Factors Affecting the Efficiency of Hydrophobic Coatings—Experience from Application on Sandstone

: Protecting stone on facades or exterior art works from deterioration is primarily about protecting them from rainwater. Hydrophobic coatings are widely used for this purpose. Here, two factors affecting the long-term efficiency of some coatings applied on stones were investigated: the number of coating layers and the curing time after their application. Tests of water absorption by capillarity, absorption at total immersion in water, and a visual check of the penetration depth have been carried out. The coating’s efficiency coefficient C ef was defined as the ratio of the maximum water absorption of a treated sample to an untreated one. Two commercial silicon-based coatings were applied on the highly porous Ho ř ice sandstone alternatively. Curing times of 2 days vs. 2 weeks, and 2 coating layers vs. 3 layers were compared. The experiments showed that the coating’s efficiency is affected more by the curing time than by the number of applied coating layers. The curing time of 2 days after coating’s application is too short, but 2 weeks proved to be sufficient for both tested coatings. There was no big difference regarding the number of coating layers; two layers seem to be sufficient if a long rain-free curing time can be guaranteed.


Introduction
Rainwater represents one of the negative climatic factors in construction activities.The presence of water in any form in any building material is unfavorable [1,2].Regarding civil engineering, it is necessary to protect buildings from the effects of water and dissolved salts in order to extend their service life [3][4][5].Hydrophobic protective coatings (known as "impregnation") should create an impermeable layer on the surface and prevent the penetration of water into the material.They are multi-purpose and can be applied on concrete, bricks, plasters, as well as on natural stone used in exteriors (facades, sculptures, monuments).The reduction (or even total elimination) of capillary absorption of wind-driven rain is the most crucial function of hydrophobization [6].Rainwater cannot moisten such surfaces for a long time and will flow down the surface relatively quickly in the form of droplets.
Researchers have worked on protective coatings for building materials around the world for many decades and the problem is still challenging.Coatings' efficiency and durability are the most discussed issues.The most complex evaluation of the durability of impregnation coatings was done on buildings in Munich, Germany [7][8][9].A study of different stone types was carried out in France after 20 to 30 years of such treatment [10].The different external conditions and the broad variety of studied stone material and coating chemistry made the correlation of local results impossible.However, important general conclusions were stated in that study: (a) penetration depth of the impregnation is the most important factor; (b) efficiency and durability of the coatings depend on the stone type; (c) every impregnation should be tested prior to its technical application; (d) even after such a long time, coatings can still be efficient, especially if polysiloxane-based ones were used [10].Similarly, supporting findings were published [11] on hydrophobic effects lasting more than 20 years if correctly applied.
Impregnation can also contribute to the strengthening of softer materials, even if this is not its main function.Additionally, a higher frost resistance of impregnated highly-porous tuffs was presented when compared to untreated ones in the freeze-thaw tests [12].
There are many different hydrophobic products available on the market today that have been researched and can be applied to natural stones [6,[13][14][15].Silicon-based systems are the most popular water repellents in use.However, in individual experimental studies, there are significant differences in the test conditions: different protective products, types of treated building material (mortar, brick, stone, etc.), methods of coating application, external or laboratory research conditions, drying/curing time of the coating after its application, and the like.All this makes a formulation of general conclusions about the use of protective coatings on stones difficult.
Many authors agree that a sufficient depth of penetration into the treated material is the key requirement for achieving the desired durability of a coating [13,16,17].The depth of penetration depends on several parameters, such as porosity of the treated material, water content or degree of saturation, contact time of the substrate with a coating, and the related consumption of a coating [18].It turns out that the time of coating hardening, or so-called curing time, also plays an important role in the durability of the coating.Curing time is the time period of undisturbed action of the coating after its application to the material.This is the aspect which the present paper is focused on.
Partial results of laboratory testing of two siloxane-based hydrophobic coatings are presented in this paper.The treated rock is well-known as Hořice sandstone, which has been widely used in the construction industry as building, decorative, or sculptural stone for a long time.In the presented research, attention is paid to the effect of the curing time of used coatings.Two-day and two-week hardening of the coatings are compared under the same conditions of coating application.Moreover, the effect of the number of applied coating layers was also assessed.
The efficiency of the coatings applied on the sandstone under the specified conditions was evaluated by checking the depth of penetration into the treated sandstone, as well as by water absorption tests.Mainly tests of the absorption by capillarity were carried out, approved as a suitable method in previous research [19,20].

Materials
The Hořice sandstone, quarried in the Czech Republic, was chosen due to its high porosity (resulting in both fast water infiltration and high hydraulic conductivity), as well as due to its mechanical stability in water.This sandstone was widely used for constructions in the Czechoslovak Republic in the 20th century, as well as for reconstruction works after the Second World War.It was the most widespread decorative stone in Slovakia [21], used on public representative buildings, bank buildings, hospitals, statues, and monuments (Figure 1).Many of these buildings are already in need of restoration.The Hořice sandstone is a fine-grained sandstone, variable in color, mostly light beige to yellowish brown, with irregular rusty streaks of limonite pigment.Quartz is the dominant mineral (>90%); other minerals such as glauconite, feldspars, mica, or heavy minerals are minor.The sandstone contains intergranular cement, which is a mixture of kaolinite, illite, and fine quartz grains [22].The basic physical properties of sandstone are as follows: bulk density ρd = 1930 kg•m −3 , total porosity n = 26%, open porosity no = 18.3%, water absorption NWAI = 9.5-12.2%,and uniaxial compressive strength UCS = 29.9MPa [20].The rock material for the research was taken directly from the Podhorní Újezd quarry in the Podkrkonoší area, which is currently the largest sandstone quarry in the Czech Republic.Test cubes with an edge length of 50 mm were sawn from the sandstone blocks taken from the quarry.The efficiency of two chemical products, Antipluviol S and Funcosil SNL, was evaluated.Both may be classified as silicon-based impregnation coatings and are currently available on the market.The product Antipluviol S from the producer MAPEI S.p.A. (Milan, Italy) is a colorless hydrophobic transparent liquid, silane and siloxane based.Funcosil SNL, produced by Remmers Gruppe AG (Löningen, Germany), is a low-molecular siloxane that reacts in the building material with the atmospheric moisture to water-repellent polysiloxane.According to the producer's recommendation, both coatings are also suitable for use on natural stone.They are instantly ready for the application, with no need to dilute or to modify them.Some of their technical specifications are given in Table 1.

Methods
To characterize the pore-size distribution of the sandstone, mercury intrusion porosimetry (MIP or Hg porosimetry) was used.
Two factors were evaluated that can affect the efficiency of the coating: a different number of applied coating layers and different curing (or "hardening") time.The curing can be defined in general as a chemical and/or physical process that is generating the required properties or desired effect.Here, curing time is the minimum time span be-tween applying the coating and reaching the intended effect.However, the curing time of the coating is rather questionable, i.e., when is the curing really completed.Therefore, the efficiency of 2-day-and 2-week-long curing of the coating was compared either with two or three layers of coating (Table 2).Coating layers were applied using a brush, at two-hour intervals according to the producer's recommendations.

Number of Coating Layers
Curing Time 2 Days 2 Weeks (14 Days) To determine the efficiency of applied coatings, the penetration depth was checked and water absorption tests under atmospheric conditions were used.
To determine the penetration depth of a waterproofing coating, the authors of [23] tested four different indicators mixed with the coating before its application.However, to avoid any change of the coating's properties, a different procedure was preferred here.The penetration depth of impregnation liquids was determined on prismatic samples with dimensions of approximately 25 mm × 25 mm × 50 mm by a simple visual test according to [24], where a similar procedure was used on concrete surfaces.After the specified curing time, those coated prisms were sawed perpendicularly to the coated surface into two cubic halves.The fresh cut surface was put in contact with blue-ink-colored water for 1 s to make the coating's penetration depth visible.Since the hydrophobic zone did not absorb as much colored water as the rest of the sample, it created a visible light rim around the dark blue center of the cut.The thickness of this rim was evaluated visually and measured by a caliper.The penetration depth of the coating was measured on sandstone with two and three layers of coating.
The efficiency of the coatings was assessed by capillary absorption tests.Water absorption by capillarity (WAC) is one of the ways water penetrates from the surroundings into the building material.In this process, water is driven by capillary forces (i.e., capillary rise) in open pores of certain small sizes (capillaries) into the material, even against gravity.The test was carried out according to the procedure defined in the technical standard EN 1925 [25].Samples dried to a constant mass (at 70 °C) were immersed 3 ± 1 mm deep into water in a plastic tank on a non-oxidizing grid.The mass increase was due to absorbed water, measured in given time intervals (in minutes): 1, 3, 5, 10, 15, 30, 60, 120, 180, 240, 360, 480, 1440, and 2880 (i.e., 48 h).According to the standard, the test should end at this point, i.e., after 2 days.However, samples were left in the tank for whole 30 days to see what happens.Results were expressed as WAC = f(t) curves.WAC (g•m −2 ) was calculated as the mass of water absorbed by the sample per unit area at time t as follows: where mt (g) is the mass of the sample determined at the time t (s) since the first contact with water, md (g) is the mass of the dry sample, and A (m 2 ) is the area of the sample base in contact with water.These absorption tests were carried out on untreated samples (i.e., without any coating), as well as on samples coated according to the testing program shown in Table 2.There were three replicates for untreated samples and 2x/2d samples, but four replicates for 2x/2w and 3x/2w samples in each series (Table S1).After 48 h the water absorption rate dropped to <1% of the last measured sample mass, so the apparent "steady state" or "constant mass" was reached as defined in the technical standard.However, samples were allowed to saturate for 30 days to see what happened.After this time, the maximum water absorption by capillarity (i.e., final specimen moisture due to capillary rise) NWAC was calculated as follows: where mmax (g) is the mass of the sample saturated with water only through capillary rise to a constant mass and md (g) is the mass of the dry sample.
In the next step, the samples were completely immersed in water (the water reached 10 mm above the surface of the samples) for a period of 30 days, until the samples were saturated to a constant weight.Subsequently, the water absorption by complete immersion at atmospheric pressure (i.e., final moisture after complete immersion when full saturation is already expected) NWAI was determined according to the technical standard EN 13755 [26] as follows: where msat (g) is the mass of the sample saturated with water to the constant mass and md (g) is the mass of the dry sample.
From the water absorption values NWAC or NWAI calculated according to (2) or (3), the efficiency coefficient of the coating Cef can be expressed (or calculated) as follows: where NWACt and NWAIt (%) is NWAC or NWAI of treated sample and NWACu and NWAIu (%) is NWAC or NWAI of untreated sample.
Values of Cef can vary between 0 and 1.The closer the Cef value is to 0, the higher the efficiency of the hydrophobic coating.
After carrying out all absorption tests, the treated samples (from 2x/2w and 3x/2w series) were dried at a temperature of 70 °C.Next, the final dry masses of those samples were determined and compared with their dry masses before treatment and testing.The aim of this procedure was to determine the retention or loss of the coating on the samples by the leaching action of water.

Application and Penetration of Coatings
Applied coatings did not create any crust on the stone surface.The Funcosil SNL coating did not even change the appearance of the stone and preserved its original color.The Antipluviol S slightly changed the color of the stone after applying two layers; the rusty smearing of the sandstone was highlighted (Figure 2).According to Munsell Color Charts [27], the color of the sandstone changed from very light brown (10YR8/2) to yellow (10YR8/6).The waterproofing effect of the coating could be seen already by dropping water onto the coated sandstone surface.A magnified photograph of the water droplet (Figure 3) revealed a water/sandstone contact angle of ca.120°, which proves the surface is hydrophobic [16].The difference of the water-to-sandstone contact angles with and without coating is very clear from Figure 4. Water infiltrated immediately after contact with untreated sandstone; no contact angle could be measured/observed.Droplets on the coated samples show similar angles for both coatings.
The pore-size distribution of the Hořice sandstone is presented in Figure 5.A unimodal size distribution of pores is typical for this rock.The main pore-diameter range is between 100 nm and 100,000 nm.Such pores can be classified as micropores or capillary pores [28].Micropores hold maximum water by capillary forces and lead to high capillary rise.The impregnation depth of the tested coatings on the sandstone is visible in Figure 6.The blue indigo color intensity was used to visually distinguish three zones: 1-hydrophobic zone (original color of the stone), 2-partly hydrophobic zone (light blue color), 3-not impregnated (i.e., hydrophilic) zone (dark blue color).Because of the smooth transition of the color, border lines were drawn very subjectively and the penetration depth cannot be exactly quantified.The penetration depth is irregular even in one sample, probably due to local differences in the grain size and porosity.Thin sections could explain more, but they were not available.The samples from the series 2x/2d showed significantly weaker impregnation than the samples 2x/2w with both types of coatings.A very thin and irregular hydrophobic zone 1 could be recognized only in the sample coated with Funcosil SNL (1 to 2 mm).The color becomes darker towards the sample center (zones 2 and 3).The sample with Antipluviol S shows only colored zones, where the irregular zone 2 is ca. 3 mm; the rest is fully saturated with the ink water, i.e., hydrophilic.
However, the hydrophobic zone 1 can be clearly recognized in all samples from the series 2x/2w.This means that the longer curing time enhanced the hydrophobic effect of the coatings.Some "hardening" of the coatings occurred with time, making the coatings less soluble and more resistant in contact with water.The infiltration of the coatings into the samples is rather irregular, which is also visible in Figure 6.This can be caused by the heterogeneity of the rock material.The same may explain the fact that Funcosil-SNL-covered samples 2x/2w show thinner zones 1 and 2 than the 2x/2d ones.
The best impregnation was achieved in the samples from the series 3x/2w.All three zones can be clearly recognized in samples with both types of coatings, whereby the hydrophobic zones are up to 3 mm thick.It means the impregnation depth of both coating types, Antipluviol S and Funcosil SNL, increases with the higher number of coating layers.It is evident from a visual comparison of samples 2x/2w vs. 3x/2w.

Efficiency of the Antipluviol S
The progress of the WAC tests in time is expressed in Figure 7 where all types of samples are represented: untreated and impregnated ones.Each series of samples is characterized by one representative line (an average of three to four tested samples).Raw data of this plot can be found in Table S1.
The untreated Hořice sandstone showed a very high absorption by capillarity.Within 1 h, WAC reached the value of 8100 g•m −2 , and the saturation continued with slower rate, reaching 10,923 g•m −2 after 30 days.
As expected from the visual tests of the coating's penetration, the coating was not water-resistant enough after 2 days of curing.From the first moment of contact with water it started dissolving or diluting, and samples 2x/2d showed WAC close to 8000 g•m −2 after 2 days in water when their test was stopped.In contrast, sample groups 2x/2w and 3x/2w show identical behavior, with very slow increase of WAC, reaching only 1100 g•m −2 at the end of the test (after 30 days).Based on this experience, 2 weeks proved to be an advisable curing time for the Antipluviol S coating which guarantees sufficient protection with a coating efficiency of 89.8% after 30 days in contact with water.Results of NWAC are compared to NWAI in Table 3.Because the data sets were too small for statistics (only 3 to 4 data per set), calculation of the standard deviation or other statistical parameters was irrelevant except the average.The efficiency of the coating is quantitatively expressed by the coefficient Cef.It is apparent that the coatings that were only cured for 2 days markedly lost their protective effect; those samples reached lower values of NWAI than the untreated sandstone, but considerably higher values than the longer cured ones.After the test of water absorption by complete immersion, samples were dried and their mass was checked once again.This weighing showed the mass increase of 0.23 g for the samples 2x/2w and 0.22 g for 3x/2w compared to their initial dry mass before the treatment with Antipluviol S (Table 4).This additional mass can be attributed to the hydrophobic coating that resisted the leaching effect of the water.Despite a small loss of the coating's mass, the protection from water is still effective, as reflected by NWAI in Table 3.

Efficiency of the Funcosil SNL
Saturation profiles without and with the Funcosil SNL coating are presented in Figure 8. Raw data from the tests are again in Table S1.The samples from the series 2x/2d reached WAC = 8030 g•m −2 within 48 h and from that moment they behaved like the untreated ones.This shows that the impregnation effect disappeared within 1 day only when the curing time was too short.Sample series 2x/2w and 3x/2w yielded quite similar saturation profiles.Certain increases in saturation rates after 8 h indicate partial loss of the hydrophobic effect.However, at the end of the tests (30 days), WAC equals only 8600 g•m −2 .This is a value that was already noticed in untreated samples after 2 days.Comparing these two sample sets, coatings with 2 weeks of curing brought a decrease of 20% of the WAC found in untreated samples.Results of NWAC are compared to NWAI in Table 5. Funcosil SNL totally lost its protective effect when cured only for 2 days; those samples finally reached similar values of NWAI to the untreated sandstone (the small absorption increase is within the accuracy limits of the method).However, this coating type seems to be less effective than Antipluviol S after 2 weeks of curing time, as fully immersed samples contained a higher amount of water.Already the capillary absorption was much higher, which is visible when comparing Figure 7 and Figure 8.The efficiency of 20% that was reached after 2 weeks of Funcosil SNL curing and observed in the test of water absorption by capillarity disappeared after 30 days under water; the hydrophobic effect was lost and NWAI of fully saturated samples became close to that of untreated samples.The dry masses of both series, 2x/2w and 2x/3w, are also the same or even lower than before treatment.All of this indicates that Funcosil SNL was significantly dissolved and washed out when fully immersed in water for 30 days (Table 6).

Discussion
Sufficient penetration of the hydrophobic medium into the material that should be protected is the key factor of coatings' long-lasting efficiency.It is necessary to find out how much of certain coating types is needed to reach the required penetration depth.With such knowledge, both longevity and cost savings can be secured [17].However, there is still no standard method of how to evaluate the penetration depth [24].The procedure used in this research is the simplest and the most often applied method [6,14,15,24,29].It showed deeper penetration when more coating layers were applied.Similar statements were published indicating that deeper penetration can be achieved either by applying more of the impregnation medium or by increasing its concentration [6,13,15].The producer of Antipluviol S also informs the users that the efficiency of the coating depends on the porosity of the treated material and on the amount of the applied coating medium.He recommends a penetration depth of 4 mm.However, as is evident from Figure 6, this was not reached with either of the recommended two or three layers of coating.This may explain why higher penetration with three layers did not show much better efficiency than with two layers-it was still insufficient and more layers should be applied on this very porous stone.However, deeper penetration does not necessarily mean higher efficiency [14].
The curing time proved to be more important for the coating's efficiency than the number of coating layers (at least in this case study).Unfortunately, no recommendation regarding the curing time is given for the tested products.Two weeks are recommended for another hydrophobic product called Wacker [6].Several authors confirm that longer curing time creates a better hydrophobic effect [6,13,15].In the presented study, 2 weeks were sufficient for the curing of Antipluviol S to protect the sandstone from water infiltration; the water absorption by capillarity of coated samples was circa 10% of that of untreated samples, meaning cca.90% improvement of this stone property.Under same conditions, results with Funcosil SNL were not so satisfying, reaching only 20% improvement.This coating type probably needs more layers applied and/or longer time for curing-further research is necessary.The leaflet provided by the producer states that the surface treated with Funcosil SNL must be protected from water at least for 5 h to secure the optimum impregnation.It is now evident that even 48 h is insufficient.Furthermore, if fully immersed in water, the coating may totally disappear.This must be taken into account in case of floods-a new coating application may be needed after drying the flooded walls to protect them from rain again.However, the flood will probably not take 30 days as the laboratory tests did, which should be regarded.The producer of the coating Funcosil SNL claimed it had a positive impact upon frost resistance, which was not tested in this study.
Based on the laboratory comparison, Antipluviol S can be recommended for the protection of stone facades of buildings, as well as of stone statues and monuments in exteriors in climates with heavy and frequent rains.Before its application, the weather forecast must be taken into account to secure a curing time of 2 weeks without rain.In less humid climates, Funcosil SNL can be useful as protection against occasional and moderate rain.
Of course, producers also work on the research and development of their products.In this way, shortly after our research, the producer of Antipluviol S changed the information leaflet.In the new leaflet, 0.8 kg•m −2 is given as the assessed amount for the application to reach sufficient saturation, whereby no breaks for drying should be left between individual coating layers.In this way, much more chemicals will be consumed.The consumption in the presented study was some 0.1 to 0.3 kg•m −2 , following previous recommendations.Now, not only the economic, but also the environmental impact must be considered.All such coatings are declared as dangerous, highly flammable, and harmful for environment.At the university, further research will be directed towards the minimum consumption necessary for an optimum effect.
Finally, some remarks as to the comparability of the presented research with similar reports [6,[13][14][15]: the testing conditions and treated stones were too different.In the presented study, curing time proved to be the most important factor.While 2 days and 2 weeks were compared here, other authors used more than 2 weeks for the curing, and some of them even tried much longer times.This makes any comparison with published reports impossible.

Conclusions
Two types of hydrophobic coatings, Antipluviol S and Funcosil SNL, were applied on the highly porous Hořice sandstone to test the impact of both the coating's curing time and the number of applied coating layers.Two and three layers were compared.The coating medium penetrated deeper with three layers, but without any significant effect upon the efficiency of the coating compared to two layers.The curing time of 2 days turned out to be insufficient for both coating types.The curing time of 2 weeks was long enough, leading to ca. 90% reduction of the water absorption by capillarity with Antipluviol S. It represents a good protection against water absorption.Reduction of this parameter with Funcosil SNL reached only 20%.Notably, the test of water absorption by capillarity proved to be suitable for a fast first check of the coatings' efficiency.Two days of this test is sufficient for that purpose; it is not necessary to perform longer testing.Though the water content absorbed by capillarity constantly increased (with rates <1% per day) for the next 28 days in all stone samples, the non-resistance of the protective coating could be recognized after 48 h in water.
The higher efficiency of Antipluviol S was also confirmed by the underwater saturation over 30 days.The coating cured for 2 weeks remained in the treated sandstone even after such a long leaching time.This can be seen on the higher dry masses of those samples (series 2x/2w and 3x/2w) compared to initial dry masses, i.e., before the application of Antipluviol S. Identical treatment and testing of samples with Funcosil SNL showed the loss of the already limited protective effect after 30 days of full immersion in water.The total water absorption was not reduced, nor was any dry mass increase observed when compared to untreated samples.Despite 2 weeks of curing, the Funcosil SNL coating must have been dissolved and leached out.This makes its application in humid climates with frequent heavy rains questionable.Resources in the Podhorní Újezd quarry guarantee future production and application of the popular Hořice sandstone as building and decorative stone.Therefore, findings presented in the paper are directly applicable in practice.

Figure 1 .
Figure 1.Two of the many buildings in Bratislava (Slovakia) with Hořice sandstone facades.

Figure 2 .
Figure 2. Color changes of the Hořice sandstone with coatings.(a) Untreated sample, (b) sample with Funcosil SNL coating, and (c) sample with Antipluviol S coating.

Figure 3 .
Figure 3.Estimated contact angle between a water droplet and the sandstone surface coated with Funcosil SNL.

Figure 7 .
Figure 7. WAC curves-samples with Antipluviol S coatings compared to untreated ones.

Table 1 .
Technical data of tested coatings and recommendations for their use.

Table 2 .
Designation of the sample series according to test conditions.

Table 3 .
Efficiency parameters of Antipluviol S.

Table 4 .
The mean dry mass of the coating remaining in the samples after all water absorption tests-Antipluviol S.

Table 6 .
The mean dry mass of the coating remaining in the samples after all water absorption tests-Funcosil SNL.