Evaluation of Calcarenite Degradation by X-ray Photoelectron Spectroscopy Analysis inside the Rupestrian Church of San Pietro Barisano (Matera, Southern Italy)

Abstract

We report on the XPS analysis of degraded surfaces inside San Pietro Barisano, the rupestrian church carved into the calcarenite rock of ancient Matera, which has been a UNESCO World Heritage Site since 1993. As reported in previous works, the “Sassi” district and the park of rupestrian churches were available as open laboratories for the National Smart Cities SCN_00520 research project dedicated to the sustainable recovery of this remarkable architectural heritage. In that context, XPS functionality was shown to reside in the possibility of analyzing surfaces by feasible sampling, acquiring spectra without any preliminary sample treatment, and processing data using a well-established curve fitting procedure. The obtained results allowed us to identify the degradation products of the investigated surfaces, thus contributing to defining a diagnostic framework for subsequent actions. Accordingly, the samples here considered, collected from the internal wall surfaces of the church, were all analyzed in comparison with the reference calcarenite, and the XPS results were evaluated as a function of local environmental factors and the historical context of the church itself. The final aim was to provide, for each sample, the most representative indicator(s) of biotic and/or abiotic degradation for reliable use, in a multidisciplinary context, in planning care interventions for building heritage.

1. Introduction

Monumental heritages are subject to the degradation processes of their constituent building stones, leading, over time, to progressive aesthetic and structural damage with the inevitable loss of their value. The various forms of stone degradation are due to physical, chemical, and biological phenomena linked to factors such as changes in temperature, solar light, and wind incidence, permeation of water containing organic and inorganic substances, biological attacks, salts crystallization, mechanical stress, and other effects [1,2,3,4]. Therefore, knowledge of the history and location of each monument is extremely important for the evaluation of its state, specifically related to synergistic actions of environmental factors, past and present, such as internal erosion due to biological colonies previously settled and already extinct in porous stones [5,6,7,8,9].

The state-of-the-art diagnostics and preservation of cultural heritage refers to the advancement of pioneering monitoring techniques and maintenance strategies based on the interrelation of green methodologies and new technologies respondent to the following overall requirements: monumental health and the accurate mapping of degradation; remote sensing and imaging at different scales and dimensions, digitization and data storage, and the dynamic use of technological databases; functional consolidation of monuments to prevent further deterioration and bioactivity; and enhancement from the perspective of safety and the circular cultural economy [10,11,12,13,14,15,16].

The experimental work presented here adds to previous contributions (vide infra) and diagnostic activity related to the research project “Smart Cities, Communities and Social Innovation, SCN_00520”, funded by the Italian Ministry of Research and University, titled “Product and process innovation for sustainable and planned maintenance, conservation, and restoration of cultural heritage”, whose objectives were in accordance with the mentioned BB CC requirements.

As part of the research project, the rupestrian churches carved into the limestone of the archeological park of Sassi di Matera (Matera, Basilicata Region, Southern Italy), represent important sites for investigating degradation phenomena. In particular, factors affecting the degradation in the Church of San Pietro Barisano, one of the most important rupestrian churches of the archeological park, were studied using multidisciplinary and multi-technique approaches [17,18,19].

For instance, the characterization of the biological patina covering the hypogeum walls of “San Pietro Barisano” by surface techniques X-ray Photoelectron Spectroscopy (XPS), Scanning Electron Microscopy (SEM), and biological assays, allowed for clarifying the role played by colonizing microorganisms, both prior to and following treatment with natural biocides (e.g., glycoalkaloids) and, more generally, the correlation between microbial colonization and surfaces degradation [20].

The aforementioned multi-investigative approaches were in fact appropriately applied there to characterize solid fragments sampled from hypogeum walls. Surface-specific XPS analysis was combined with SEM equipped with EDS to detect X-ray emission from underlying elements, thus achieving complementary information for correlation with microorganism identification and their in-depth colonization via biological analysis.

In the wide field of material science [21], XPS provides surface composition at the nanometer scale, and its contribution is thus crucial in determining the surface-dependent properties of materials in various environments inducing corrosion, adhesion, catalysis, biological interactions, etc. Considering its analytical depth, XPS is often proposed for the analysis of layered samples, such as archaeological finds, pigments in paintings, etc., in combination with microscopy and spectroscopy techniques of different categories and depth of analysis to achieve multi-dimensional information from the valuable merging of data that mutually confirm the overall outcome [22,23].

In this work, the XPS investigation is extended to a series of samples further collected from the internal wall surfaces of the San Pietro Barisano church, from the underground to the upper level, in order to identify chemical indicators of degradation depending on the internal location.

XPS analyses of ten powdered samples taken from the surface of internal walls appearing visually degraded are compared, taking raw calcarenite as the reference sample. The results demonstrate the ability of XPS to derive the composition of these homogenized samples and to associate the various classes of identified chemical compounds to sample locations, particular microclimatic conditions, and the state of the walls’ surfaces.

Three of the ten analyzed samples (#1, #3, and #5), collected along contiguous zones of the same wall, but differently degraded, were previously analyzed [24] using exactly the same experimental procedure reported below.

Therefore, for the completeness of information, the results reproduced from reference [24], published under the CC-BY-N-ND license, which allows the authors to retain copyright, have been included in this comparative monitoring of the San Barisano’s interiors.

The following sections highlight the correlation between the surface data and data from parallel experiments carried out in the same nave of the church, which support and confirm the indicators of calcarenite degradation derived by XPS.

2. Materials and Methods

2.1. Study Site

The “Sassi” district of Matera city, declared a UNESCO heritage site in 1993, is an example of a prehistoric rupestrian village located in Southern Italy (Basilicata Region), which represents an awesome rock-cut settlement, where architecture is intermingled with geological and geomorphological features of the area. The Church of San Pietro Barisano, originally called “San Pietro de Veteribus”, is the largest rupestrian church in the city of Matera (Figure 1). Subsequent archaeological investigations and renovations have brought to light the current appearance of the church with a new façade (dated 1755), a three-nave structure, and the hypogeum level used in the past for draining corpses.

As already stated, underground walls were first investigated for biodegradation, using combined techniques, to understand the correlation between chemical composition and microbial activity in the covering patina, grown in the climatic conditions of this closed environment [20].

In this work, we considered the upper level of the church consisting of three naves divided by carved pillars supporting round arches and six altars also crafted from tuff, as shown in the map in Figure 2. The high altar, dating back to the eighteenth century, is made of gilded wood. Adjacent to it, in the right aisle, a small aperture near the first altar devoted to St. Joseph allows access to a chamber with modifications dating back to the fifteenth century. This chamber was sealed off in the eighteenth century and repurposed as an ossuary.

The church presents, both outside and inside, evident phenomena of degradation in the calcarenite rock composing the structure, aggravated by biodeterioration mainly inside. For example, the internal walls selected for diagnostic sampling (Figure 1d and Figure 2) show chromatic alterations, with the presence of patinas, efflorescence, alveolizations, and detached fragments, as do other areas inside the church.

A wide portion of the right nave from the St. Joseph altar to the main altar (I–V in the map) was monitored in this work, individuating the sampling points to be analyzed, hereafter specified.

2.2. Sample Collection

The sampling collection performed in this work was based on the identification of the most degraded zones by visual inspection, as is marked in the map in Figure 2.

The samples were carefully collected from the walls using a steel spatula to ensure only the removal of surface patinas or other degradation products. Care was taken to preserve the deeper layers to avoid damaging the underlying structure.

The systematic collection of powdered samples, which were gently removed and homogenized in an agate mortar for XPS analysis, aimed to detect any eventual differences in their chemical composition depending on the internal location and specific environmental conditions [19,20,24]. In particular, samples #1, #2, #3, #4, and #5 (Figure 3 in this work and reference [24]) were taken from a degraded wall located on the right side of the main altar [24]. Two other samples, #6 and #7, were taken from a column next to the previous wall, which showed signs of frescoes that are now no longer visible (Figure 4, per left). On this column, the surface showed detachments and concretion phenomena because of water infiltrations and probable rising humidity.

Sample #8 was instead taken from the side altar of Saint Joseph, located near the entrance to the right-side nave (Figure 4, upper right).

Next to the altar of Saint Joseph, an access gate was opened in 1995 to an ossuary created in the final part of a completely frescoed sixteenth-century chapel. The room was characterized in the past by the presence of numerous frescoes, now in a state of decay, beneath which extensive biological colonization is visible. Sample #9, shown in Figure 4 in the lower left, was taken from this room, with the sampling area chosen beside the fresco of San Canio Vescovo where a surface bio-cleaning operation was carried out with a hydrogel based on sodium alginate (5% by weight) containing Ca(ClO)₂ (0.4% by weight) [25,26].

Finally, sample #10 was taken from the pillar column, and placed as the crow flies between altars III and IV, showing a mottled reddish surface (Figure 4, lower right).

2.3. XPS Analysis

The sampled powders collected for XPS analysis were properly stored in Eppendorf tubes. Before analysis, samples were first homogenized in an agate mortar and then pressed onto a double-sided copper tape, properly fixed on a steel sample holder, and directly introduced in the analysis chamber of the XPS spectrometer. The spectrometer Phoibos 100- MCD5 (SPECS) was operated at 10 kV and 10 mA, in medium area (diameter = 2 mm) mode, using MgKα (1253.6 eV) and AlKα (1486.6 eV) radiations, with pressure better than 10⁻⁹ mbar, during spectra acquisition.

Wide spectra were acquired in FAT (Fixed Analyzer Transmission) or FRR (Fixed Retarding Ratio) modes with channel widths of 1.0 eV. The use of a double anode (Al/Mg) proves to be appropriate for better identification of XPS signals, varying in kinetic energy (KE) with the source employed, and of X-ray-induced Auger signals, dependent only on atomic relaxation following photoemission and, therefore, present in spectra at the same KE (eV) with both sources [27,28]. High-resolution spectra were acquired only in FAT mode, with a constant pass energy of 9 eV and channel widths of 0.1 eV, using the X-ray source deemed most suitable from the preliminary wide comparison. The detailed regions were “curve-fitted” using the Googly program, which allows for the evaluation of intrinsic (detected peaks) and extrinsic (background) features of XPS spectra [29,30]. As detailed in the mentioned references, Googly associates each peak with its own background in the form of polynomial tails. In this way, their relative intensity can be adjusted to better determine the in-depth position of elements in layered samples and/or the simultaneous presence of different elements in the same energy region as those considered for the homogenized powders in this work. The peak areas and positions (Binding Energies, BE) resulting from the best curve-fitting of each detailed region were, respectively, normalized using proper sensitivity factors and referenced to the C1s aliphatic carbon, as an internal standard set at 285.0 eV [27,28].

Sample characterization was accomplished by relating the curve-fitting results to the standard compounds analyzed, by accessing the online XPS database [31] and the relevant literature data. In addition to the corrected BE values, the Auger parameter (α’), which combines photoelectron and related Auger peaks, was considered, where possible, to confirm the chemical assignments, given its independence from the reference energy level and surface charge of the non-conducting samples [27]—Appendix 4. For semi-quantitative analyses, the percent atomic composition (At%), summarized in pie charts for each sample analyzed, was calculated by checking the normalized peak areas of the curve-fitted regions and verifying their matching, in compliance with the accuracy limits of the XPS technique (+/−10%, see Section 3).

3. Results

In reference [24], pages 91–104, the XPS analysis, as detailed above, is reported for samples #1, #3, and #5, whose composition was obtained by curve-fitting the main spectral regions of each sample detected in their wide spectra. The results of the three samples (shown in more detail for sample #1), summarized in tables (corrected BEs, normalized areas, and chemical group assignments) and therein displayed on pie charts, are included here as part of the sample collection indicated in the map in Figure 2. The investigation of the entire collection aims to determine any differences in the composition of the samples depending on their position inside the right nave of the church.

It is well reported [32] that calcarenitic rock is mostly composed of calcite (in some cases, it is associated with aragonite). Other components (magnesium oxide, quartz, aluminum oxide, and other inorganic ions) may be present in different, minor percentages. Thus, to start with an almost raw rock that could be considered “not or only incipiently degraded”, we first analyzed the surface powder collected from a fragment of a calcarenite block extracted from a (not specified) local quarry, in the construction site of the archaeological park of “Sassi di Matera”. In Figure 5, the wide spectra, registered in FAT mode with both sources, are superimposed and show Auger peaks at identical kinetic energies, while the photoelectron peaks are shifted at 233 eV, as the difference in energy of the two sources. Since the same elements are detected in both wide spectra, only those marked in black, acquired with AlKα, are reported. Of these, the main detailed regions of each element, acquired at higher resolution and curve-fitted, are grouped in Figure 6.

The curve-fitting results helped to identify the chemical states of resolved peak components and their relative intensities and thus to derive the surface composition of calcarenite, as specified hereafter.

The same set of acquisitions was then repeated with MgKα to verify the concordance in the composition of the calcarenite powder with both sources. The curve-fitted regions with MgKα are not reported in the figure, but the curve-fitting results are aligned in green for comparison in

Table 1. Curve-fitting results of detailed regions acquired with AlKα (black) and MgKα (green).

Region	Normalized Area	BE-Corrected (eV)	Assignments
C1s
1	276/392	282.9/282.3	C-C
2	2868/4096	285.0	C-C
3	476.5/598.3	287.5	C=O, O-C-O
4	2317.3/2540.4	289.7/289.8	(CO3)2−
K2p	94.1/116.4	293.8 (2p3/2)	mineral oxide
Mg KLL	Only considered corrected KE max	1179.9	α’ = KEKLL + BE1s = 2484.7 indicative of MgO in clay minerals
Mg1s	167.6	1304.8
Mg2p	224.2	49.9
Ca 2p
1,3	107.1/138.8	345.1/345.2 (2p3/2)	Ca/CaO
2,4–6	2184.5/2506.85	347.2/347.4 (2p3/2)	CaCO3 + SU1,2
Si 2p
1	105.3/67.7	99.85/100.5	SiC/SiOx
2	868.7/1125.23	103.0	SiO2/silicates
Al 2s	635.2/533.5	119.6/119.5	Al2O3
O1s
unresolved components	total area 10,733.3/12,407.6	531.9/532.0	Metal oxides CaCO3, O-R

as well as the correspondent At% composition, displayed in the graphs of Figure 7.

In the pie charts, as partly performed in reference [24], components not belonging to the main structure of calcarenite, calcium carbonate, or “derived” calcium compounds, are grouped as “oxides/silicates” and/or “soluble salts” based on the current results.

The pie charts in Figure 7 indicate the same calcarenite components with both X-ray sources with only some variations in their relative intensity (At%) within the limits of XPS accuracy (+/−10%) [27,28], mostly because of an increase of the C-C component with MgKα, likely induced by the possible degradation of organic compounds under prolonged irradiation, as already investigated in [33].

Therefore, taking all the considered aspects into account, these two referent compilations coherently show the calcarenite’s surface to be composed of carbonate and phyllosilicate structures dominated by carbon-containing contaminants. Thus, both sources can be used to interpret the degradation paths of the ten samples considered here. However, to facilitate data processing, spectra acquisition with AlKα was favored to quantify magnesium, which is present in all samples with different intensities, using the curve-fitted Mg1s region [27,28]. The acceptance of results is based on, similar to previous work, the partial mass balances (i.e., Ca:CO₃ equal to unity for the primary calcarenite component), on the overall neutrality balance, and the goodness of all curve-fitted regions. In this regard, the total area of the O1s peak, in this case, fitted with only one component, comprising all unresolved chemical states, should correspond to the total oxygenated contributions derived from other regions, assigned to oxygenated compounds, taking into account the stoichiometric coefficients, within the reported limits of XPS accuracy.

The results, summarized and displayed in pie charts for all the samples numbered in Figure 2, were obtained using the same procedure reported here for the reference calcarenite and in reference [24].

For sample #1, taken at the rightmost side of the surface wall in Figure 3 (evidently degraded in a non-uniform way along the direction of sampling), the At% distribution in Figure 8, left, shows a higher amount of carbonaceous components of different nature including 59% of C-C types (carbides, aliphatics, aromatic, etc.) and 25% with other functionalities (C-S, C-O-C, C-O, C-N, C=O, etc.). The presence of calcium sulfate (7%) and calcium oxalate (4%) in the dark green patina of sample #1 can be associated with calcium carbonate transformation due to sulphation and biological metabolites (organic acids), respectively, as well documented for phenomena related to monumental degradation [34,35]. The Auger parameter of magnesium (α’ = 2484.5 ± 0.5 eV), in the range of values including hydrated MgCl₂, MgO, and silicate minerals of the Na/Mg/Si₄O₁₀(OH)_n-type, suggests that chlorides, silicate components (SiO_x), and other alkaline earth metals are linked together, according to the XPS database [31]. These results are compatible with the sample location in Figure 3. The sampled dark green degraded layers did not contain carbonates, and a total percentage of 5% was reached by silicates/inorganic salts.

Sample #2 was obtained from an area close to sample #1. Figure 8, right, shows, in the At% chart, the organic contamination of the same entity of sample #1, while calcium oxalate prevails, reaching 9%. In this case, moreover, species in the silicate group (such as SiOx, CaO, and NaCl) are not identified, which were replaced by magnesium chloride (7%), presumably linked to biological activity, the transformation of inner carbonate/dolomite structures into oxalate and calcium sulfate (5%), and/or other degrading interactions This result is similarly expressed by sample #1 [34,35], but more significantly.

Sample #3 (Figure 9 left) was taken in an area visually attenuated in color, i.e., the dark green layers are thinner and certainly collected together with uncolored sublayers less affected by degradation. This assessment is corroborated by the lower percentage of sulfate (1%) and the concurrently higher percentage of calcium carbonate (23%), the primary component of the calcarenite material. Likewise, the proportion of silicates rose to 13%, encompassing a notable amount of potassium in addition to calcium as intercalated fluxing ions. The persistent high percentage of carbonaceous components likely indicates the presence of biofilm residues, which are possibly no longer biologically active given the discoloration, embedded within the green patina that has spread from the area of samples #1 and #2.

Sample #4 (Figure 9 right) was taken from a lumpy whiter area where the presence of biological patina was no longer visible. From the At% results, however, we note the low percentage of calcium carbonate (6.3%) and the conspicuous presence, beside sulfates (nearly 8%), of K⁺/Na⁺ chloride, nitrites, and nitrates up to 20%, i.e., soluble salts likely deriving from efflorescence phenomena. The whitish color of the surface at this sampling point may in fact indicate salts and dehydrated oxides deposited from water permeated from the ground or the ceiling and subsequently evaporated, creating a vapor pressure difference between the hydrated compound and atmospheric water vapor. Thus, in this case, the strong hydration and changes in relative humidity (RH) of the surroundings contribute physically to the wall degradation, not favoring bio-colonization but salt deposits given the percentage of organic carbon (46%), which is lower than the previous samples, and the visual absence of the dark green patina.

Regarding sample #5, the last of the series in Figure 3, located next to the high altar, the analytical findings reported in Figure 10, left, show similarity either with sample #3 (for the same lower quantity of sulfates (1%) and an even increase in the percentage of calcium carbonate (26%) with the appearance, again, (about 4%) of the oxide/silicate group) and with the closest sample #4, given the absence of the dark green patinas on both surface zones and presence of soluble salts once more linked to efflorescence phenomena triggered by the permeation and/or evaporation of water into the masonry on this specific site. As for the nearby samples, the high percentage of total organic carbon (61%) may again suggest the presence of biofilm residues and other contaminants, also considering the exposition of the wall in Figure 3 to the passage of visitors toward the main altar of the church.

Samples #6 and #7 were taken from a column next to the previous wall where signs of the presence of deteriorated frescoes were still evident (Figure 4 upper left). This location shows wrinkled walls with here and there surface detachments and concretion phenomena, also because of probable water infiltration/rising humidity.

Figure 10, right, shows the distribution of the different groups identified in sample #6 expressed as At%. The percentage of organic carbon (nearly 64%) is probably also linked to the presence of paint residues on the surface. The second most abundant species (20%) is calcium carbonate, bringing only residual sulfates (1%), followed by abundant soluble salts and magnesium oxide (7%), probably also migrating to the surface with rising water.

Sample #7 was taken in the zone below that of sample #6, as indicated in Figure 4 in the upper left, affected by concretions. In this case, in the distribution graph in Figure 11, left, the percentage of carbonate, magnesium, and soluble salts is even more noticeable. Unlike the efflorescence phenomenon found in sample #4, composed of extensive whitish areas because of the high percentage of soluble salts (including nitrates and nitrites) transported by water and precipitated in the form of crystalline powders, the concretion identified in sample #7 is of limited extension in the form of a compact stalactite.

Sample #8 was taken from the side altar of St. Joseph (Figure 4, upper right) located near the entrance to the right-side nave (point I in the plan, Figure 2).

From a first visual analysis, the presence of black crusts is clearly visible, especially covering the base and the right corner of the altar where the sampling was performed. As expected for this sample, XPS analysis did not identify any calcarenite components, as shown in the graph in Figure 11, right. Given the high levels of organic carbons (79%) with oxygen and nitrogen functionalities, the presence of calcium oxalate and thiosulfate can certainly be linked to the presence of a biological crust, probably produced by a rather large number of organisms including bacteria, fungi, lichens, and higher plants, which may have also incorporated particulate matter from the church floor. The presence of nitrogenous salts, such as nitrites and nitrates could, also be an indication of bacterial activity. However, the black patina resembling a tarry substance was very easily removed and did not respond to the cleaning treatment reported hereafter for sample #9.

Sample #9 was taken from the room adjacent to the altar of St. Joseph, Figure 2. In the past, the room was characterized by the presence of numerous frescoes that are now in a significant state of decay; in the area below them, there is an extensive dark green biological patina. In this room, beside the fresco of San Canio Vescovo, visible in the upper corner of Figure 4, lower left, a cleaning treatment with a bio-gel, made of sodium alginate with incorporated calcium hypochlorite, was tested on a selected stained zone shown in the lower part of the picture. Sampling in this “treated” area also aimed to verify the conditions for restoring the limestone surface after the cleaning operation in comparison with other treatments of the green patina in rupestrian churches, based on gel support, considered within the SCN_00520 project [17,18,20,24].

As per visual impression, the surface was reported to be completely cleaned of the patina, still present around the sampled, treated area [17,18]. Based on the results of the XPS analysis, shown in Figure 12, left, we find, in fact, the presence of the carbonate component (CaCO₃ of the underneath calcarenite) at a similar percentage to the reference sample in Figure 7. The discoloration of the green patina demonstrates the effectiveness of the bio-cleaning; however, the high quantity of functionalized carbon, almost half of the total organic component (61%), is probably due to biofilm residues that were not completely removed and possibly trapped during the treatment, given the porosity of the wall.

Finally, sample #10 was chosen mainly because of the reddish appearance of the column seen in Figure 4, lower right, from which it was taken, which was visibly not too degraded.

In reality, the red stains flashing on the surface were not specifically identified, probably because their sampling was a negligible part of the sampled totality. The better state of the column seems to be confirmed by the At% composition shown in Figure 12, right, which shows all the components of the calcarenite block in Figure 7 with analog intensity and only a few percent of soluble salts, although present and considered as warning indicators, as well as the high amount of functionalized carbon in sample #9, covering the silicate group of the underneath calcarenite.

4. Discussion

Regarding the obtained results, it is worth noting that the fraction of organic carbon (understood as the sum of all the organic carbon (C-C) and functionalized C-O/CN groups) is predominant in nearly all the considered samples consisting of surface layers with adsorbed organic contaminants. The ubiquitous presence of aliphatic carbons allowed us to set the correct energy scale by referring to aliphatic C1s at 285.0 eV. The relative distribution of the functionalized carbons is seen to vary more on the relevant pie charts. The maximum percentage of oxygen- and nitrogen-containing carbon is in sample #8 (altar of St. Joseph), where black oily crusts are observed, and in samples #1 and #2 (portion of the wall with extensive dark green patina). In these samples, the carbonate structure of calcarenite is absent, being transformed into calcium sulfate and calcium oxalate. The presence of oxalate (C1s component at around 289 eV) is probably derived from the microbial metabolism present in the patinas, as found in previous investigations carried out in the church’s hypogeum [20].

In the other samples, the carbonate value (C1s component at around 290 eV) varies in percentage depending on the concomitant presence of sulfate, oxides, and efflorescence products. The minimum value of carbonate is in sample #4, where sulphation and soluble salts prevail. The values of carbonate are maximized in samples #3, #9, and #10, which are also visually cleaner, with no or a minimum percentage of sulphation and soluble salts. The cleanliness of samples #3 and #10 allows for the appearance of silicates, which are probably obscured in sample #9 by the considerable amount of functionalized carbon, as in sample #8.

Sample #7 shows 27% of carbonate with the minimum values of organic carbon balanced; however, the greatest presence of magnesium oxide/silicates (18%) with sodium salts (10%) and sulfate (15%) in the confined efflorescence agglomerate is visually compact. As demonstrated for composite building materials [36], possible treatments of the calcarenite walls with plastering materials of different textures and porosities would have effects on the evaporation rate of the saline solutions in the composition and form of efflorescence deposits.

Sulfate with S2p3/2 at an average value of 169.3 eV (only at 168.7 eV as thiosulphate in sample #8) was also highlighted in all samples with values ranging from 1% to 15%, except in sample #9. This is probably because it washed away, if present like gypsum, during the cleaning operation carried out on the wall.

Other components constantly present in the percent distribution of the ten samples considered are the oxides included in the silicates group in the pie charts. Of these, dehydrated CaO and, particularly, MgO are often visible even in the absence of the silicate matrix, sometimes being transformed in chlorides and then included in the group of soluble salts.

To better understand the obtained results, other experiments carried out in the SCN_00152 project on related topics are considered together with data from the literature [17,18,20,21,22,23,24].

X-ray diffractometry (XRD) investigation on four detached fragments from the internal walls of the church, on the right side wall of the apse affected by efflorescence phenomena [18], showed that calcite (CaCO₃) and dolomite (MgCa(CO₃)₂) are the prevalent minerals in the stones, with a lesser presence of gypsum (CaSO₄), quartz (SiO₂), nitratin (NaNO₃), and akermanite (Ca₂MgSi₂O₇).

The XRD results, shown in Figure 3 of reference [18], thus indicate that components adjunctive to those characteristic of calcarenite rocks [32] are present in the deeper layers of church San Barisano’s masonry walls. In particular, the calcium compounds mostly present in those fragments are likely derived from calcarenite dissolution, leading, in various paths and successions, to those identified by XPS in the surface layers.

Regarding calcium sulfate, it is known that it may exist in the following three hydration levels: anhydrous CaSO₄ (anhydrite), hemihydrate CaSO₄·½H₂O (bassanite), and dihydrateCaSO₄·2H₂O (gypsum). Thus, given the presence of gypsum in the fragments investigated by XRD, referred to as CaSO₄ regardless of its hydration state, it can be deduced that the damage induced in the walls of the San Barisano church is properly linked to its crystallization cycles (hydrated/dehydrated states) within the porous calcarenitic rock, the latter playing a fundamental role in determining the extent, and type of degradation. The presence of other salts, in particular, deliquescent salts, allowing greater solubility of CaSO₄ and their joint migration towards the wall surface, accentuate the structural damage.

The efflorescence phenomena, a diffuse weathering form producing deposits of soluble salts and dehydrated oxides along the pores and extruded on surfaces, as identified and detected by XPS in this work, are reported for similar monumental masonry made of calcarenitic stones [37].

The main cause of deterioration and bio-deterioration is thus the migration of water inside the wall structure combined with excursion in internal humidity and airborne contaminants, which is also introduced by visitors and different human activities inside/outside the church extended over time. The migration of water brings dissolved salts as precipitates to the surface after water evaporation, and the damage produced by these efflorescence phenomena should be investigated together with probable causes connected to the defective sewage systems, underground tunnels, and locks, possibly produced by past interventions on the flooring, etc. [38]. Furthermore, both temperature and humidity play a decisive role in the development of mold, bacteria, and biodeteriogens, as found in various biological investigations [39,40].

As shown in previous work [17,18,20,24], because of local and microclimatic conditions that favor bioactivity, calcium oxalate is detected by XPS in samples made of dark-green patina, which is metabolically produced by a number of organisms including bacteria, fungi, lichens and higher plants [41].

As part of the SCN project, ecological approaches were tested for the control of biodegradation using natural biocide extracts of spontaneous plants and secondary metabolites less toxic to human health [42].

Unfortunately, the recolonization of biological agents may occur, and attention should be paid after bio-cleaning operations to consolidate treated surfaces with protective products suitable to contrast/slow down adverse factors. For example, fast re-hydration, an increase in carbon dioxide, and atmospheric particulate matter in an “indoor” environment are also significantly influenced by the presence of visitors.

Another important factor to consider for church conservation is the porosity of calcarenite, which certainly makes the operation of eliminating biodeteriogens and completely cleaning the surfaces more complex.

Comparing the At% distributions of the ten samples (Figure 8, Figure 9, Figure 10, Figure 11 and Figure 12), with that of the calcarenite block considered as reference sample #0 (Figure 7), the main chemical constituents of each sample can be highlighted (Figure 13). This allows us to associate significant indicators to the type of damage, characterized by the given location.

The following deductions can be drawn by examining the histogram in Figure 13, which is useful for identifying damage indicators:

• The absence of carbonates and the presence of oxalate/sulfate with a higher amount of functionalized carbon are indicative of biodegradation (as for samples #1, #2, and #8) favored by humidity and climate conditions, especially in areas with poor air exchange;
• The presence of soluble salts, prominent in sample #4, is indicative of efflorescence phenomena because of water migration inside the walls;
• The presence of sulfate and oxides, particularly magnesium oxide, can be linked to the presence of dolomite and other minerals in detached fragments, indicative of subsequent degradation processes of calcarenite;
• The presence of mixed components is indicative of progressive degradation paths that have occurred over the centuries because of the consecutive/concomitant effects of contaminants of both abiotic and biotic origin and interventions and activities carried out on site over time [6,43,44].

From the above considerations, it is clear that XPS has the capability to identify the chemical descriptors associated with the various types of sample degradation.

The XPS indicators will serve to integrate the database of the technological platform developed within the scope of the project SCN_00152 [24] pp. 31–44, which will be available online as a useful support to users operating in the field of cultural heritage.

5. Conclusions

Nowadays, the conservation of cultural heritage requires the development of increasingly innovative, effective, long-lasting, and less expensive protection strategies, with the use of Information and Communication Technologies (ICTs).

From this point of view, the interdisciplinarity of the SCN_00520 project pursued these goals in advance. In that context, XPS, as a surface-specific technique, provided considerable cognitive support for the characterization of surface samples collected from internal walls of rupestrian churches. The identification of organic and inorganic components and their association with the multiple causes of degradation, i.e., intrinsic porosity of calcarenite, water permeation, and other environmental factors favoring the bio-receptivity and colonization of adaptive organisms, contributed to the diagnostic investigation.

Summarizing the results obtained in this work inside San Barisano church, the compositions of degraded surfaces, as identified by XPS, confirmed the importance of climatic parameters such as temperature, humidity, lighting, external contaminants, and air quality inside the church, especially in more closed areas. The necessity of a broader control of their influence on porous calcarenite was foreseen within the diagnostic scope of the SCN project by implementing the sensor network placed inside and outside the church, which aimed to compare ex situ analysis with the output of in situ sensors recorded by the technological platform. The main objective was to define the threshold level of biotic and abiotic degradation below which risky situations are reduced and the masonry walls better preserved.

Another important confirmation emerged from the use of combined techniques with different analytical depths. The correspondence of various calcium compounds, as identified by XPS, on the wall surfaces with those provided by XRD, along with efflorescent salts and mixed oxides driven by water permeation, as well as the presence of oxalate and organic metabolites related to the microorganisms detected in previous work by SEM/EDS and biological essays, all proved to be very important in identifying the current state of the entire masonry and forecasting curative interventions for the UNESCO site preservation.

For these reasons, the XPS results of the present investigation will be combined with the results from other churches and calcarenitic artifacts located in different areas of the archaeological park (work in progress) to obtain a larger set of data for statistical processing. Principal component analysis (PCA) will then be applied to discriminate the variables that define the degradation descriptors better in the specific situations of the case studies considered.

In the future, the implementation of the database, accessible online, will be useful in the field of cultural heritage as a technical guide for the continuation of this study and research project and similar projects dedicated to the recovery of monumental sites.