsum and calcium oxalate show a decrease, also hematite was not detected. Point 3 has a different appearance other than point 1, the surface shows a dark orange color and the texture is loose and powdery, the XRD result revealed calcite, some weddellite, feldspar, and minor hematite. Halite was present at point 3 and absent at points 1 and 2.
Calcium oxalate is itself colorless but when organic compounds, mainly derived from the metabolic activity of lichens, fungi and bacteria (from oxalic acid which reacts with calcite to form a thin calcium oxalate film) and other mineral grains (such as quartz, feldspar) are present, the patina may acquire a yellowish-brown hue (Ion et al. 2017). Thus, the color on this balustrade could possibly originate from calcium oxalate mixed with other mineral grains such as soil dust. The stone surface oxalates were generally considered to be the result of lichen excreted oxalic acid that reacts with calcium in the substrate (Del Monte et al. 1987), however, a few recent studies reported that various bacteria can produce oxalic acid even without lichen microbiomes, for instance, Pseudomonas fluorescens, Burkholderia, Bacillus and C. jiangningensis JN53, etc. (Palmieri et al. 2019; Hess et al. 2008; Cheng et al. 2017). In fact, Bacillus (phylum Firmicutes) and Burkholderia (phylum Beta-Proteobacteria) were found in the Batalha Monastery bio-deteriorated stone and the atmospheric environment in the previous study by the authors (Ding et al. 2021). Thus, although no lichen crust was present on this balustrade sample, the possibility of bacteria producing oxalic acid and causing mineral weathering cannot be excluded.
Calcium oxalates may also be deriving from vehicular exhaust emissions (Kawamura and Kaplan 1987) suggested that incomplete combustion of aromatic hydrocarbons (such as benzene, toluene, naphthalene) in the car engines could generate diacids including oxalic acid, cis-unsaturated acids and aromatic acids. Due to the polar nature of these dicarboxylic acids, they preferentially associate with moisture and would react with calcareous stone and form calcium oxalates. As a matter of fact, the Portuguese national highway IC2 run through Batalha only 50 meters west of the monastery, so atmospheric vehicular pollution could not be ruled out as a cause for the development of oxalate patinas on the limestone surface. And the benzoic acid found by Aires-Barros et al (2001) on the external walls of this monument, makes this hypothesis more plausible.
Comparison between the two samples
Though both named “orange patinas”, there are significant differences between the two samples investigated. Table 3 lists these differences of these two samples with respect to composition, appearance and decay patterns. For the ornament “S-116”, the orange layer follows the external shaping, for example, the ridging stripes spreading out from the center, which does not exist on the substrate. For the Royal Cloister balustrade, the orange patina is closely adherent to the limestone substrate instead of a visible separate layering present on “S-116”. And there is an abundant weddellite on the balustrade surface, which was not found on the sample “S-116”.
The decay pattern of these orange layers on the two samples are also distinct. On the ornament “S-116”, the surface layer is flaking, on the white substrate under the peeled-off lamination, higher content of halite was detected. This indicates a typical salt decay process: salt solution evaporated and crystallized at the interface between the coating and the substrate, leading to detachment eventually (Duffy et al. 1996). In this case, the salt responsible for decay is mainly halite (NaCl), its precipitation leading to salt decay in the monastery may probably derive from ground-moisture capillary rise. Considering the Royal Cloister balustrade, the patina is lost preferentially at the ridges and bulges but is more preserved in flat areas where marks of scratches can also be seen. Three possible explanations can be suggested: a) the evaporation rate of salt solutions is higher at the geometrically protruded parts due to a larger specific surface area, resulting in higher crystallization pressure inside the stone pores and more severe decay (Rodriguez et al. 1999); b) protruded areas endure more mechanical abrasion from rain, wind, sands, and human activities, thus accelerating the decay processes; or c) the orange patina was originally formed irregularly on the substrate.
Combing the results obtained, preliminary sketch (Fig. 8) was made to elucidate the different characteristic of the two samples investigated.
Comparison with the previous research
Through the above-mentioned results and analysis, it can be concluded that the orange layer on the Royal Cloister balustrade is similar to the findings by Aires-Barros et al (2001). Such patina is compatible with a “scialbatura” nature for its composition, texture and color, as described by Lazzalini and Salvatori (1989). No titanium oxide and fluorite presence as reported in previous research (Fassina 1995). The existing results are in favor of the conclusion that the presence of calcium oxalates on this monastery could be linked to either vehicular exhaust emissions (Kawamura and Kaplan 1987) or by bacterial activity (Palmieri et al. 2019), but not to oxalic production by lichenous colonization (Del Monte and Sabbioni 1987).
The orange coating on the ornament “S-116” sample matches the description of that on the original apostle statues from the church doorway (Rattazzi et al. 1996). Considering the ornament “S-116” and the apostle statues were both crafted in the 15th century, it is feasible that the same coating procedure was used, despite they were located in different sites of the Batalha Monastery. Although it is commonly believed that in this type of surface layers, the gypsum occurring on calcareous stones is a reaction product of calcareous or silicate stones in a SO2 polluted urban atmosphere (Schiavon 2007), the possibility of artificial gypsum coating cannot be completely excluded (Sánchez 2009). In fact, the homogeneity of the layers and its mixing with hematite, may support the conclusion of the intentional application of a pigmented plaster.