Deterioration of building materials and artworks in the ‘Santa Maria della Stella’ church, Saluzzo (Italy): causes of decay and possible remedies
Roberto Giustetto, 1,2 Elena Maria Moschella, 1 Mariano Cristellotti,3 Emanuele Costa1 1Department of Earth Sciences, Università di Torino, via Valperga Caluso 35, 10125 Torino (Italy)
2NIS - Nanostructured Interfaces and Surfaces, via Quarello 11, 10135 Torino (Italy)
3Restorer, Operative Director Cristellotti & Maffeis S.r.l., Loc. Ceretto 9/a, 12024 Costigliole Saluzzo (CN; Italy) *Corresponding author: roberto.giustetto @ unito.it – tel. +39-011-6705122
Abstract
An in-depth scientific survey brought to understand the decay mechanisms affecting the ‘Santa Maria della Stella’ church in Saluzzo. Rainwater seepages permeating the vault and interiors caused: i) epsomite growth as interstitial columnar crystals (causing pictorial coat detachment) or superficial, floury-like efflorescences; ii) formation of nesquehonite/hydromagnesite crusts on wall paintings; iii) nitratine growth causing pigment staining and detachment. These processes involve selective Mg2+ mobilization from magnesian-lime mortars and bacterial-induced formation of
nitrates from guano, with consequent precipitation of degrading salts. The study confirms how characterization of all decaying agents is fundamental to plan a viable Cultural Heritage conservation and restoration.
Key-words: Magnesian lime; Efflorescence; Incrustation; Epsomite; Nesquehonite; Dypingite;
Hydromagnesite; Nitratine.
Introduction
The use of artificial building materials for masonry constructions has been known since the ancient Egypt. Mortars (for binding blocks and/or filling/sealing irregular gaps) and plasters (for coating walls and ceilings) are composed of an inorganic binder mixed to a bunch of aggregates (sand and pebbles). The former undergoes particular transformations when melt to water, conferring plasticity
during laying and sturdiness afterwards. Different additives can be added to the mixture depending on the type of compound, its installation and setup and environmental conditions. In historic mortars and plasters, the inorganic binder is generally based on calcium or magnesian limes – or a mixture of them both, sometimes with gypsum.
Calcium lime is obtained after calcination of crushed limestones or chalks, mainly composed of calcite (CaCO3) or gypsum (CaSO42H2O) respectively. The so obtained quicklime (CaO) is then
slaked by water addition, with formation of portlandite [Ca(OH)2]. High-Ca lime is not hydraulic
and sets very slowly through carbonation with atmospheric CO2 and moisture, thus (re-)forming
calcite. Magnesian lime can be obtained from the calcination of dolomitic limestones, essentially made of dolomite [CaMg(CO3)2]. European Standard EN 459-1 (2001) defines dolomitic limes as
consisting of CaO/MgO or Ca(OH)2/Mg(OH)2, with no additions of hydraulic or pozzolanic
materials. The production of Mg-lime slightly differs from that of high-Ca lime, as the calcination of dolomitic limestone needs lower temperatures (510-750°C) than calcitic limestone (900°C) (Oates, 1998). Also, the slaking/carbonation of Mg-lime is different as MgO shows slower
hydration rates than CaO in presence of water due to particle size distribution and stirring (Lanas & Alvarez, 2004a). Besides, formation of newly-crystallized dolomite, coupled to calcite, enhances the system complexity changing both its mineralogy and microstructure (Kumar et al., 2007). The properties of magnesian lime-based materials, compared to high-Ca mortars, are not yet established. Some authors claim they have scarce hydraulicity, producing poor quality materials submitted to fractures (Cowper, 1927). Others state that they represent high quality compounds with hydraulic properties (Burn, 1871). Recently, Chever et al. (2010) proved that masonry mortars obtained from Mg-limes should shrink further, need less water and have higher mechanical strength than the analogous Ca-lime-based ones, not leading to cracking. This higher strength might also be favoured by calcite formation through de-dolomitization due to carbonation of portlandite (Lanas et al., 2004). Other experimental evidences suggest that magnesian lime mortars would perform
superiorly in areas subject to presence of moisture, due to lower capillary raise, porosity and absorption (Pavia et al., 2005).
Some lime-based mortars basically maintain unaltered their aesthetic and mechanical features over hundreds of years whereas others undergo severe decay even after few decades, requiring proper interventions. One deterioration source is represented by mineralization of salts in the form of efflorescences (Morillas et al., 2015). Although facades are usually exposed to worst damages, often degradation phenomena involve also the building interiors. In addition to the materials quality, these processes may also concern environmental and/or anthropic parameters, such as urban/industrial pollution.
This study deals with the decay affecting the building materials and artworks of the ‘Santa Maria della Stella’ church in Saluzzo, North-western Italy (Fig. 1.a). An in-depth scientific survey was performed to try and understand the causes of deterioration, consisting in diffuse efflorescences and incrustations covering the walls surface and/or pervading the mortars and plasters interstices. Finally, a viable modus operandi for a restoration intervention is proposed.
Experimental and case study
Case study and materials
(INSERT FIGURE 1)
The Church of ‘Santa Maria della Stella’ (‘St. Mary of the Star’) is part of a complex comprehending two buildings, a monastery and facilities grown around an original nucleus, now incorporated in the monastic choir, built in 1611 for the ‘Clarisse’ nuns of Rifreddo (Comba, 1999). The monastery was abandoned at the beginning of the XIX century due to an edict of the French National Constituent Assembly which suppressed all monastic and religious orders. In 1873, after a temporary location by the Confraternity of the Red Cross, a small group
of Jesuits purchased the building (Monnier, 1928) and held it until 1958. The complex was then abandoned and used as a storehouse of furniture and fittings until today.
In the period between the Napoleonic suppression and its abandonment, the complex underwent several modifications for what concerns the wall decorations and architectural elements, which caused the superposition of several pictorial coatings and ornaments. The dome, though somewhat restored in late 1800, has maintained its original decoration together with three mural paintings in the central nave and three pendentives reporting the effigies of the Fathers of the Church. Lack of maintenance and abandonment brought to the breaking of the roof and coverings, causing rainwater infiltrations. The facade and most of the interiors, especially the central nave and the vault, underwent an intense decay due to weathering conditions with damages to mortars, plasters and mural paintings. In 2013 the “Fondazione Cassa di Risparmio di Saluzzo”, current owner of the structure, embarked on a recovery and restoration project aimed to render this Church its original beauty and importance.
The worst situation involves the central vault, inner portion of the cupola and adjoining apse, positioned almost 20 m above the floor. Only minor signs of decay were visible before 1995, but nowadays the vault and apse are mostly covered by saline blooms and/or mineral
incrustations on those areas exposed northbound (Fig. 1.b). Similar inconveniences affect the supporting pillars and adjoining walls, though the crusts tend to gradually fade proceeding towards the ground (Fig. 1.c). The incrusting materials represent both an aesthetic and structural damage, hiding the decorative motifs and also favouring deterioration of mortars, plasters and artworks, often causing fragments detachment.
Three different kinds of alterations were individuated, marked by specific features
distinguishable even to the naked eye on different areas of the vault, pinnacles, architrave and pillars. Their distribution is indicated in Fig. 2 (a: vault; b: northbound-oriented side wall) with separate colours: i) green areas indicate diffuse growth of whitish, translucent saline
preparatory level; in the orange areas, analogous floury-efflorescences homogeneously cover the surface of some columns, capitals and above architrave; ii) in the red areas, a peculiar whitish crust inhomogeneously covers the surface of mural paintings on a pinnacle surrounding the cupola; iii) the blue regions indicate presence of discontinuous crusts on an arch beside the vault, causing detachment and fall of the pictorial coat.
(INSERT FIGURE 2)
More than 50 representative specimens (surface < 1 cm2) were collected from all degraded
areas. Sampling was performed in agreement with the regulations for conservation and restoration of Cultural Heritage (NorMal 1/80, 1980). To study their stratigraphy, some specimens were incorporated in cold-polymerized polyester resin and the resulting pellets cut transversally and polished with respect to the sample surface.
Methods
A multidisciplinary archaeometric approach was adopted to study the constituent materials of the decorative apparatus and their degradation products. By combining all information, a thorough stratigraphic description of the deteriorated building materials and artworks was obtained. Incidental presence of organic matter was also checked to evaluate presence of biodegrading agents.
A stereoscopic photo-microscope Leica MZ16 was used for microscopic observations. Scanning Electron Microscopy was performed with a SEM Stereoscan 360, Cambridge Instrument, coupled with an EDS Link Pentafet Oxford instrument equipped with a ‘thin window’ detector allowing qualitative/quantitative chemical analyses of light elements (down to Boron).Working parameters: acceleration voltage 15 kV; working distance 25 mm; probe current 1 nA; spectra acquisition time varying from 60 to 300 s. Standardization was performed using a pure Co specimen. Data were collected on polished, carbon coated samples, processed
with the Inca 200 Microanalysis Suite Software, version 4.08 and calibrated on natural mineral standards using the ZAF correction method.
X-ray powder diffraction (XRPD) data were collected on specimens hand ground in an agate mortar and mounted on a zero-background Si-monocrystal flat sample holder, using an automated PW3050/60 PANalytical X’Pert-PRO diffractometer in Bragg-Brentano geometry, with - setup, Real Time Multiple Strip detector and Cu-K radiation. Data were analysed in the 3-70° 2 range with the Diffrac Plus (2005) software (EVA 11,00,3).
FT-IR and Raman spectra were collected on a Brucker Vertex 70 FT-IR, equipped with an ATR attachment, and with a Horiba Jobin Yvon HR800 micro-Raman system excited with a Nd laser emitting at 532 nm (green) equipped with an Olympus BX41optical microscope, respectively. Scattered signals were collected in the 100-1800 cm-1 range, 2m lateral
resolution, by calibrating on the 520.6 cm-1 absorption band of Si and using a CCD detector at
180° with respect to the incident beam. The Labspec 5 software was used for smoothing and baseline. Data were interpreted with the Fityk software (Wojdyr, 2010) using the RRUFF database (Downs, 2006).
Results
Optical and electron (SEM) microscopies allowed a sharp description of the samples stratigraphy (deeper layers and superficial coatings), identifying also those agents responsible for the decay and their relationships with building materials. In addition, microprobe (EDS) analyses allowed semi-quantitative estimates of their chemical composition. XRPD data were collected separately on each stratigraphic layer, so to identify all related materials and degradation byproducts. FT-IR and Raman spectroscopies were used to investigate further aspects. The obtained results are described hereafter for each kind of alteration.
Floury efflorescence on the vault, pillars and walls
Two distinct stratigraphic sequences were observed on vast areas of the northbound-oriented side wall and vault (Fig. 1.b and green areas in Fig. 2.a and b) and on some columns, capitals and architrave on the same side wall (Fig. 1.c and orange areas in Fig. 2.b).
(INSERT TABLE 1)
a. Those areas coloured in green in Fig. 2, where swelling of pictorial coat/decorations and eventual detachment has occurred, show a peculiar stratigraphy. Optical microscopy shows that in these specimens, representative of the mechanical decay affecting most wall
paintings on the cupola, saline efflorescences with translucent/crystalline aspect expand between the deeper (mortar with fine grained aggregates) and the more superficial layers (pictorial coat and underlying plaster finish; Fig. 3.a). SEM images in back-scattered electrons (BSE; Fig. 4.a) confirm that these salts consist of columnar crystals growing perpendicular to the stratification, whose interstitial expansion favours mechanical separation of the surface from the deeper substrata. This process implies gradual swelling of the superficial layers, detachment and fall to the ground. EDS analyses proved these crystals to be essentially made of Mg sulphate, identified by XRPD as epsomite
(MgSO4•7H2O; Fig. 5.a).
The superficial paint coat is usually formed by sparse and finely ground pigment grains (i.e. green and/or red ochres) showing no apparent decay, sometimes dispersed in Ca and Mg-carbonates matrices (‘a calce’ glazing) or bound to organic materials (identified as egg proteins by ATR FT-IR; ‘tempera’ glazing). Beneath the pictorial coat, a thin (50 m) Ca/Mg-carbonate preparatory level is occasionally observed, seldom substituted by a Ca-sulphate layer identified by XRPD as gypsum. This is followed by an underlying plaster
finish (≈ 1 mm thick) made of calcite, subordinate dolomite and no (or few) aggregates (Fig. 4.a).
The deeper mortar always shows presence of a calcitic binder, at times coupled to
subordinate dolomite (ratio 2:1), in which a typical silicate sand is dispersed with grains of different nature (Table 1). ATR FT-IR detected no organic matter; small S amounts, revealed by EDS throughout all the profile, possibly indicate ubiquitous presence of sulphates.
(INSERT FIGURE 3)
b. The orange area in Fig. 2.b is marked by the growth of analogous powder-like saline efflorescences, forming a thick and translucent upholstery homogeneously coating all underlying layers (Fig. 3.b). When observed with SEM, these superficial efflorescences consist of small (few m) crystal aggregates with typically rounded surfaces (Fig. 4.b), identified by EDS and XRPD as epsomite (Table 1). However, no pictorial coat nor preparatory level is found underneath the floury salts but a coarse mortar at times topped by a white plaster finish, with no remarkable sign of degradation. The mortar, similar to that described above, consists of a whitish carbonate cement in which abundant millimetric to sub-millimetric silicate aggregates are dispersed, occasionally topped by a thinner layer with fine-grained, darker grains. The plaster consists of a Ca/Mg-carbonate binder
(essentially calcite with 40-50 % dolomite) incorporating tiny grains with the same composition. Again, gypsum is sporadically detected (Table 1).
(INSERT FIGURE 4)
The surface of those specimens collected from the damaged pinnacle (red areas in Fig. 2.a and b) shows discontinuous crusts with sub-millimetric thickness, growing on the pictorial coat(s) (Fig. 3.c). Although appearing homogeneous in BSE-SEM images (Fig. 4.c), these
incrustations contain Mg, O, C (the last one inferred by comparison with the analogous peak on the Co sample used for chemical standardization) and Ca traces at EDS, suggesting a possible carbonate and/or oxalate composition. XRPD proved them to be made of a mixture of
nesquehonite (MgCO3•3H2O), subordinate hydromagnesite [Mg5(CO3)4(OH)2•4H2O] (Fig. 5.b)
and occasional dypingite [Mg5(CO3)4(OH)2•5H2O], together with minor quartz, calcite,
dolomite, muscovite and plagioclase, related to the underlying substrata.
The stratigraphic sequence is peculiar, as beneath the pictorial coat(s) no plaster finish nor preparatory level are observed, but a coarse mortar including abundant millimetric to sub-millimetric aggregates (quartz, micas, feldspars, etc.; Table 1) in a scarce, carbonatic cement (calcite, with few or no dolomite).
(INSERT FIGURE 5)
Paint detachment on the arch mural paintings
Most wall paintings on an arch beside the vault (blue area in Fig. 2) are damaged by loss of pictorial coat and colour fading. EDS and XRPD proved the blue pigments to be made of ultramarine blue (possibly sodalite) and bone black grains, dispersed in a Ca/Mg-carbonate matrix. BSE-SEM images show presence of small rhombic-shaped crystals (≈ 20 m; Fig. 4.d) in the pictorial coat and preparatory level, undistinguishable from euhedral calcite at the optical microscope. These crystals, containing Na and N at EDS, were identified by micro-Raman as nitratine (NaNO3; Fig. 6). Both layers also show presence of small Mg/Ca sulphates amounts,
Underneath the pictorial coat lies a thin ( 50 m) calcite-rich preparatory level, followed by a thicker (1.5 mm) white plaster finish where a Ca/Mg-carbonate binder includes translucent calcite and dolomite aggregates. A coarse mortar lies underneath, made of a carbonate binder and millimetric to sub-millimetric silicate aggregates (mainly quartz, feldspars and micas; Fig. 5.d).
(INSERT FIGURE 6)
Discussion
Several degradation processes are active in the ‘St. Maria della Stella’ church, consequent to the state of neglect affecting the building whose maintenance and refurbishment has been ignored in the last decades. A fundamental role is performed by water, in the form of rain, moisture or capillary climbing, which painstakingly infiltrated all degraded materials, favouring a mechanical weakening and supporting those chemical reactions responsible for degradation.
Two out of three decay mechanisms involve minero-chemical aspects, namely the formation of Mg salts – sulphates and carbonates. This is strictly related to the composition of binders in mortars and plasters, obtained by magnesian limes with co-occurrence of calcite and dolomite. Also, aggregates in the white plaster finish were derived from crushing of marbles or
calcitic/dolomitic rocks. Despite coexistence of Ca and Mg, only the latter participated to the reactions which formed degradation byproducts, whereas the former was not actively involved. However, the action of biological agents is not to be ignored. Both aspects must be considered to plan a viable restoration intervention.
Sulphate attack is a well-known process, leading generally to gradual transformation of calcite into gypsum due to the action of H2O (rain, condensate or moisture) and sulphur oxides (SO2 or
SO3, pollutants in the air, soil and rainwater resulting from combustion of fossil fuels), with
formation of H2SO4 and consequent sulphate production. Gypsum causes both aesthetic and
structural damages to building materials and Cultural Heritage, forming superficial and/or interstitial crusts whose detachment leads to swelling and uncovering of incoherent substrata. Reiteration of this process brings, in time, to complete destruction of the artefacts (Lal Gauri et al., 1989; Moussa et al., 2009).
Though it is commonly accepted that gypsum should be the only stable compound formed in the mortar/SO2 interaction, previous experiences show that other sulphates (soluble or unstable)
can also appear. In Portland cement, thaumasite and ettringite may form after reaction between C-S-H and sulphates in presence of carbonate ions in wet environments (Hooton & Thomas, 2002). For what concerns dolomitic lime mortars, their high Mg content of makes them particularly susceptible to dissolution by sulphate attack after acid rain (including dry
deposition) and pollution (Chever et al., 2010), bringing to formation of Mg-sulphates. Though this problem has been known for centuries, the studies published so far are contradictory: it is still unclear whether formation of Mg-sulphates should be considered common or atypical. Dolomitic limes tend to harden less readily than other mortars due to a delayed rate and scarce extent of carbonation upon curing (Maurenbrecher & Rousseau, 2000). This causes slaking of periclase (MgO) to brucite [Mg(OH)2] and carbonation to magnesite (MgCO3) or other Mg
hydroxycarbonates to be slower than the analogous cycle involving Ca compounds (quicklime portlandite calcite; Lanas & Alvarez, 2004a). The cured dolomitic mortars therefore contain for a longer time these Mg-based precursors, which are sensitively less stable and more reactive than calcite to environmental sulphur compounds (Klemm & Siedel, 2002), thus favouring selective mobilization of Mg. Such a release could also be favoured by dissolution of dolomite, with or without sulphate ions (Caner et al., 1985). Later (partial) evaporation leads to
crystallization of hydrated Mg-sulphates, whose quantities depend on the reactants amounts (function of diffusion-controlled solubility) and the time in which they become available (depending upon their dissolution rate; Morse & Arvidson, 2002; Berner, 1978). At room temperature and relative humidity (R.H.) > 85% epsomite is the stable form, which in drier conditions dehydrates to form hexahydrite (R.H. ≈ 60%), starkeyite (30%) and kieserite (Juling et al., 2004; Kramar et al., 2011; Hartshorn, 2012). Other parameters, such as presence of organic additives and solution viscosity, play a crucial role in the crystallization rate and pore size in which crystals grow, with consequent damages to host materials (Ruiz-Agudo et al., 2007a; 2007b, 2008).
The composition of the analysed mortars/plasters is quite usual and traceable to local sources. Sulphate attack, however, only involved Mg mobilization and crystallization of epsomite. Scarce gypsum was presumably added on purpose to favour the plaster grip rather than originating from sulphation, as hinted by its regular stratigraphic distribution. Presumably water containing dissolved SO2/SO3 permeated the church vault from seepages due to rain
percolations from the fissured roof and coverings, gradually dripping on pillars and walls. Capillarity may have favoured superficial spreading once H2O reached mortars/plasters due to
connected porosity. These sulphate-containing solutions infiltrated through the deeper layers, favouring complete solubilisation of the precursors (periclase and brucite) or products of Mg carbonation (magnesite and possibly dolomite, as well as Mg-hydrocarbonates) and promoting eruption of the epsomite efflorescences (Lopez-Arce et al., 2009). The solubility in H2O of
Mg-carbonates in the 20-25°C range is significantly higher than that of analogous Ca compounds (0.106 g/l vs. 0.014 g/l; Siedel, 2003). Selective Mg mobilization is also favoured by the related sulphates being deliquescent compounds, with higher water content and moisture sensitivity than Ca ones, which causes their crystals to undergo reiterated solubilisation/re-crystallization cycles reshaping their morphology (epsomite solubility at 20°C = 720 g/l). Absence of Mg-containing precursors in the analysed deeper layers – an occurrence in contrast to what usually
observed in ancient mortars (Montoya et al., 2003; Lanas & Alvarez, 2004a; Villaseñor and Price, 2008) – is consistent with the advanced degradation of the church vault and walls, which feasibly brought to their complete consumption. The susceptibility of these precursors to acidic solutions is known to be enhanced by presence of calcite, abundant in these mortars (though Mg2+ concentration is higher than Ca2+; Hartshorn, 2012). Detection of dolomite is probably to
be ascribed to aggregates of dolomitic limestones; however, presence of Ca2+ may also
contribute to re-crystallization of secondary dolomite as a carbonation product in the deeper layers. Such an evolution, although unlikely, would justify lack of gypsum. Weathering of dolomitic limes does not always imply sole crystallization of epsomite: contextual presence of Mg (epsomite and/or hexahydrite) and Ca (gypsum) sulphates was in fact often reported (Lopez-Arce et al., 2009; El Gohary, 2011; Kramar et al., 2011).
Mg-sulphates are commonly used in standard tests to assess the durability of stone under sulphate attack and considered extremely detrimental for the conservation of building materials and wall paintings. Epsomite, in particular, is very hygroscopic and due to its high solubility it repeatedly dissolves and reforms at the interface with atmosphere at R.H. > 83%, migrating and distributing throughout the material pores. Re-crystallization upon drying causes extensive weathering due to expansion and leaching, leading to crack development throughout the bulk stone and reducing the bond strength between masonry units (Seeley, 2000; Juling et al., 2004; Lopez-Arce et al., 2008; Zehnder & Schoch, 2009; Balboni et al., 2011). In the studied context, high R.H. values are not only favoured by daily and seasonal variations but also by rainwater permeating the building materials. These conditions allowed two distinct but connected crystallization events for epsomite to occur, each with typical morphology, consistently with the partition proposed by Ruiz-Agudo et al. (2007a): i) columnar interstitial crystals (usually termed ‘sub-efflorescences’; Fig. 4.a) grown between the deeper (mortar) and superficial layers (plaster and pictorial coating), exerting a pressure which leads to widening of cracks and fissures, superficial swelling and eventual detachment of fragments; ii) bunches of anhedral
small crystals (so-called ‘wax-like aggregates’) grown on the wall surfaces, with rounded faces symptomatic of continuous reshaping due to high deliquescence (Fig. 4.b). These differently shaped crystals possibly grew in consecutive steps, following a sequence which is reversed with respect to that proposed by Ruiz-Agudo et al. (2007a). The columnar crystals came first, in lower supersaturation conditions after successive imbibition-drying cycles, leading to swelling of the superficial layers and macroscopic crack-widening of the materials. These processes not only altered the porosity and pore size distribution of the materials, but coupled to reiterated freeze-thaw cycles (implying cyclic expansion/contraction of calcite) typical of northern Italy winters, brought to the detachment and fall of the superficial coatings (Steiger et al., 2008; Martínez-Martínez et al., 2013). Wax-like aggregates came later, showing non-equilibrium morphologies typical of crystals formed at a very high supersaturation and growing directly on the deeper substrata (mortars), in those areas of the walls (orange in Fig. 2.b) where the mechanical action of the first generation crystals had already caused loss of the superficial layers.
Formation of hydrated Mg-carbonates
Presence of nesquehonite and hydromagnesite in the whitish crusts grown on the damaged pinnacle wall paintings is intriguing. These phases, together with others, are known to represent average constituents of ancient dolomitic lime mortars (Montoya et al., 2003; Cardoso et al., 2014): formation of Mg-hydrocarbonates, in fact, represents an intermediate (or final) step in the carbonation of these materials. Hydromagnesite may form as a stable phase together with magnesite from curing in a humid, CO2-rich atmosphere of Mg-precursors such as periclase
and brucite (Bruni et al., 1998; Dheilly et al., 1999; Lanas et al., 2005). A similar process may involve crystallization of nesquehonite (Lanas & Alvarez, 2004b). However in this case study these minerals, rather than being confined in the deeper layers, form superficial crusts on the
pictorial coat, an occurrence at times reported in literature (Bruni et al., 1998; Kloprogge et al., 2003). This could be explained by two different mechanisms.
a) CO2-rich aqueous solutions, streaming and/or flowing from leakages in the
badly-maintained church roof, infiltrated the deeper mortars/plasters causing mobilization and solubilisation of Mg2+ ions from precursor phases, such as brucite or periclase. Capillarity
and interconnected porosity allowed these Mg-enriched solutions to reach the pinnacle surface, where they precipitated giving off CO2 at room temperature and causing
crystallization of neomorphic nesquehonite. Though these solutions contained also Ca2+
ions, the higher Mg2+ concentrations inhibited, at least partly, the nucleation and growth
rate of Ca-carbonates (Lippman, 1973). This is consistent with the reflections of hydrated Mg-carbonates being more intense than those of calcite in XRPD (Fig. 5.b). This
neomorphic nesquehonite, forming one or more intermediate phases (dypingite and
possibly other transitory Mg-hydrated carbonates), gradually converted to hydromagnesite – a process occurring in aqueous medium by dissolution-precipitation step(s), influenced by high temperatures (52-65ºC) (Langmuir, 1965; Davies & Bubela, 1973; Villaseñor & Price, 2008) and Ca-carbonate dissolution kinetics (Hopkinson et al., 2008).
b) Alternatively, though less likely, formation of hydrated Mg-carbonates could result from
an intervened consolidation attempt using unfit materials, such as ‘Water Glass’. This term indicates various types of soluble Na and/or K-silicates, used in the last century for surface consolidation and waterproof of lime mortars, plasters, wall paintings and stone
monuments (Schiessl, 1985). Later it was discovered that these materials, after few years, form a hard superficial layer sometimes leading to accelerated scaling. By reaction, ‘Water Glass’ gradually releases Na and K-carbonates which, in acidic atmosphere, combine with autochthonous salts in ancient walls forming neomorphic alkali nitrates and sulphates. The
K/Na-nitrates crystallize easily under high R.H. values (75-96%), as do Mg (and Na) sulphates, thus implying that use of alkali materials not only supplies more salts to the walls, but also converts less harmful salts into more dangerous ones. Besides, these transitory phases tend to gradually transform into nesquehonite, hydromagnesite and calcite – final products of the whole process, whose low solubility favours completion of the reactions (Arnold & Zehnder, 1991).
Though the first hypothesis may find an objection in the scarce solubility and dissolution rate of nesquehonite and hydromagnesite (Königsberger et al., 1999), it still seems the more plausible. This is basically due to: i) absence of historical or analytical evidence about the building having undergone restoration/consolidation with ‘Water Glass’; ii) the tendency for hydrated Mg-carbonates (especially hydromagnesite) to be susceptible to acidic sulphate solutions (Hartshorn, 2012), such as those which permeated the church roof and walls originating, in other areas, sulphate efflorescences (see section 4.1.1).
Whatever its origin, it is well known that nesquehonite can readily be altered into
hydromagnesite consequent to temperature raise or arid conditions (Ferrini et al., 2009). At room (or slightly higher) temperature, this process happens via formation of several metastable, less hydrated phases depending on R.H. and partial CO2 pressure, though it remains unclear
which step provides formation of which phase, as several mechanisms seem plausible (Davies & Bubela, 1973; Hopkinson et al., 2008). At higher temperatures (> 100°C), thermal
decomposition causes this transformation to be quite straightforward (Lanas & Alvarez, 2004b; Vágvölgyi et al., 2008).
While hydromagnesite should represent the final decomposition product, one of the
intermediate phases is dypingite, sometimes found in the studied crusts. This phase may form after calcination/CO2 leaching process of magnesite (Canterford et al., 1984) as well as
suggests that the decay proceeded at different rates in different areas, sometimes allowing formation of transitory phases or else leaving only the initial reagents and final byproducts (assuming that, in the end, only hydromagnesite should residue). As nesquehonite-to-hydromagnesite transformation was only superficial (and not in the deeper layers), no
detrimental effect due to shrinkage after water loss was observed – a phenomenon reputed to seriously damage building materials (Villaseñor & Price, 2008).
Formation of nitrates (biochemical deterioration)
Presence of nitrates in the superficial layers of wall paintings on the arch suggests a further kind of alteration occurred, involving also a biological aspect. Nitrates have become a major problem for the decay of ancient building materials, due to intervened changes in
environmental conditions. They are mostly found in close proximity to cultivated plots of land and decomposing organic matter (Siedel, 2000; Maguregui et al., 2008). In particular, Na and K-nitrates show equilibrium R.H. between 75 and 96%, crystallizing easily in humid
environments (Arnold & Zehnder, 1991).
Nitratine is iso-structural to calcite but has a significantly higher solubility in H2O. In some
cases it forms rather easy to eliminate efflorescences, whose damaging potential is weaker than sulphates (Feilden, 2003). Formation of nitratine may be caused by microbial activity
(Turcanu-Carutiu & Ion, 2014) as well as organic matter decomposition (El Gohary, 2011). Besides, nitratine may represent an intermediate step in the decay consequent to ‘Water Glass’ application, which finally brings to formation of hydrated Mg-carbonates (see section 4.1.2). Due to its deliquescence, it can absorb water from environmental moisture forming a Na-nitrate gel. Nitratine crystallization occurs at very high R.H. values (equilibrium: 75.4% at 20°C; Arnold & Zehnder, 1991) and its detection in the pictorial coat and preparatory levels confirms the extremely high water content permeating the whole stratified sequence.
In the studied case, huge availability of soluble and volatile elements such as Na (as NaCl) and N (as ammonia) can be explained by presence of relevant amounts of pigeon excrements (guano) on the roof surmounting the cupola and apse. The metabolic activity of nitrificant and nitroso bacteria in situ possibly oxidized ammonia to nitrous and nitric acid, while rainwater infiltrating through the badly maintained roof and coverings brought to Na+ mobilization. An
acid leachate was therefore formed, which gradually infiltrated the deeper wall layers until reaching those substrata under the pictorial coat, causing nitrates precipitation. In agreement with previous studies, in the analysed samples the concentration of nitrates increases from the deeper plaster to the painted surface (Török et al., 2011). The evaporation rate and location of salt crystallization also depends on the environmental R.H.: the more the crystallization proceeds beneath the surface, the more damage is produced (Arnold & Zehnder, 1991). In addition, pigeon droppings may also contain other salts (such as sylvite, Ca/K-sulphates, aphthitalite, apatite, and weddellite; Gomez-Heras et al. 2004) potentially dangerous due to acid attack.
Local formation of nitratine had multiple negative effects on the artworks: its crystallization caused cracking of the painted surfaces and pigment detachment due to mechanical pressure as well as stains on the pictorial coat, thus implying both structural and aesthetic damages.
Besides, organic compounds contained in guano provide useful nutrients for microorganisms, thus paving the way for a further microbial attack (Zehnder & Schoch, 2009; Török et al., 2011; Morillas et al., 2015).
Guidelines for a restoration intervention
Preliminary conservation measures aimed to overcome the progressing decay are concerned with the problem of keeping the building waterproof. All detected degrading byproducts clearly indicate that the main problem is represented by the bad building conditions – especially the
roof, subject to rainwater infiltrations from leakages which cause the walls to be saturated by H2O and very high R.H. values to persist inside the church. It is therefore fundamental, first of
all, to reconstruct and insulate the roof and coverages thus ensuring impermeability of the interiors, and drain H2O from walls to reduce R.H.. After that, a rigorous temperature and R.H.
monitoring would help in preventing further degradation (Price, 1993).
Once this major requirement is satisfied, an effective restoration of the decorative apparatus should be planned. Expansive salts such as sulphates and/or nitrates must be accurately
removed from walls and mural paintings, mechanically (where the artwork is compromised) or by applying superficial compresses. Desalination by reiterated application of superficial
poultices (i.e., wet cellulose saturated with distilled H2O) represents a viable procedure to
remove salts from porous materials preserving the underlying substrata, exploiting ion diffusion and capillary transport (Siedel, 2000). Epsomite efflorescences may also be cleaned by
spreading organic additives such as amino tri-methylene phosphonic acid (ATMP),
diethylenetriaminepentakis (methylphosphonic acid) (DTPMP) and poly(acrylic acid) sodium salt (PA), which act as growth inhibitors and habit modifiers (Ruiz-Agudo et al., 2007b; 2008). For what concerns the removal of hydrated Mg-carbonate crusts, use of acidic compounds is not recommended due to their distribution on pictorial coats realized “a calce”. Alternatively, micro-air-abrasive mechanical systems should be preferred.
After salt removal, the walls should be consolidated for both aesthetic reason and to ensure conservation of the structure. This could be done by using Silane based products, aimed to eliminate or reduce capillary sorption of water, efficient in increasing compressive strength, modulus of rupture, and abrasion resistance.
An in-depth scientific survey was performed to check the state of conservation and progressive decay of the ‘St. Maria della Stella’ church, in Saluzzo. The constituent materials and decorative apparatus underwent serious aesthetic and structural damages which, if neglected, might
inexorably compromise any chance of intervention to restore the building. The main cause of decay is represented by rainwater infiltrating through the ramshackle roof, windows and
coverages, which causes salt precipitation and increases R.H. – a condition favoured by scarce air recycle and presence of a groundwater aquifer. The church, in fact, is located at the foot of a slope which drains water from a nearby hill. Moreover, part of its floor is below the street level thus favouring humidity increase under the paving.
The scientific investigation proved that in a restricted area various decay mechanisms can coexist. In the studied case, these processes caused selective Mg2+ mobilization and precipitation of related
sulphates and carbonates, not involving analogous Ca2+ salts. Besides, formation of nitrates
consequent to bacterial metabolism accounted also for biological decay.
This study confirms how the accurate characterization of all degrading agents represents a fundamental step to plan a viable conservation of Cultural Heritage and restoration campaign.
Acknowledgements
Special thanks go to the “Fondazione Cassa di Risparmio di Saluzzo”, which partially financed this project.
References
Arnold, A. & Zehnder, K. (1991). Monitoring wall paintings affected by soluble salts. In: S. Cather ed., The conservation of wall paintings. Proc. Symposium organized by the Courtauld Institute of Art and the Getty Conservation Institute, July 13-16, 1987, London: pp. 103-135.
Balboni, E., Espinosa-Marzal, R., Doehne, E., & Scherer, G. (2011). Can drying and re-wetting of magnesium-sulfate salts lead to damage of stone? Environmental Earth Sciences, 63: 7-8, 1463-1473.
Berner, R.A. (1978). Equilibrium, kinetics, and the precipitation of magnesian calcite from seawater. American Journal of Science, 278: 1475-1477.
Bruni, S., Cariati, F., Fermo, P., Pozzi, A., & Toniolo, L. 1998. Characterization of ancient magnesian mortars coming from northern Italy. Thermochim. Acta, 321: 161-165.
Burn, R.S. 1871. Masonry, bricklaying and plastering. Donhead, Shaftesbury, 2001.
Caner, E.N., Demirci, S., & Turkmenoglu, A.G. 1985. Deterioration of dolomite by soluble salts in Divrigi Mosque-Turkey. In: Felix G, ed., Proc. Vth International Congress on deterioration and conservation of stone, Lausanne: pp. 299-305.
Canterford, J.H., Tsambourakis, G., & Lambert, B. 1984. Some observations on the properties of dypingite, Mg(CO3)4(OH)25H20, and related minerals. Mineralogical Magazine, 48: 437-442.
Cardoso, I, Macedo. M.F., Vermeulen, F, Corsi, C., Santos Silva, A., Rosado, L., Candeias, & A., Mirao, J. 2014. A multidisciplinary approach to the study of archaeological mortars from the town of Ammaia in the Roman province of Lusitania (Portugal). Archaeometry, 56(1): 1-24.
Comba R. 1999. Il monastero di Rifreddo e il monachesimo cistercense femminile nell'Italia occidentale (secoli XII-XIV). Atti Convegno di Studi Staffarda-Rifreddo, 18-19 May 1999, Cuneo. Società per gli studi storici, archeologici ed artistici della provincia di Cuneo, Comune di Rifreddo.
Chever, I., Pavìa, S., & Howard., R. 2010. Physical properties of magnesian-lime mortars. Materials and Structures, 43: 283-296.
Cowper, A.D. 1927. Lime and lime mortars. Special report no. 9, Department of Scientific and Industrial Research. Donhead, Shaftesbury.
Davies, P. J. & Bubela, B. 1973. The transformation of nesquehonite into hydromagnesite. Chem. Geol., 12: 289-300. Dheilly, R.M., Bouguerra, A., Beaudoin, B., Tudo, J., & Queneudec, M. 1999. Hydromagnesite development in magnesian-lime mortars. Materials Science and Engineering, A268: 127-131.
Diffrac Plus Evaluation Package 2005. Copyright © SOCABIM, 1996-2005.
Downs, R.T. 2006. The RRUFF Project: an integrated study of the chemistry, crystallography, Raman and infrared-spectroscopy of minerals. Program and Abstracts of the 19th General Meeting of the International Mineralogical Association, Kobe, Japan. O03-13.
El Gohary, M. 2011. Chemical deterioration of Egyptian limestone affected by saline water. International Journal of Conservation Science, 2(1): 17-28.
EN 459-1 2001. Building limes – part 1: definitions, specifications and conformity criteria. European Committee for Standardisation, CEN.
Feilden, B. 2003. Conservation of historic buildings, 3rd ed., Architectural Press, New York, p. 107.
Ferrini, V., De Vito, C., & Mignardi, S. 2009. Synthesis of nesquehonite by reaction of gaseous CO2 with Mg-chloride solution: its potential role in the sequestration of carbon-dioxide. Journal of Hazardous Materials, 168: 832-837. Gomez-Heras, M., Benavente, D., Alvarez De Buergo, M., & Fort, R. 2004. Soluble salt minerals from pigeon droppings as potential contributors to the decay of stone-based Cultural Heritage. European Journal of Mineralogy, 16(3): 505-509.
Hartshorn, H. 2012. Dolomitic lime-mortars: carbonation complications and susceptibility to acidic sulphates. Master Thesis, Graduate School of Architecture, Planning and Preservation, Columbia University, USA.
Hooton, R.D. & Thomas, M.D.A. 2002. The use of limestone in Portland cements: effect on thaumasite form of sulfate attack. Portland Cement Association. PCA R&D Serial No. 2658.
Hopkinson, L., Rutt, K., & Cressey, G. 2008. The transformation of nesquehonite to hydromagnesite in the system CaO-MgO-H2O-CO2: an experimental FT-Raman spectroscopic study. Journal of Geology, 116: 387-400.
Juling, H., Kirchner, D., Brüggerhof, S., Linnow, K., Steiger, M., El Jarad, A., & Gülker, G. 2004. Salt damage of porous materials: a combined theoretical and experimental approach. In: Kwiatkowski, D. & Löfvendahl, R.. eds., Proc. 10th International Congress Deterioration and Conservation of Stone, 27.6-2.7.2004, Stochholm, Sweden: pp. 187-194.
Klemm, W. & Siedel, H. 2002. Evaluation of the origin of sulphate compounds in building stone by sulphur-isotope radio. Geological Society, London, Special Publications, 205: 419-429.
Kloprogge, J.T., Martens, W.N, Nothdurft, L., Duong, L.V. & Webb, G.E. 2003. Low-temperature synthesis and characterization of nesquehonite. Journal of Materials Science Letters, 22(11): 825-829.
Königsberger E, Königsberger L, & Gamsjäger H 1999. Low-temperature thermodynamic model for the system Na2CO3-MgCO3-CaCO3-H2O. Geochim. Cosmochim. Acta, 63: 3105-3119.
Kramar, S., Mladenović, A., Pristacz, H., & Mirtič, B. 2011. Deterioration of the black Drenov Grič limestone on historical monuments (Ljubljana, Slovenia). Acta Carsologica, 40(3): 483-495.
Kumar, G.S., Ramakrishnan, A., & Hung, Y.T. 2007. Lime calcination. In: Wang, Lawrence K., Hung, Yung-Tse, eds., Handbook of Environmental Engineering, vol. 5, Advanced Physicochemical Treatment Technologies. Humana Press, Totowa, NJ: pp. 611-633.
Lal Gauri, K., Chowdhury, A.N., Kulshreshtha, N.P., & Punuru, A.R. 1989. The sulfation of marble and the treatment of gypsum crusts. Studies in Conservation, 34(4): 201-206.
Lanas, J. & Alvarez, J.I. 2004a. Dolomitic limes: evolution of the slaking process under different conditions. Thermochimica Acta, 423: 1-12.
Lanas, J. & Alvarez, J.I. 2004b. Dolomitic lime: thermal decomposition of nesquehonite. Thermochimica Acta, 421: 123-132.
Lanas, J., Perez Bernal, J.L., Bello, M.A., & Alvarez, J.I. 2004. Mechanical properties of naturally hydraulic lime-based mortars. Cem. Concr. Res., 34: 2191-2201.
Lanas, J., Perez Bernal, J.L., Bello, M.A., & Alvarez, J.I. 2005. Mechanical properties of masonry repair dolomitic lime-based plasters. Cem. Concr. Res., 36: 951-960.
Langmuir, D. 1965. Stability of carbonates in the system MgO-CO2-H2O. J. Geol., 73: 730-754. Lippmann, F. 1973. Sedimentary carbonate minerals. New York, Springer.
López-Arce, P., Doehne, E., Martin, W., & Pinchin, S. 2008. Magnesium-sulfate salts and historic building materials: experimental simulation of limestone flaking by relative humidity cycling and crystallization of salts. Construction and Building Materials, 58(289-290): 125-142.
López‐Arce, P., Garcia‐Guinea, J., Benavente, D., Tormo, L., & Doehne, E. 2009. Deterioration of dolostone by magnesium-sulphate salt: an example of incompatible building materials at Bonaval Monastery, Spain. Construction and Building Materials, 23: 846‐855.
Maguregui, M., Sarmiento, A., Martínez-Arkarazo, I., Angulo, M., Castro, K., Arana, G., Etxebarria, N., & Madariaga, J. 2008. Analytical diagnosis methodology to evaluate nitrate impact on historical building materials. Analytical and Bioanalytical Chemistry, 391: 1361-1370.
Martínez-Martínez, J., Benavente, D., Gomez-Heras, M., Marco-Castaño, L., & García-del-Cura, M.A. 2013. Non-linear decay of building stones during freeze-thaw weathering processes. Construction and Building Materials, 38: 443-454.
Maurenbrecher, A.H.P. & Rousseau, M.Z. 2001. Review of factors affecting the durability of repointing mortars for older masonry. 9th Canadian Masonry Symposium, Publ. Inst. For Research in Construction, Univ. New Brunswick, Canada: p. 12.
Monnier G. 1928. I Gesuiti a Saluzzo (1660-1773); Memorie storiche del Collegio della Compagnia di Gesù, Saluzzo, appendice sulla Nuova Residenza fino al 1900, Padova.
Montoya, C., Lanas, J., Arandigoyen, M., Navarro, I., García Casado, P.J., & Alvarez, J.I. 2003. Study of ancient dolomitic-mortars of the church of Santa María de Zamarce in Navarra (Spain): comparison with simulated standards. Thermochimica Acta, 398: 107‐122.
Morillas, H. Maguregui, M., Trebolazabala, J., & Madariaga, J.M. 2015. Nature and origin of white efflorescence on bricks, artificial stones, and joint mortars of modern houses evaluated by portable Raman-spectroscopy and laboratory analyses. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 136: 1195-1203.
Morse, J.W. & Arvidson, R.S. 2002. The dissolution kinetics of major sedimentary carbonate minerals. Earth-Science Reviews, 58: 51–84.
Moussa, A.M.A., Kantiranis, N., Voudouris, K.S., Stratis, J.A., Ali, M.F., & Christaras, V. 2009. The impact of soluble salts on the deterioration of Pharaonic and Coptic wall paintings at Al Qurna, Egypt: mineralogy and Chemistry. Archaeometry, 51(2): 292-308.
NorMal, 1980. CNR Centro di Studi di Milano e Roma sulle cause di deperimento e sui metodi di conservazione delle opere d’arte – ICR– Raccomandazioni Normal. Alterazione dei materiali lapidei e trattamenti conservativi - Proposte per l’unificazione dei metodi sperimentali di studio e di controllo.
Oates, J.A.H. 1998. Lime and limestone: chemistry and technology, production and uses. Wiley-VCH, Germany. Pavıa, S., Fitzgerald, & B., Howard, R. 2005. Evaluation of properties of magnesian-lime mortar. In: Brebbia C.A., Torpiano A., eds., Structural studies, repair and maintenance of heritage architecture, IX, Malta, June 2005, Vol 83, WIT Transactions on the built environment. WIT Press, Southampton: pp 375-384.
Power, I.M., Siobhan, A.W., Thom, J.M., Dipple, G.M., & Southam, G. 2007. Biologically induced mineralization of dypingite by cyanobacteria from an alkaline wetland near Atlin, British Columbia, Canada. Geochemical Transactions, 8(13), doi:10.1186/1467-4866-8-13.
Price, C.A. 1993. Preventive conservation of salt-contaminated masonry in the Wakefield Tower, HM Tower of London. Institute of Archaeology Bulletin, 30: 121-133.
Ruiz-Agudo, E., Martìn-Ramos, J.D., & Rodriguez-Navarro, C. 2007a. The role of saline solution properties on porous limestone salt-weathering by magnesium and sodium-sulphates. Environ. Geol., 52: 269-281.
Ruiz-Agudo, E., Mees, F., Jacobs, P., & Rodriguez-Navarro, C. 2007b. Mechanism and kinetics of dehydration of epsomite crystals formed in the presence of organic additives. J. Phys. Chem. B, 111: 41-52.
Ruiz-Agudo, E., Putnis, C.V., & Rodriguez-Navarro, C. 2008. Interaction between epsomite crystals and organic additives. Crystal Growth & Design, 8(8): 2665-2673.
Schiessl, U. 1985. Stereochromie. Zur Enrwicklung der Maltechnik mit Alkalisilikaten und der gleichzeitigen Enrwicklung der Irrwege der Putz- und Steinkonservierung damit seit 1825. In Historische Technologie und Konservierung von Wandmalerei. Bern: Schule für Gestaltung. 158-168.
Seeley, N. 2000. Magnesian and dolomitic-lime mortars in building conservation. Journal of Architectural Conservation, 2: 21‐29.
Siedel, H. 2000. Effects of salts on wall paintings and rendering in the Augustusburg Castle (Saxony). Proc. 6th International Congress Applied Mineralogy, Göttingen July 2000, Balkema/Rotterdam.
Siedel, H. 2003. Dolomitkalkmörtel und Salzbildung an historischer Bausubstanz. In: Leitner, H., Laue, S., Siedel, H., eds., Mauersalze und Architekturoberflächen. Hochschule für Bildende Künste, Dresden: pp. 57-64.
Steiger, M., Linnow, K., Juling, H., Gülker, G., El Jarad, A., Brüggerhoff, S., & Kirchner, D. 2008. Hydration of MgSO4•H2O and generation of stress in porous materials. Crystal Growth & Design, 8(1): 336-343.
Török, Á., Galambos, É., Józsa, Zs., Kriston, L., Bóna, I., Csányi, E., Szemerey-Kiss, B., & Méreyné-Bán, B. 2011. Salt efflorescence and subflorescence in Baroque frescos and the role of bat droppings in the decay wall paintings. In: Ioannou, I., Theodoridou, M., eds., Salt Weathering on Buildings and Stone Sculptures, Limassol, Cyprus: pp. 97-104.
Turcanu-Carutiu, D. & Ion, R.M. 2014. Pre-restoration investigations of the Basarabi chalk monument; diagnosis, treatment and implications. European Scientific Journal, 3: 124-134.
Vágvölgyi, V., Frost, R.L., Hales, M., Locke, A., Kristóf, J., & Horváth, E. 2008. Controlled rate thermal-analysis of hydromagnesite. Journal of Thermal Analysis and Calorimetry, 92(3): 893-897.
Villaseñor, I. & Price, C. 2008. Technology and decay of magnesian-lime plasters: the sculptures of the funerary crypt of Palenque, Mexico. Journal of Archaeological Science, 35: 1030-1039.
Wojdyr, M. 2010. Fityk: a general-purpose peak-fitting program. J. Appl. Crystallogr., 43: 1126-1128.
Zehnder K. & Schoch O. 2009. Efflorescence of mirabilite, epsomite and gypsum traced by automated monitoring on-site. Journal of Cultural Heritage, 10: 319-330.
Tables
Mineral phase Crystal-chemical formula Alteration
Floury efflorescence on
vault, pillars and walls Whitish incrustation onpinnacle mural paintings on archPaint detachment from Degradation byproducts epsomite MgSO4•7H2O X nesquehonite Mg(HCO3)(OH)•2(H2O) X hydromagnesite Mg5(CO3)4(OH)2•4(H2O) X dypingite Mg5(CO3)4(OH)2•5(H2O) X nitratine NaNO3 X
Superficial pictorial coat
sodalite Na8Al6Si6O24Cl2 (absent) X
calcite CaCO3 X X dolomite CaMg(CO3)2 X X quartz SiO2 X X albite NaAlSi3O8 X X anorthite CaAl2Si2O8 X muscovite KAl2(Si3Al)O10(OH,F)2 X talc Mg3Si4O10(OH)2 X lizardite Mg3Si2O5(OH)4 X gypsum CaSO4•2H2O X
Preparatory level and/or plaster finish
calcite CaCO3 X (absent) X
dolomite CaMg(CO3)2 X X quartz SiO2 X X gypsum CaSO4•2H2O X X thermonatrite Na2CO3•H2O X X Mortar calcite CaCO3 X X X dolomite CaMg(CO3)2 X quartz SiO2 X X X gypsum CaSO4•2H2O X
albite NaAlSi3O8 X X X anorthite CaAl2Si2O8 X orthoclase KAlSi3O8 X X muscovite KAl2(Si3Al)O10(OH,F)2 X X X paragonite NaAl2(Si3Al)O10(OH)2 X X X clinochlore (Mg,Fe)5Al(Si3Al)O10(OH)8 X X antigorite (Mg,Fe)3Si2O5(OH)4 X X hematite Fe2O3 X
Table 1. Mineralogical composition, inferred by XRPD, of the different stratigraphic layers related to all investigated kinds of alteration.
Figure captions
Figure 1. a) Façade of the ‘Santa Maria della Stella’ church (upper right square: Northern Italy
map with location of Saluzzo); b) view of the church vault: inner cupola, pinnacles and apse; deteriorated areas appear on the right; c) pillars with upper portions coated by a white-floury efflorescence.
Figure 2. Scheme of: a) the vault; b) northbound-oriented side wall. Different kinds of alteration
are rendered with different colours. Green: growth of interstitial columnar crystals causing swelling/detachment of the superficial layers. Orange: presence of a superficial floury efflorescence. Red: superficial whitish crusts on the pinnacle. Blue: pictorial coat detachment from wall paintings on the arch.
Figure 3. Stereo-microscopic images of the different alteration processes: a) interstitial growth of
columnar, translucent crystals between the superficial pictorial coat/preparatory level and deeper mortar; b) superficial floury efflorescence on a fragment of the side wall; c) whitish crust on a fragment of the damaged pinnacle; d) stratigraphic sequence on a fragment from the damaged arch, from top to bottom: superficial blue pictorial coat, greyish preparatory level, thick white plaster finish and coarse mortar with darker grains.
Figure 4. SEM BSE images of the different alteration processes: a) interstitial columnar epsomite
crystals growing perpendicular to the stratification, favouring detachment of the superficial layers; b) superficial wax-like aggregates of epsomite crystals; c)
Nesquehonite and hydromagnesite crusts on the damaged pinnacle; d) stratigraphy of a fragment from the damaged arch: nitratine crystals (dark grey with rhombic outline) appear in the superficial pigmented layer and preparatory level.
Figure 5. XRPD patterns of: a) superficial floury efflorescence on the side wall; all reflections
related to epsomite (Eps) with minor quartz (Qz) and calcite (Cal); b) whitish crust on the damaged pinnacle, made of nesquehonite (Nsq) and subordinate hydromagnesite (HMgs) with minor quartz (Qz), calcite (Cal), muscovite (Ms), plagioclase (Pl) and talc (Tlc).
Figure 6. Raman spectrum of the rhombohedral crystals in the pictorial coat on the damaged arch
