On the Mechanisms of Marble Deterioration of Antonio Canova Cenotaph in Santa Maria Gloriosa dei Frari Basilica in Venice

Abstract

The funerary cenotaph dedicated to Antonio Canova in the Frari basilica was erected in 1827. Since the beginning, some alteration features were recorded. In the last decades, some areas showed a sharp increase in deterioration processes due to large pieces of marble details missing from the statue surfaces. Macroscopic observation of the marble surface showed different forms of alteration as well as the massive presence of salt efflorescence. The main aim of this paper is to assess if there is a relationship between the decay observed and the presence of salt efflorescence, to subsequently ascertain the source of salts, and consequently to propose how to intervene to stop any further cause of alteration. In order to assess the relationship between the different types of alteration macroscopically observed, some samples were taken from the specific areas showing significant alteration features. Optical (OM) and scanning electron microscopic (SEM) observations associated with energy dispersive analysis (EDS) allowed us to explain the stages, each one corresponding to different features, through which the exfoliation and lamination of surface scales have been taking place. Moisture content in the brick structure was in the range of 17–26% until 140 cm of height. Above this height, moisture content is decreasing, and the maximum height of the capillary rise front is between 200 and 250 cm. In these areas, ions coming from the foundations of the monument deposit salt crystals within marble at a sub-surface level, causing the detachment of marble surface layers. In order to stop the rapidly increasing rate of decay observed over the last decades, it has been recommended to remove the statues from the basement and to insert a damp-proof course to prevent any further capillary rising damp. For the removal of embedded salts in the statues, the immersion of the removed statues inside deionized water baths has been recommended.

1. Introduction

The cenotaph dedicated to Antonio Canova was built in 1827 inside the Frari Basilica on the left nave, against the south side wall (Figure 1) [1]. After Canova’s death, a public subscription was launched to collect funds, which led to the erection of a funerary monument to celebrate the artist and to provide a resting place for Canova’s heart. The work’s execution was assigned to Canova assistants who realized the funerary project prepared by Canova to celebrate the painter Titian. The funerary monument was realized according to Canova’s style, which takes into account the use of a pyramid shape in the theatrical arrangement of the melancholy figures in a sad procession.

The cenotaph had manifested some alteration features three decades after its construction. In particular, the alterations reported at that time concerned the formation of stains on the surfaces of the marble elements and some fractures in the stone ashlar coverings at the basement of the monument [1]. The first photographic documentation dates back to the beginning of XX century, during some important conservation works carried out to re-instate the stability of the monument. The survey of the conservation state of the monument had highlighted static problems and the presence of some yellowish and grayish areas. At that time, the marble surfaces were not affected by any material loss. In the last quarter of XX century, some deterioration features have been observed. In particular, some areas were affected by salt efflorescence and the detachment of small surface pieces of marble. The people responsible for the conservation of the monument decided to make a restoration intervention using acrylic resin without a preliminary identification of the causes of deterioration [2].

Almost two decades after the aforementioned restoration, the situation had further worsened (Figure 2a–d) and the Superintendence for Architectural and Landscape Heritage of Venice had planned to intervene with a conservation project that would be preceded by a diagnostic investigation aimed at understanding the causes that had significantly accelerated the further loss of marble surface portions.

It is important to point out that based on the consolidated experience gained from studies carried out since the 1970s in numerous Venetian churches and historic buildings, significant losses of material located at a certain height from the ground were observed [2,3]. These alterations were generally due to the salt crystallization, whose saline species, in relation to their mobility, migrate vertically to different heights before crystallizing [3,4,5,6,7]. Measurements carried out in many Venetian buildings confirm the different salt distribution according to elevation; salts deriving from chlorides and nitrates were found at greater heights than those deriving from sulphates due to their different mobility [8,9,10]. Crystallization can occur on the surface, forming white efflorescence, or inside the substrate, nearby the surface, forming non-visible salt crypto-efflorescence or sub-efflorescence [11]. These last are more dangerous because growing inside the capillaries, they exert a pressure on the capillary walls, breaking them.

2. Previous Studies Carried out in 2009–2010

Macroscopic observation of the frequent formation of droplets [12,13] on some areas of the group of sculptures led the authorities responsible for their conservation to plan a survey to investigate the microclimatic conditions that could be responsible for the phenomena observed (Figure 3).

Measurements of surface temperature on different areas of the statues’ surfaces have revealed the slow change of surface temperature with respect to the surroundings’ air temperature fluctuations [14]. The thinner parts of statues, such as the arms, legs, and hands, show a faster change in surface temperature in relation to sudden neighbouring air temperature fluctuations. One of the factors responsible for this behaviour is the ventilation, which is lowering the surface temperature due to the heat exchange with the air of the church [15].

The thermal behaviour of the cenotaph was also investigated using a thermographic survey [15]. Measurements of the pyramid basement and of the lower steps showed a higher temperature than the other areas, and in particular, the temperature decreased progressively with elevation from the church floor (Figure 4).

This temperature decrease seems to be in relation to the reduction in the thickness of the monument structure. The higher temperature measured in correspondence to the pyramid basement (greater thickness) in contact with the church floor indicates that the heat source could be attributed to the presence of higher moisture content.

Another issue regarding the lower temperature of the statues compared to the supporting steps was taken into consideration. This different behaviour can be ascribed to the construction typology of the statues, which have a lower mass with respect to the steps of the pyramid, and consequently, they have a lower humidity content.

Thermographic investigations of the roof and upper parts of vertical walls indicated that water infiltrations were absent [15]. A thermographic survey, in a transient heating regime, was also carried out on the detachment areas visually observed. This technique allows us to individuate deterioration areas which were not visible because they are in a latent phase which precedes the detachment which will take place successively.

The most important finding from the above investigations was the monitoring of detachment areas which have a greater extension than those visually perceived, i.e., in some areas of the Sleeping Genius’s head (Figure 5 a,b). These areas are characterized by a pronounced convexity (strongly curved surfaces) in which the more frequent evaporative cycles are taking place, and they are responsible for a heavier deterioration process compared to the nearby flat areas or areas with a lesser degree of curvature [3].

Microclimatic studies on the monument established that the droplet formation ascribed to condensation processes is not very frequent. According to these studies, biological colonization is responsible for the increase in volume which is detaching thin marble layers from the substratum [14].

As this biological mechanism seemed not to be reliable, the people responsible for the conservation asked for the reliability of this hypothesis to be verified.

3. Present Survey

Our diagnostic plan was designed to identify the causes of alteration taking into account the building construction techniques of the monument. In fact, many problems were probably related to the structure, which is supporting all the weight of the marble blocks of the steps, the pyramid, and the sculptures. As the brick structure is in close contact with the moisture of foundations, it was also decided to investigate the water content in order to evaluate if the capillary rise is influencing the presence of water droplets on the marble surface (see Figure 3), and if it may be related to salt efflorescence occurrence.

The investigations above discussed established that droplet formation could be ascribed to the migration of capillary rising damp in correspondence to high tide events. However, the droplets percolating on the surface did not show the formation of salts along the edges of their percolation path, as has been frequently observed in the church of Santa Maria dei Miracoli [16,17] (see Figure 6). As the conclusions of previous studies were not explaining the sharp increase in deterioration processes observed in different areas, that has caused the loss of large pieces of marble details from the surface of the statues. The superintendents responsible for funerary monument conservation decided to carry out further investigations that could provide useful input for the preparation of the restoration project.

In order to satisfy this condition, our diagnostic plan was focused on studying the existence of the capillary rise phenomenon in relation to the structure characteristics of the cenotaph. At the same time, it was investigated the reason why some areas are more damaged than other ones.

It was surprising that the areas where droplets are present are not very damaged whereas some other parts have suffered serious damage. For instance, the chest of the Sleeping Genius or the robes of the Painting and Architecture statues (Figure 2 a,b) show the presence of salt crystals associated with a marked exfoliation of thin marble layers, while the droplets present on the face of the Sleeping Genius were not causing any damage.

4. Materials and Methods

Areas presenting different alteration features were investigated by sampling small micro-flake specimens. Eleven samples from decayed areas were taken according to standard EN 16085 [18], which was delivered by European Committee TC 346 Conservation of Cultural Heritage. Exfoliation scales representative of decayed areas were taken by using a scalpel, and ensuring they were taken as small as possible. Samples were labeled and transferred in the laboratory within a plastic bag. The localization of the most significant samples is reported in Figure 7, Figure 8, Figure 9 and Figure 10.

In the laboratory, samples were incorporated inside a resin cross-section [19]. An optical microscope (OM) and environmental scanning electron microscope, ESEM Quanta 200, equipped with a dispersive energy micro-analyser, EDAX Falcon, were used. Mineralogical petrographic analysis at OM in polarized light (Nikon alphaphot 2) was also carried out [20]. The morphological characteristics of micro-flake samples were observed by examining 3D samples using secondary electrons of environmental scanning electron microscopy (ESEM). The presence of organic compounds was determined by the micro-FTIR (Fourier transform infrared spectroscopy) micro-ATR (attenuation total reflexion) mode.

Nineteen bulk samples, taken by drilling cores on bricks, were used to determine moisture and salt content to investigate the rising damp process. Anion concentrations such as sulphates, nitrates, chlorides, and oxalates were measured by ion chromatography (IC) Dionex ICS-90 following the European standard EN 16455 [20]. Water content was determined by the gravimetric method, according to EN 16882 [21,22,23,24].

Istrian stone is a microcrystalline limestone with a color variable from gray-green to pale yellow. It is a micritic limestone or mudstone rock formed by the diagenesis of calcareous mud, composed of very fine-grained calcite crystals (<4 μm). It is a very compact stone with very rare small pores (Figure 11), with a few sedimentations, and stylolite dissolution planes, as well as sedimentary joints, are sometimes very frequent, with small deposits of clay minerals often occurring along them, colored by yellow ochre.

Marble sample 2 (Figure 7d) was analyzed at OM in polarized and transmitted light (Figure 12). It is a metamorphic rock with a granular texture, compact isotropy, and calcite granules in the range of 0.06–0.76 mm (granoblastic polygonal structure). It is formed by a pure metamorphosized limestone in the Apuan Alps. Its name is Carrara marble.

5. Results

5.1. Structure Underlying Carrara Marble Statues and Steps and Related Moisture Content

In Venice, the masonry brick structure of historic buildings and churches has an important role in the transmission of moisture from the foundations toward the marble applied as coverings of the brick structure.

As there was no information in the archives regarding the construction typology of the masonry structure supporting marble, it was necessary to make an inspection by removing the lumachella ashlars) and by drilling the compact Istrian stone (thickness 22 cm) supporting the first marble step. The drilling inspection of the monument structure shows a rather regular brick structure, which in turn is in contact with the lagoon water (Figure 13).

The moisture content in the brick structure supporting the marble steps and the statues was measured at various heights (

Table 1. Percentage of water content (gravimetric method EN 16882), chlorides, nitrates, and sulphates (ion chromatography method EN 16455) at different heights.

Heights (cm)	H2O (%)	Chlorides (%)	Nitrates (%)	Sulphates (%)
10	26.1	0.01	0.01	0.04
80	25.8
90	21.3	0.05	0.06	0.02
110	18.5
130	17.2
140	19.7
185	8.5	0.87	0.18	0.04
235	2.4	0.51	0.39	0.07
335	2.3	0.30	0.21	0.02

).

On the church floor, bricks are completely saturated with a moisture content of around 26%. This high-water content is maintaining rather constant (21–26%) up to 90 cm of elevation. Above this height, moisture content is slowly decreasing up to 130 cm of elevation (17.2%). Above this height, the water content is decreasing more rapidly, and at the pyramid base (200 cm), it is reduced to 8.5%. Above the pyramid base, at 235 cm of elevation, the moisture drops furtherly to 2–3%, that is corresponding to the maximum height of the capillary rise front (Figure 14).

In contrast with the high moisture content of the brick structure, the Istrian stone blocks, in contact with the brick structure, have a very low moisture content, approximately 50 times lower. This is due to the different porosity of the two materials. From the measurements taken, it is clear that the very porous brick structure is close to moisture saturation. Moisture drops dramatically in the Istrian stone blocks (22 cm thickness) because the latter, having a very low porosity, absorbs limited quantities of water. The same happens for marble. The larger extension of deterioration areas is irregularly localized, approximately from 150 to 250 cm.

In Table 1, at a higher level (185–235 cm), chlorides and nitrates are present in significant quantities with respect to sulphates, thus confirming the results obtained by other authors [4]. The chloride concentration trend reported in Figure 15 shows their accumulation in correspondence to the water capillary rise front where the lower moisture content (2.3–2.4%) occurs.

From the data, the capillary rise phenomenon is active up to an elevation of approximately 180–250 cm, where the solution is enriched in chlorides at a sub-surface level, and sodium chloride crystallizing is causing the exfoliation and disintegration of the surface marble layer.

5.2. Macroscopic Features of Marble Surface

The observed alteration features can be grouped according to the following classification:

(a)

Rough whitish areas. The surface of the marble is rough, opaque, and generally the white-gray marble surface is turning whitish due to the change in the refractive index consequent to the formation of the exogeneous crystals responsible for the detachment of thin marble layers. These alteration features are present quite widely on the marble steps and on some areas of the sculptures (Figure 16a).

(b)

Areas with exfoliation of superficial marble scales. The alteration process is more advanced than that before described, and exfoliation of superficial marble layers is detaching from the substrate (Figure 16b).

(c)

Areas with missing portions of marble. This is a more advanced decay in which the material has already detached. These visible deterioration forms are present on the Sleeping Genius (Figure 2a), on the statues of Painting and Architecture (Figure 2b), and on some parts of the steps.

(d)

Areas with reddish circular spots (Figure 9a). These appear localized in some areas, and the surface is rough and opaque as in the whitish areas before described (Figure 10).

(e)

Areas of surface turning to grayish, characterized by marble smooth surface without any evident alteration (Figure 17).

5.3. Optical and Scanning Electron Microscope Analyses

In order to explain the different alteration features macroscopically observed, some small samples (see Section 4) were taken from the different areas and analyzed at the optical and scanning electron microscope.

Type: (a) rough and whitish areas (Figure 16a). Cross-section analyses show a fair amount of interstitial micro-cracks caused by the sodium chloride crystallization. This is the initial stage of deterioration in which calcite crystals detach from each other, and the structure becomes less compact and the marble appearance changes from a white-gray to whitish color (Figure 18a,b).

In some areas, the intergranular interstices had extended longitudinally to form a line of discontinuity almost parallel to the external surface (Figure 19). This stage represents a level that precedes the exfoliation of thin marble layers.

Type: (b) areas with exfoliation of superficial marble scales (multilayer). Progressive marble detachment flakes are observed (Figure 7a,b and Figure 20a). Examination of a cross-section of a multilayer exfoliation scale shows eleven flake layers (Figure 20a). Among the various layers, sodium chloride crystals were detected, and they are responsible for the scale’s detachment.

Microanalysis also emphasizes the presence of potassium chloride and gypsum. Sodium chloride crystals were observed in between the exfoliation layer (Figure 21a–c). They are responsible for the detachment of the thin marble layer.

Type (d) Areas with reddish circular spots (Figure 9a,b). The presence of hydrated iron oxides was identified, most likely generated by the oxidation of the iron linchpin that was used to fix the marble blocks. According to some authors, iron is firstly oxidized to ferrous ion (Fe II), which, in the presence of moisture, forms a mixture of ferrous and ferric hydroxide. The hydrated iron oxide migrating in the carbonate substrate reacts with the carbonate anions of the matrix, forming colloidal oxy-hydroxy ferric carbonate Fe^(III)₆ O_(2+X) (OH)_(12+2x) · (H₂O)x (CO₃), which in turn diffuses through the carbonate matrix, and when it reaches the saturation condition, a red precipitate of lepidocrocite (γFeOOH) takes place [25,26,27]. A radial diffusion of this colloidal solution gives the red-orange circular spots (Figure 9 a,b), often with particular shapes, such as Liesegang rings [28]. The formation of Liesegang rings represents a phenomenon of periodic precipitation, observed both in simulated chemical systems constituted by ions scattered in a colloidal medium in a gel state, and in natural systems (sedimentary rocks). Such phenomenon causes the formation of colored ring bands more or less regular. When observed using an optical microscope, the colloidal iron oxides diffuse in the crystal interstices and cover the calcite crystal (Figure 22).

Type: (e) areas of surface turning grayish (Figure 15). The cross-section optical microscopic observation shows a compact internal structure which indicates a good state of conservation. On the surface, the dark color is due to wax remains of a past treatment (Figure 23). On the same surface, a micro FTIR survey was carried out. Traces of acrylic and siloxane compounds were detected (Figure 23). The wax is probably ascribed to ancient treatments, while the synthetic products are ascribed to the 1993 intervention and to some maintenance not reported in the archive documents.

6. Discussion

According to the results obtained, the marble steps and the pyramid basement are supported by a brick structure whose foundations are in direct contact with lagoon water. The capillary rising damp migrating from the foundations is transporting sodium and chloride ions through bricks and marble porosity. With elevation, the driving force is decreasing, and at a certain level, the capillary rising front is stopping because the driving rising force is completely counterbalanced by the water vapour evaporation rate, and the ion concentration in the solution is reaching the oversaturation conditions with consequent salt crystallization at the sub-surface level (crypto-efflorescence) (Figure 18 and Figure 19).

The increase in the volume of the crystals exerts considerable pressure within the capillaries close to the surface, causing the breaking of their walls with the detachment of thin marble flakes, followed by a progressive disintegration of the marble’s calcite crystals. The areas presenting exfoliation and detachment of thin marble layers occur irregularly according to the structure and texture of the material [2,8].

The same mechanism is active all over the monument, and the deterioration features observed represent the different stage of deterioration recorded during the survey.

The ESEM-EDS analysis carried out on different alteration features, macroscopically observed, was a powerful tool to explain the mechanism responsible for deterioration in different areas.

Whitish rough surface areas of marble show an accumulation of sodium chloride at the sub-surface level. The whitish color due to a less dense substrate is ascribed to a change in the refractive index. These areas are quite widely present on the marble steps and in some parts of the sculptures (Figure 10 and Figure 14). Deterioration processes in these areas are at an initial stage.

Another alteration feature, macroscopically observed, is characterized by partially detached marble flakes (multilayer exfoliation) from the substrate showing a more advanced stage of deterioration (second level of deterioration). Crystallization initially occurs near the external marble surface and continues towards the internal layers, separating marble flakes mostly in parallel layers (crypto-efflorescence) (Figure 18).

A third stage of deterioration is characterized by the complete missing of marble portions as it is visible on the Sleeping Genius (Figure 2a) and Painting and Architecture statues (Figure 2b–d). These more damaged areas are corresponding to the maximum height of the capillary rise front (between 150 and 250 cm), which presents the highest salt crystallization pressure.

A specific alteration feature is represented by the formation of reddish circular spots which are localized in some areas, and the surface is rough and opaque, similar to the whitish areas before described (Figure 9a,b). These alterations are ascribed to linchpin oxidation to colloidal iron oxides, which slowly diffuse through the porosity of marble.

SEM observations highlight a dark film covering the marble surface (Figure 23), ascribed to wax application as periodic maintenance operations were usually carried out in the past.

The FTIR technique was useful to investigate the presence of an organic-based treatment (Figure 23).

7. Conclusions

To summarize, the main cause of deterioration is the large amount of moisture coming from the brick structure, which is transporting the soluble salts from the foundations in contact with the lagoon water. The marble steps and pyramid, being closely in contact with the brick structure, are continuously embedded by soluble salt solutions. The water evaporation from the surface of the marble steps causes the enrichment in salt concentration nearby the marble surface, and when ion saturation conditions are reached, crystallization takes place. Mechanical tension at the sub-surface level is breaking the capillary walls.

The steps of the cenotaph are between 100 and 200 cm, and they are affected by the initial stage of deterioration (Figure 18a,b) and, to a minimal extent, by the second stage of deterioration. In the case of the statues of Painting, Architecture, and the Sleeping Genius, the third stage of deterioration is observed at a height of approximately 200 cm. Other areas of deterioration at the third stage can be observed in correspondence with the draperies of the statues and the garland and other convex surfaces that are characterized by more frequent evaporation cycles than that of flat or slightly curved surfaces. The greater frequency of evaporation cycles leads to a greater number of crystallization cycles, and therefore, to an acceleration of the decay processes with greater losses of material [3].

The last observation is related to the observed turning of white-gray marble toward grayish ascribed to a chromatic alteration of the wax-based treatments. Wax application most likely dates back to the final phase of the construction of the cenotaph, and may have been periodically repeated as a maintenance practice in use in subsequent periods.

In the 1993 intervention, acrylic or acryl–siloxane resins were applied, which were identified by our survey and by the previous 2009 survey. This intervention together with the initial wax treatment exacerbated the phenomena of sub-surface crystallization, accelerating the decay processes.

The results obtained suggest that before starting any conservation intervention, it is recommended to stop the rising damp process by isolating the brick structure from the marble steps and statues. Furthermore, since the marble elements were completely salt embedded, it is necessary to remove all statues and marble elements at least up to three meters of elevation from the church floor, where deterioration areas occur. For the marble elements above three meters, it is sufficient to proceed with the removal of the surface wax.

The removed marble statues and steps should be desalinated by immersion in deionized water tanks. A similar procedure had been adopted for the marble slabs of the church of Santa Maria dei Miracoli and for the statues of San Giobbe church in Venice.

The brick structure should be dehumidified and desalinated by using poultices according to EN 17891 [29]. Successively, the brick structure should be insulated to stop the transmission of moisture from the foundations to the marble elements.

The proposed intervention is radical and certainly entailed very significant technological and structural difficulties; however, almost similar interventions had already been undertaken on other Venetian monuments.