Eco-Friendly Protective Coating to Extend the Life of Art-Works and Structures Made in Porous Stone Materials

The application of hydrophobic treatments to stone surfaces is the most common proven method to prevent, or at least limit, the degradation of stone-made constructions and artworks brought about by the ingress and action of water, in particular in the case of very porous stone materials. To avoid the use of protective products containing harmful solvents, new green products have been proposed. In this paper, an eco-friendly hydrophobic coating, based on a fluorine polymer dispersed in water, was deeply analyzed to evaluate its protective properties, especially for very porous stone substrates. To this aim, a wide characterization of treated and untreated Lecce stone elements, i.e., a stone typical of the Apulia region, was carried out to assess the optimum required amount, the effectiveness and the protective capability, even against graffiti staining, of the green hydrophobic treatment, still allowing the stone to retain adequate vapor permeability. The efficacy of the eco-friendly product was analyzed also after a short time (four weeks) of outdoor exposure. Suitable performance and short-term durability of the green hydrophobic coating were found, comparable or even greater than those reported in the current literature for other widespread commercial products, confirming the capability of the product to preserve porous stone surfaces even in absence of solvents in its formulation. The study also allowed to experiment with the “contact sponge” test as an appropriate method for evaluating the water absorption properties of the stone.


Introduction
Porous stone materials, such as Lecce stone, Noto stone, Globigerina stone, Ançã stone, Tuffeau limestone, Maastricht limestone, have been widely employed since ancient times to realize buildings, monuments, ornaments and other works of art, in UNESCO World Heritage sites [1], in many European countries [2] and particularly in the Mediterranean regions (such as the southern Apulia [3], Sicily [4], and Malta [5]). Their characteristic high porosity makes them suitable for being worked and modeled to create true masterpieces; this feature is, at the same time, responsible for their high vulnerability towards environmental agents, foremost water; consequently, they can easily undergo degradation which compromises their mechanical characteristics, functionality and external appearance.
Water can easily penetrate porous stones either in liquid or vapor forms, it can contain soluble salts, it can change its physical state according to the external temperature: all these processes are particularly damaging to the porous materials and determine their limited durability over time, especially if the stone is outdoor exposed. Therefore, the need arises to protect the stone-made historical buildings and artworks, also to preserve the Cultural Heritage, representing the local history, culture and traditions, for future generations. A fluorine-based system was chosen since such chemical products are considered excellent protective agents and are among the most widely used materials for stone protection. Coatings obtained from fluorinated polymers typically exhibit both high hydrophobic properties [22] and anti-graffiti protection [23]; being the C-F bond a strong chemical bond, good durability and chemical stability are also expected and observed [24].
A highly porous calcarenite, named ''Lecce stone'', was chosen as the substrate to test the protective product. This natural stone material originates from quarries located in the Salento region (Apulia, southeastern Italy), specifically near the city of Lecce (hence the name "Lecce stone"). Typical of Baroque architecture, "Lecce stone" has been employed in the built heritage in this area since ancient times. Still widely used in many historic and civil buildings, this material has known an increasing diffusion also out of Italy as a decorative dimension stone. Besides, due to its characteristics, "Lecce stone" can be regarded as a reference of porous materials used for monumental heritage and civil buildings in many European countries. Noto stone in Sicily [25], Globigerina limestone in Malta [26], Tuffeau limestone in the Loire Valley (France) [27], Maastricht limestone in northern Europe [28], Ançã stone in Portugal [29] are some examples of stone materials similar to "Lecce stone" for both features and processes of decay. A fluorine-based system was chosen since such chemical products are considered excellent protective agents and are among the most widely used materials for stone protection. Coatings obtained from fluorinated polymers typically exhibit both high hydrophobic properties [22] and anti-graffiti protection [23]; being the C-F bond a strong chemical bond, good durability and chemical stability are also expected and observed [24].
A highly porous calcarenite, named "Lecce stone", was chosen as the substrate to test the protective product. This natural stone material originates from quarries located in the Salento region (Apulia, southeastern Italy), specifically near the city of Lecce (hence the name "Lecce stone"). Typical of Baroque architecture, "Lecce stone" has been employed in the built heritage in this area since ancient times. Still widely used in many historic and civil buildings, this material has known an increasing diffusion also out of Italy as a decorative dimension stone. Besides, due to its characteristics, "Lecce stone" can be regarded as a reference of porous materials used for monumental heritage and civil buildings in many European countries. Noto stone in Sicily [25], Globigerina limestone in Malta [26], Tuffeau limestone in the Loire Valley (France) [27], Maastricht limestone in northern Europe [28], Ançã stone in Portugal [29] are some examples of stone materials similar to "Lecce stone" for both features and processes of decay.
"Lecce stone" is composed of calcite as the main constituent (93-97% [30]), very small quantities of clay and phosphates [31], as well as other non-carbonate minerals [32]. "Lecce stone" is frequently affected by deterioration phenomena (such as alveolization, erosion, rising damp, salt crystallization), in which water is the factor causing most of the processes of decay. The specimens of "Lecce stone" used in this study exhibited a porosity of 42% with an average pore radius of 1.23 µm and pore sizes mainly between 0.5 and 4 µm, as evaluated by Mercury Intrusion Porosimetry [20].
Prismatic stone specimens, with dimensions of 5 × 5 × 1 cm 3 , were cut by a saw from quarry blocks. Following the UNI10921 standard protocol [33], the specimens were smoothed with abrasive paper (180-grit silicon carbide), numbered using an engraver tool, washed with deionized water and cleaned with a brush in order to remove dust deposits. The stone specimens were left 24 h in the laboratory, at a temperature of 23 ± 2 • C and relative humidity (R.H.) of 45 ± 5%; then, they were completely dried in an oven at 60 • C, until the dry weight was achieved. Finally, the samples were stored in a desiccator with silica gel (R.H. = 15%) at 23 ± 2 • C.
The removal of the paint was, then, performed using a commercial chemical remover (trademark ECO 7 G, provided by C.I.R. Chimica Italiana Restauri, Arezzo, Italy) formulated for graffiti removal on stone materials, also for monuments and historic buildings. This remover, ready-to-use, is a mixture of esters, surfactants, and emulsifiers; it does not contain chlorinated solvents, all its components are declared as biodegradable. The remover is a transparent gel ( Figure 1c) having a density of 1.02 kg/L; it can be applied by either brush or paint roller. The time of action should range between 10 min and 12 h, depending on both the porosity of the substrate and the staining agent.

Surface Protection
The first stage of the research was aimed at assessing the optimal quantity of protective product to apply on "Lecce stone". Starting from the amounts suggested in the technical data sheet for porous natural stone materials (i.e., 2-5 m 2 /L), four different quantities of product were chosen. Details about the treatments are summarized in Table 1. The application by brush (Figure 1d) was used to simulate a typical procedure commonly employed in field conditions. Before the application of the product, the samples of stone were conditioned in equilibrium with the surrounding environment (24 h in the laboratory, at 23 ± 2 • C and 45 ± 5% R.H.). Only a 5 × 5 cm 2 side for each specimen was treated (namely, the side without the inscribed sample's number). After the application of the product, all the specimens were kept in laboratory conditions, at 23 ± 2 • C and 45 ± 5% R.H., for 30 days. Then, they were dried in the oven at 40 • C until a constant weight was achieved, controlling the stabilization by periodical measurements of weight.
Color measurements and water absorption by the "contact sponge" test were performed to assess the harmlessness and the efficacy of the treatments, respectively. These methods were selected to evaluate the fundamental performance of the protective coating in a quick and simple manner, using standard tests easy to apply and reproduce in the field, which have been proposed also by other research groups for in situ analysis in historical buildings, for instance, monuments of Matera [34] and Catania [35]. Contact angle measurements were also carried out to control the surface hydrophobicity.
As it will be illustrated and discussed in Section 3.1, the treatment labeled with PT 3 yielded a more suitable performance. Therefore, only the PT 3 specimens (i.e., those related to the coverage of 3 m 2 /L) were taken to complete the assessment of the protective capability of the treatment. In particular, a water vapor transmission test was performed to control the harmlessness of the treatment and its compatibility with the stone material; Coatings 2021, 11, 1270 5 of 16 capillary water absorption was also monitored to evaluate the efficacy of the applied coating against water ingress.

Anti-Graffiti Evaluation
Staining was carried out with the acrylic blue-colored paint (Figure 1e), applied on both untreated and protected stone samples. The paint was sprayed in two coats on specimens placed on a 45 • tilted surface. The lateral sides of the specimens were protected with a PET film. The distance between the sample surface and the nebulizer was about 15 cm. After the application of the paint, the samples were stored for 2 days in laboratory conditions (23 ± 2 • C, 50 ± 5% R.H.).
The sprayed paint was removed, 15 days after its application, using the chemical remover described in Section 2.1. According to the recommendation reported in the technical data sheet, the remover was applied by brush ( Figure 1f). After 25 min of action, a partial absorption into the stone structure of both the gel remover and the paint started taking place. Therefore, to avoid that the movement of the paint beneath the surface could reduce the cleaning efficacy, removal with a sponge was carried out 30 min after the application of the gel remover. Finally, the surfaces were rinsed using running tap water and a sponge ( Figure 1f). The samples, dried for 24 h in laboratory conditions (i.e., 23 ± 2 • C and 50 ± 5% R.H.), were subjected to color measurements to quantify the cleaning efficacy. Due to initial unsatisfactory results, the removal was repeated using the same procedure and the final cleaning efficacy was calculated.

Outdoor Exposure
To assess the durability of the coating against natural weathering, untreated and protected stone samples were outdoor exposed. Color measurements, water absorption by the "contact sponge" test and contact angle measurements were performed just before the exposure. The specimens were, then, placed at 45 • on a holder (Figure 1g), with the surfaces exposed to the south, on the roof of the Department of Innovation Engineering building, on the Ecotekne Campus of Salento University, Lecce (latitude = 40 • 20 02.5 N, longitude = 18 • 06 50.0 E). The exposure started in mid-May. Some of the specimens were removed after 4 weeks of exposure, and the experimental investigations were repeated on these samples.
The meteoclimatic data measured during the period of exposure by a weather station close to the site of exposure are reported in Table 2.

Experimental Investigations
Color measurements [36] were performed with a Konica Minolta spectrophotometer CM-700d; CIE Standard illuminant D65 and the target mask 8 mm in diameter were used. Ten measurements were performed on each specimen and the instrument was recalibrated to a white calibration cap before each measurement session. The color coordinates were expressed in the CIE L* a*b* (1976) color space: L*, ranging from 0 (representing black) to 100 (white), indicates the lightness/darkness; a* correspond to the red (positive values) green (negative values) coordinate; b* is the yellow/blue coordinate (with positive values related to yellow and negative to blue). The total color changes (∆E* ab ) were determined by the formula: where: ∆L*, ∆a*, and ∆b* are the variation in each color parameter. The color variations, after the coating application and in the evaluation of the antigraffiti efficacy, were calculated in comparison to the untreated surface; the color changes after the outdoor exposure were evaluated using the color parameters measured just before and after the exposure.
The water absorption was investigated by a standardized procedure [37] feasible in the field, that is, the "contact sponge" test [38]. A Petri dish, 3.5 cm in diameter, containing a moist disc-shaped sponge (area of 9.29 cm 2 ), was used. All the tests were performed on stone samples dried in an oven for 24 h at 40 • C, then, stored for 24 h in laboratory conditions (at 23 ± 2 • C and 45 ± 5% R.H.). First, 1 mL of water was added to the moist sponge inside the Petri dish and the "contact sponge" system (i.e., Petri dish, sponge and added water) was weighted. The lid was removed and the sponge inside the dish was pressed against the specimen placed almost vertically (tilted at 45 • ), to simulate a small part of a wall. To assure that the whole surface of the sponge touched the stone, a sponge, approximately 1 mm thicker than the Petri dish, was used; furthermore, the pressure suitable to provide complete contact between the dish's borders and the surface was applied. The time of contact between the wet sponge and the surfaces under investigation was fixed at 1 min. The lid was promptly replaced to avoid water evaporation and the "contact sponge" system was weighed again. The water absorption was, then, calculated as follows: where: m i is the initial weight of the "contact sponge" system; m f is the weight of the "contact sponge" system after the contact with the stone surface; A is the area of the sponge; t is the contact time. At the end of the test, the sponge was squeezed to remove the residual water, each specimen was dried in an oven at 40 • C for 24 h. Following the procedure described in the European standard [39], water-stone static contact angle measurements were carried out. A Costech apparatus (Costech International S.p.A., Cernusco sul Naviglio, Milan, Italy) was used to deposit micro-drops of deionized water on the stone surfaces, while a connected camera recorded the image of the drop; the related contact angle was calculated by means of the "anglometer 2.0" software (Costech). Thirty measurements for each specimen were performed on different positions of the surface, averaging the obtained results. The image of each drop was acquired 15 s after deposition to assure the reproducibility of the test. It is to point out that the contact angle was not determinable for the untreated "Lecce stone": due to the high porosity of such stone material, in fact, water absorption was very high and rapid, hence the drops of water were suddenly absorbed.
The water vapor transport properties of the stone were assessed by the vapor transmission test. The parameters were evaluated at 20 • C, placing the containers with the samples into desiccators with silica gel and storing them in a climatic chamber (ACS Angelantoni Climatic Systems, Mod.UY 600, Massa Martana, Perugia, Italy). Every 24 h, weight measurements were carried out to calculate the rate of vapor transport through the sample from water in the container (where the R.H. was very close to 100%) to the regulated atmosphere into the desiccator (R.H. 15%, 1 atm). Further details of the procedure are reported in a previous paper [40]. The cumulative mass decrease recorded during the test was plotted versus time and the water vapor flow rate (G) was calculated as the slope of the curve. The water vapor transmission rate (WVTR), referred to as permeability in [41], being the mass of water vapor passing through the surface unit in the unit time (24 h), was calculated by the equation: where: ∆M, expressed in g, is the weight change in the steady state water vapor diffusion flow through the specimen (achieved when the difference between two subsequent measurements is lower than 5% [41]); A is the area (in m 2 ) exposed to water vapor; t is the unit time (24 h). In this study, ∆M was calculated as the average of five consequent values of the daily difference in weight; the exposed area was 0.001611 m 2 .
The reduction of the vapor permeability (RVP) was also quantified after the application of the protective coating, using the following formula [42]: where: ∆M u and ∆M x are the weight changes in the steady state for the untreated and the treated samples, respectively.
The capillarity water absorption test was performed according to the European standard [43]. The test was carried out for 6 days. At each time of exposure, the amount of absorbed water (Qi) was calculated as follows: where: w i and w 0 are the weight of the sample at time t i and t 0 , respectively; A is the area exposed to water. Q i values were plotted versus the square root of time to evaluate the trend of water absorption. All the weight measurements were acquired by an analytical balance (Sartorius AG, Model BP 2215, Göttingen, Germany) with an accuracy of ±0.1 mg.
During the preparation of samples, the application of products and the performance of tests, the environmental conditions were continuously monitored by a thermo-hygrometer (Mod. EMR812HGN, Oregon Scientific Inc., Tualatin, OR, USA) able to collect temperature from −50 • C to 70 • C (with a resolution of 0.1 • C) and relative humidity in the range 2-98% (with a resolution of ±1%).
After the removal of the spray paint, the cleaning efficacy (CE) was evaluated as a percentage by the following Equation [18]: where: (∆E* ab ) cleaned is the color variation of the cleaned surfaces; (∆E* ab ) stained is the color variation of the stained surfaces.

Evaluation of the Optimal Amount of Product
Color parameters, water absorption measured by the "contact sponge" test, and waterstone static contact angle values, determined on the stone samples before and after the treatments performed with the eco-friendly protective product, as described in Section 2.2.1, are reported in Table 3.
After the application of the coating in different quantities, very slight and comparable variations in the color parameters were measured, regardless of the amount of treatment. The following trend was observed: not significant decreases in L*; small increases in a* and b* ( Table 3). As illustrated in Figure 2, a very low value was measured for ∆E* ab for all the samples, it did never exceed the value perceivable by the naked eye, that is, ∆E* ab = 3 [44,45]. The observed colored variations are lower than those obtained after the application of other commercial protective products on Lecce stone. ∆E* ab around 4 units was measured applying fluorine-based polymers [17]; using siloxanes as coatings, ∆E* ab higher than 2 units were found [46], while lower values were obtained only using lower but ineffective amounts of product [40]. The different amounts of protective product applied on Lecce stone caused similar negligible color changes to the stone surfaces: the lowest ∆E* ab value was measured, in fact, for PT 2 samples (i.e., the surfaces treated with the highest quantity of product). Table 3. Color parameters (L*, a*, b*), color change (∆E* ab ), water absorption by "contact sponge" test (WA), and water-stone static contact angle (α) in the samples before and after the application of the protective treatments.

Samples
Before  The water absorption, measured by the "contact sponge" test, strongly declined after the treatments; reductions above 96% was calculated in all the samples, in accordance with the results obtained during a monitoring campaign carried out on the façades of monuments, made with Lecce stone, treated with siloxane polymers [47].
It is to highlight, however, that, as reported in Table 3 and illustrated in Figure 3 the highest amount of product (PT2 specimens) did not assure the highest reduction in water absorption: decreases of 99% were measured for the samples PT3, PT4, and PT5, while a reduction of 96% was found for PT2. Similar behavior was observed in a previous study [40] when a larger quantity of applied protective product caused a poor barrier against water ingress due to the presence of fractured coatings. Cracks in coatings can be, in fact, generated during the drying/hardening [48] as a consequence of shrinkage promoted by the evaporation of water, solvents or reactions' by-products, and they are frequently The water absorption, measured by the "contact sponge" test, strongly declined after the treatments; reductions above 96% was calculated in all the samples, in accordance with the results obtained during a monitoring campaign carried out on the façades of monuments, made with Lecce stone, treated with siloxane polymers [47].
It is to highlight, however, that, as reported in Table 3 and illustrated in Figure 3 the highest amount of product (PT 2 specimens) did not assure the highest reduction in water absorption: decreases of 99% were measured for the samples PT 3 , PT 4 , and PT 5 , Coatings 2021, 11, 1270 9 of 16 while a reduction of 96% was found for PT 2 . Similar behavior was observed in a previous study [40] when a larger quantity of applied protective product caused a poor barrier against water ingress due to the presence of fractured coatings. Cracks in coatings can be, in fact, generated during the drying/hardening [48] as a consequence of shrinkage promoted by the evaporation of water, solvents or reactions' by-products, and they are frequently observed in siloxane-based coatings [49]. Anyway, in any kind of coating, the presence of fractures is strongly dependent on the coating's thickness [50], with a critical value above which the cracks start to occur [51]. As explained in Section 2.3, water-stone contact angle cannot be determined on untreated stone surfaces, while contact angles greater than 130° were measured on all the protected samples, as reported in Table 3. No clear dependence on the quantity of applied product was observed, the lowest value was measured on the PT4 sample.
On the basis of these results, the evaluation of color changes did not allow discriminating the proper amount of product. Trying to find a balance between strong reduction in water absorption and high surface hydrophobicity, the quantity corresponding to the coverage of 3 m 2 per liter of product (i.e., PT3 treatment) was selected to complete the evaluation of the protective treatment.

Harmlessness of the Treatment
Color variations as a consequence of conservative treatments on stone surfaces certainly have a great impact since they can be appreciated also by the naked eye. Changes in water vapor transport properties are other important parameters to take into consideration in evaluating the harmlessness of the treatment itself. This is because, if the permeability to water vapor is strongly reduced after the application of a coating, water may condensate at the interface between the treated and untreated stone regions, thus producing mechanical stress able to activate processes of decay [52].
With this perspective, the water vapor transport properties were evaluated on the same specimens before and after applying the PT3 treatment. The mass changes measured during these tests are reported in Figure 4: linear trends were always found, with the steady state of water vapor diffusion flow through the specimen observed from 72 h after the beginning of the test up until its end. Slight reductions were measured for the water vapor transport parameters (reported in Table 4) which are, however, comparable to those obtained for neat "Lecce stone" [40] and for the same material protected with commercial coatings; in other studies, WVTR was found approximately 180 g/m 2 × 24 h after the In the PT 2 samples, due to the greater amount of applied product, the presence of a thicker coating can be inferred: thus, an excessive accumulation of the polymer on the stone surface cannot be excluded. This condition might be the cause of coatings affected by microcracks which reduce their ability to act as a barrier against water ingress into the stone. Further investigations are in progress to confirm this hypothesis.
As explained in Section 2.3, water-stone contact angle cannot be determined on untreated stone surfaces, while contact angles greater than 130 • were measured on all the protected samples, as reported in Table 3. No clear dependence on the quantity of applied product was observed, the lowest value was measured on the PT 4 sample.
On the basis of these results, the evaluation of color changes did not allow discriminating the proper amount of product. Trying to find a balance between strong reduction in water absorption and high surface hydrophobicity, the quantity corresponding to the coverage of 3 m 2 per liter of product (i.e., PT 3 treatment) was selected to complete the evaluation of the protective treatment.

Harmlessness of the Treatment
Color variations as a consequence of conservative treatments on stone surfaces certainly have a great impact since they can be appreciated also by the naked eye. Changes in water vapor transport properties are other important parameters to take into consideration in evaluating the harmlessness of the treatment itself. This is because, if the permeability to water vapor is strongly reduced after the application of a coating, water may condensate at the interface between the treated and untreated stone regions, thus producing mechanical stress able to activate processes of decay [52].
With this perspective, the water vapor transport properties were evaluated on the same specimens before and after applying the PT 3 treatment. The mass changes measured during these tests are reported in Figure 4: linear trends were always found, with the steady state of water vapor diffusion flow through the specimen observed from 72 h after the beginning of the test up until its end. Slight reductions were measured for the water vapor transport parameters (reported in Table 4) which are, however, comparable to those obtained for neat "Lecce stone" [40] and for the same material protected with commercial coatings; in other studies, WVTR was found approximately 180 g/m 2 × 24 h after the application of fluorine-based products [17], while ranging between 166 and 188 g/m 2 × 24 h, depending on the applied amount of product, using siloxane-based protective coatings [40].  The permeability decrease (reported as RVP in Table 4) was to be very low, with the measured reduction well below the threshold of 20% accepted as tolerable for treated stone materials in buildings and monuments [53]. These results gave evidence of the harmlessness and compatibility of the applied treatment with respect to the "Lecce stone". Table 4. Water vapor flow rate (G) and water vapor transmission rate (WVTR) in the untreated and protected stone samples; the reduction of the vapor permeability (RVP) after the protective treatment is also reported.

Barrier against Water Ingress
Capillary water absorption curves, determined on the stone samples before and after the PT3 treatment, are compared in Figure 5.  Table 4. Water vapor flow rate (G) and water vapor transmission rate (WVTR) in the untreated and protected stone samples; the reduction of the vapor permeability (RVP) after the protective treatment is also reported.

Untreated
Protected The permeability decrease (reported as RVP in Table 4) was to be very low, with the measured reduction well below the threshold of 20% accepted as tolerable for treated stone materials in buildings and monuments [53]. These results gave evidence of the harmlessness and compatibility of the applied treatment with respect to the "Lecce stone".

Barrier against Water Ingress
Capillary water absorption curves, determined on the stone samples before and after the PT 3 treatment, are compared in Figure 5. Referring to the specimens without the protective coating, most of the water absorption (i.e., 75%) occurred within the first 10 min of exposure; at the end of the test (after 7 days of exposure), the total amount of absorbed water was 3.24 kg/m 2 ; this result is comparable to that reported in the literature for this uncoated stone, that is, water absorption very low at the beginning of the water exposure [54], achieving values of approximately 4 kg/m 2 [55] or above [56]. After the application of the treatment, the water absorption was strongly limited, especially in the early steps of the test. Over the whole experiment performed on the protected specimens, the absorption values were substantially lower than those measured on the uncoated samples of Lecce stone. Only 55% of the absorbed water was gained after 24 h of exposure; at longer times, the differences among treated and untreated specimens were reduced, with the total quantity of water absorbed by capillarity at the end of the test of 2.88 kg/m 2 , that is, −11% in comparison to the result found for the samples without protection. In other studies, at the end of the test (i.e., 8 days), where commercial siloxane-based polymers were applied on Lecce stone the capillary water absorption still remained very low (about −90% in comparison to the untreated stone [46]), while using commercial fluorinated coatings the protective action against water ingress was almost completely lost after the first 24 h [18].
The observed behavior is in agreement with the strong reduction in water absorption measured by the "contact sponge" test. Although this latter experiment evaluates only the short-term water uptake, it has similar sensitivity if compared to the capillary rise method [57] and other well-established tests for evaluation of water absorption [38]. Therefore, such a result confirms that significant information about the water absorption behavior can be obtained by a simple and cheap "contact sponge" test, easily feasible also in the field [47].

Anti-Graffiti Action
In addition to the good protective performance illustrated in the previous sections, the applied product acted also as a good anti-graffiti since the cleaning of the stone surfaces stained by spray paint was more effective in the area where the protective coating was applied.
A first cleaning step allowed achieving a CE value of 80% for the treated samples, while CE below 50% was obtained on the stained neat stone subjected to the same procedure of paint removal. In fact, CE = 80% is the lower threshold to judge as adequate Referring to the specimens without the protective coating, most of the water absorption (i.e., 75%) occurred within the first 10 min of exposure; at the end of the test (after 7 days of exposure), the total amount of absorbed water was 3.24 kg/m 2 ; this result is comparable to that reported in the literature for this uncoated stone, that is, water absorption very low at the beginning of the water exposure [54], achieving values of approximately 4 kg/m 2 [55] or above [56]. After the application of the treatment, the water absorption was strongly limited, especially in the early steps of the test. Over the whole experiment performed on the protected specimens, the absorption values were substantially lower than those measured on the uncoated samples of Lecce stone. Only 55% of the absorbed water was gained after 24 h of exposure; at longer times, the differences among treated and untreated specimens were reduced, with the total quantity of water absorbed by capillarity at the end of the test of 2.88 kg/m 2 , that is, −11% in comparison to the result found for the samples without protection. In other studies, at the end of the test (i.e., 8 days), where commercial siloxane-based polymers were applied on Lecce stone the capillary water absorption still remained very low (about −90% in comparison to the untreated stone [46]), while using commercial fluorinated coatings the protective action against water ingress was almost completely lost after the first 24 h [18].
The observed behavior is in agreement with the strong reduction in water absorption measured by the "contact sponge" test. Although this latter experiment evaluates only the short-term water uptake, it has similar sensitivity if compared to the capillary rise method [57] and other well-established tests for evaluation of water absorption [38]. Therefore, such a result confirms that significant information about the water absorption behavior can be obtained by a simple and cheap "contact sponge" test, easily feasible also in the field [47].

Anti-Graffiti Action
In addition to the good protective performance illustrated in the previous sections, the applied product acted also as a good anti-graffiti since the cleaning of the stone surfaces stained by spray paint was more effective in the area where the protective coating was applied.
A first cleaning step allowed achieving a CE value of 80% for the treated samples, while CE below 50% was obtained on the stained neat stone subjected to the same procedure of paint removal. In fact, CE = 80% is the lower threshold to judge as adequate for the removal of the staining agent [18]. These results were also evident to the visual inspection by the naked eye: as observable in Figure 6, the paint was partially removed and only in limited areas of the untreated stone surfaces; on the other hand, most of the protected stone surface appeared without the blue paint, even if some parts were still affected by the staining agent.
Coatings 2021, 11, x FOR PEER REVIEW 13 of 18 for the removal of the staining agent [18]. These results were also evident to the visual inspection by the naked eye: as observable in Figure 6, the paint was partially removed and only in limited areas of the untreated stone surfaces; on the other hand, most of the protected stone surface appeared without the blue paint, even if some parts were still affected by the staining agent. A second removal step was performed in order to try to improve the cleaning. Actually, the second cleaning resulted in more effectiveness on both coated and uncoated specimens. However, the paint was well visible on the untreated stone surfaces, with a CE value that increased to 63%, i.e., a value still not acceptable; CE above 90% was calculated for the protected samples, confirming an effective stain removal, as also clearly evident to the visual observation ( Figure 6). It is to point out that for Lecce stone surfaces, protected with other commercial anti-graffiti systems (either siloxane or fluorine-based) and stained with blue acrylic spray paint, the CE values did not exceed 75% [18].
As described in Section 2.2.2, the time of remover's application was restricted to 30 min because, during the first cleaning (after 25 min), a partial absorption into the stone of both remover and paint was observed. Using the double procedure, the penetration of the paint inside the stone was avoided or strongly restricted, at least. Therefore, the cleaning performed in two shorter steps, rather than only one remover's application twice longer, was successful and probably more effective. On the other hand, the penetration of the staining agent into the pores of the substrate can compromise the cleaning efficacy, especially in a porous stone material, such as Lecce stone [23].

Durability to Short-Term Natural Aging
The changes in the color parameters as a consequence of the 1-month outdoor exposure resulted very low, as reported in Table 5. Small increases in L* and decreases in a* and b* were, in fact, measured. Similar variations were found irrespective of the performed treatment: only the change in b* was higher for the protected samples with respect to the untreated ones. A lower b* parameter contributed to obtaining ΔE*ab more pronounced in the protected samples, achieving values above 3 units. As already reported in Section 3.1, although the latter result can mean color changes perceivable by the naked eye, it is to take into account that in the field of restoration of heritage buildings, a total color difference smaller than 5 units is judged acceptable in evaluating the effects of stone A second removal step was performed in order to try to improve the cleaning. Actually, the second cleaning resulted in more effectiveness on both coated and uncoated specimens. However, the paint was well visible on the untreated stone surfaces, with a CE value that increased to 63%, i.e., a value still not acceptable; CE above 90% was calculated for the protected samples, confirming an effective stain removal, as also clearly evident to the visual observation ( Figure 6). It is to point out that for Lecce stone surfaces, protected with other commercial anti-graffiti systems (either siloxane or fluorine-based) and stained with blue acrylic spray paint, the CE values did not exceed 75% [18].
As described in Section 2.2.2, the time of remover's application was restricted to 30 min because, during the first cleaning (after 25 min), a partial absorption into the stone of both remover and paint was observed. Using the double procedure, the penetration of the paint inside the stone was avoided or strongly restricted, at least. Therefore, the cleaning performed in two shorter steps, rather than only one remover's application twice longer, was successful and probably more effective. On the other hand, the penetration of the staining agent into the pores of the substrate can compromise the cleaning efficacy, especially in a porous stone material, such as Lecce stone [23].

Durability to Short-Term Natural Aging
The changes in the color parameters as a consequence of the 1-month outdoor exposure resulted very low, as reported in Table 5. Small increases in L* and decreases in a* and b* were, in fact, measured. Similar variations were found irrespective of the performed treatment: only the change in b* was higher for the protected samples with respect to the untreated ones. A lower b* parameter contributed to obtaining ∆E* ab more pronounced in the protected samples, achieving values above 3 units. As already reported in Section 3.1, although the latter result can mean color changes perceivable by the naked eye, it is to take into account that in the field of restoration of heritage buildings, a total color difference smaller than 5 units is judged acceptable in evaluating the effects of stone treatments [58] as well as the stability over time of the conservative coating [59]. Since the variations in the color parameters were comparable in both the sets of samples (protected and neat stones), the color changes were probably due to the deposition of airborne particles on the surfaces, a phenomenon occurring also in the unprotected samples, rather than to the presence of the coating or to the polymer's degradation, taking also into account the shorter period of exposure (i.e., 1 month). On the other hand, color variations in stone samples treated with a fluorinated polymer were already connected to the presence on the surface of particles from the outdoor environment [60]. Table 5. Color parameters (L*, a*, b*), color change (∆E* ab ), water absorption by "contact sponge" test (WA), and water-stone static contact angle (α) in the samples before and after the application of the protective treatments.

Samples
Before The water absorption, measured by the "contact sponge" test, remained unchanged after 4 weeks of outdoor exposure (Table 5): the coating, therefore, was still able to act as a barrier against water ingress into the stone; the observed small differences were in the range of the experimental error.
Finally, a slight decrease in water-stone contact angle was found after the short-term natural aging, as can be observed from data reported in Table 5. The measured variation was comparable to those reported in other studies for fluorinated protective products on Lecce stone under simulated solar radiation [61], as well as on specimens of compact stone materials exposed outdoor [62].
These results suggest that the polymer coating was not degraded under these conditions of exposure, therefore, their initial properties and performances were retained at shorter exposure times. Further experiments are underway to evaluate the durability of the coating after longer exposure times.

Conclusions
An eco-friendly hydrophobic coating, based on a fluorine-based polymer dispersed in water, was analyzed as surface treatment able to protect a very porous stone, i.e., Lecce stone.
The first part of the investigation was devoted to the optimization of the amount of the product to apply on the surface of Lecce stone elements. The comparison of results of color measurements, water absorption through "contact sponge" tests and contact angle analyses, carried out on the Lecce stone specimens before and after the treatments performed with the protective product under investigation, allowed to identify the quantity of product producing the best balance of hydrophobicity properties and water absorption characteristics, i.e., 3 m 2 per liter of product. This amount was, then, employed for the pursuance of the study. The different tested amounts of the product applied on Lecce stone, on the other hand, produced only negligible modifications of the stone color.
In accordance with the "contact sponge" tests, the water absorption through capillary rise tests confirmed that the ingress of water was greatly reduced when the green coating was applied on the stone surface in comparison with un-protected specimens. Furthermore, the selected amount of protective product proven to produce only slight reductions in the water vapor transport properties of the treated stone with respect to the untreated specimens, allowing the water eventually contained in the substrate to come out and not cause damage to the stone.
As an additional advantage of the green product under analysis, it was also able to behave as a good anti-graffiti treatment: the cleaning of the Lecce stone surfaces, previously marked by a paint, in fact, had better success if the protective coating was applied.
The short-term outdoor durability of the coating applied on the Lecce stone surface was assessed by repeating the same described tests on treated and untreated stone elements after a four-week exposure in Lecce, Italy. It was found that the initial properties and performances of the eco-friendly product remained substantially unchanged after a month of outdoor exposure. Further studies are in progress to assess the durability of the coating after longer exposure times.
The study allowed also confirmation that the "contact sponge" tests can be used as an effective monitoring method to evaluate the water absorption of conservation/protection treatments.
Finally, it is to point out that some questions still remain unclear and require additional investigations. In particular, further experiments are necessary to verify the presence of microcracks in the coatings which might be the reason for lower ability to prevent water absorption. In addition, the evaluation of the coating's durability also needs to be broadened; in this regard, tests performed after longer outdoor exposure times are in progress.