Laboratory Assessment of Selected Protective Coatings Applied on Two Sandstone Types

Abstract

The efficiency of three commercial hydrophobic coatings applied on two types of sandstones was subjected to the tests of water absorption by capillarity, water absorption by complete immersion, and frost resistance. Two days of the curing time of coatings on samples were chosen for the laboratory tests. The effects of the used coatings were different (ambiguous). Two coatings proved their effectiveness and thus, relevance of their use, but different coatings were effective on each type of sandstone. The coating F showed the best efficiency for the Hořice sandstone and coating A for the Malé Skalky sandstone. The third testing coating H was ineffective on both types of sandstone. The water absorption tests by capillarity showed a loss of the coating effectiveness over time. By completely immersing the samples in water, all three coatings lost their effectiveness and the hydrophobic effects disappeared. Applied protective coatings helped to preserve the frost resistance of the Hořice sandstone when looking at the uniaxial compressive strength after 25 freeze–thaw cycles, regardless of their fading hydrophobic effect. In addition, a penetration depth test for Hořice sandstone was performed. The impregnation depth reflects the effectiveness of the coatings and confirmed the results obtained by the water absorption tests by capillarity. The findings presented in the article are also useful in practice, as Hořice sandstone will be mined and frequently used in the future as well.

1. Introduction

Natural building stone, either used as a bearing construction element or as a decorative stone element, weathers in the same way as in the rock mass it was taken from if exposed to the climate factors. Any decay of culturally significant stone artifacts represents an irretrievable loss of our heritage and history [1]. Water is the most important decay factor. Cyclic temperature changes, including freezing pore water, leading to its volume increase at the transition to the solid phase followed by increasing pore water pressures, accelerate the destruction by cracking and extending the older cracks [2,3]. Different salts dissolved in the penetrating water can be another negative factor. Wet building stone is also susceptible to mildew.

There are several ways water can enter the stone at the construction sites: hygroscopic sorption and condensation of water vapour, capillary uplift, and direct rain, which is the most critical one. If rain is combined with strong wind, mechanical impact is added [3]. Wind can also transport fine solid particles mixed with the rainwater, resulting in surface abrasion or even gradual disintegration, especially in sedimentary stones with weak bonds, such as sandstones and limestones [4]. Water absorption by capillarity (called also sorptivity) enables water penetration into the building material even against the gravitation due to the capillary forces [5]. The resulting amount of water in the porous building material also depends on the temperature, as shown by experiments [6]. The water absorption coefficient rises with increasing temperature for bricks and stones.

To prevent water from penetrating the building stone means preserving it from gradual degradation to extend its service life. This is the primary goal of protective coatings applied on building facades. Hydrophobic impregnation creates liquid-water-impermeable surfaces on the natural building stone and prevents spontaneous water absorption by capillarity, as stated by [7]. Their findings also show that hydrophobic treatment makes stone nearly impermeable to liquid water when evaluating the samples with water absorption tests by capillarity, but it is still permeable to water vapour. Moreover, the treated stone’s impermeability grows after exposure to water. In addition, the water repellent agent appears to spread progressively in the material for a long time after the hydrophobic treatment, yielding high final impregnation depths. These findings confirm that water repellent agents successfully hydrophobize the tested materials with a liquid-water-tight but vapour-open hydrophobic layer. This payer goes deep into the material without notably changing its pore-size distribution. The same is known from the impregnation of shoes or textiles. Wind-driven rain creates water drops remaining shortly on the stone surface and running quickly down the wall. Hydrophobic impregnation coatings change the surface to keep water and aggressive water-soluble salts, such as chlorides and sulfates, out of the substrate [8]. Water is a polar liquid, and building materials have mostly a negative surface charge, attracting the polar water molecules, i.e., they are hydrophilic. Through the electro-chemical compensation of the negative surface charge by the coating, the surfaces become neutral and non-attractive, i.e., hydrophobic for the polar water molecules [7]. The application of hydrophobic coatings is discussed more than any other preventive measures in the restoration of historic stone objects. It is fully rejected by some opponents who mean the hydrophobic effect is irreversible, which can complicate some future treatment. However, research on buildings in Germany has proven that the impregnation does not affect later restoration procedures negatively; corrective plaster and paintings hold well without immediately peeling off after their application. Ref [9] published that the hydrophobic effect of properly applied silanes and siloxanes was evident for 20 years and longer. Ref [10] pointed out the improper protection by synthetic resin and cement mortar applied on a dolomite cornice of a valuable historic building in Spain. As [11] emphasizes, all impregnation products must be tested in the laboratory before their application on historic buildings. Based on their results, silicone resins in aqueous solution represent the best choice for protection treatment of tuff stones, as they showed a good compromise between ecological compatibility, reduction of absorbed water and yellowing of the original substrate and resistance to weathering after UV radiation. The current market offers a huge number of chemical products for hydrophobic impregnation. Ref [7] listed 77 commercial water-repellent agents suitable for masonry, concrete, mortar, and mineral substrates. Most of them, up to 67%, are silicon-based systems suitable for building stones.

In this paper, the results of laboratory pilot research of three commercial impregnation coatings applied on two types of sandstone used as building and decorative stone are presented. The efficiency of the coatings was assessed by the test of the water absorption by capillarity, water absorption by total immersion, and by the frost-resistance test simulating weathering conditions. In the future, this research will continue with a larger set of rocks to allow a better statistical analysis.

2. Materials and Methods

2.1. Materials

For their high porosity and stability in water, sandstones are very suitable materials for observing the water penetration into the building stone and thus, also for the assessment of the efficiency of impregnation coatings. Sandstones are also commonly available in Slovakia and widely used as building and decorative stones.

Two different types of sandstone were selected for this research. Geological origin, colour, and structural characteristics of both sandstones are summarized in

Table 1. Characteristics of the studied sandstones.

Sandstone Type	Hořice (HS)	Malé Skalky (MS)
Quarry location	Podhorní Újezd, Czech Republic	Chtelnica, Slovakia
Geological unit	Bohemian Cretaceous Basin [12]	Dobrá Voda Depression [13]
Age	Mesozoic–Upper Cretaceous [12]	Terciary–Neogene [13]
Lithological type	Quartz sandstone/Arenite	Carbonate sandstone
Colour in fresh state	Very pale brown (10YR8/2) with characteristic darker rusty streaks [14]	Very pale brown (10YR8/2), homogeneous color [14]
Texture	Fine grained	Fine grained
Isotropy/Anisotropy	Isotropic	Isotropic
Figure of sandstone

.

The first, called “Hořice sandstone“(further HS), is a well-known and popular Czech rock type used for ages in construction for interior and exterior decorations. It is the most frequent decorative stone on public buildings, statues, and monuments in Slovakia. Its consumption culminated in 1950–1960 due to the reconstruction after the Second World War [15]. Rock blocks were taken from an active quarry in Podhorní Újezd, the biggest recent sandstone quarry in the Czech Republic. Quartz is the dominant mineral (>90%); other minerals are minor (glauconite, feldspars, mica, heavy minerals).

The second sandstone type was taken from an abandoned Malé Skalky (further MS) quarry in Chtelnica, a location in the Malé Karpaty Mts. in the western Slovakia. It is suitable for similar applications as the Hořice sandstone. It was used as building and decorative stone in the past and can be found on many historic buildings as well as on monuments and graves in western Slovakia. The well-rounded clastic grains consist mostly of carbonates with a very minor content of quartz (ca. 2%) or cherts. The intergranular cement consists of carbonates, less often of a blend of clay and carbonate [16].

As they are stable in water, water absorption was used as a criterion to assess and compare the efficiency of the selected impregnation coatings.

The dimensions of the monolithic blocks taken in both quarries were ca. 30 cm × 30 cm × 20 cm. From them, samples were cut in the form of cubes with 50 mm long edges (Figure 1). Three prismatic samples 50 mm × 25 mm × 25 mm were also cut solely for the first attempt to test the coating’s penetration depth.

2.2. Methods

Three types of commercial impregnation coatings were compared by laboratory tests: coating A, coating F, and coating H. Coating A is a product of the Mapei Company (Milan, Italy). It is based on siloxane resins. Coating F from the Remmers Company (Löningen, Germany) is a low-molecular alkylalkoxysiloxane. Its penetration is excellent due to the low-molecular structure. Inside the rock, it reacts with the pore vapour yielding the hydrophobic polysiloxane. Both coatings are ready-to-use; no dilution is necessary. The third coating, coating H from the Phase Restauro Srl Company (Treviso, Italy), is based on silane monomers and must be diluted to the ratio coating: distilled water = 1:9 before application.

All three coating types are colourless and transparent. The coating is absorbed and not visible on the surface; it does not significantly change colour so it does not form any film on the surface. They are hydrophobic (i.e., water-repellent), liquid (no gel or cream), do not modify the external appearance or the color of the substrate, and at the same time, do not reduce the transpiration of the treated material, as stated by the producers. They belong into the silicon-based agents group. Silicon-based systems are the most popular water repellents in use. Tested coatings are all recommended as suitable for treating natural stone.

The producers recommend two layers of the coating applied by brush on a dry, clean surface of the treated material 2 h apart. This creates an impermeable surface for liquid water. After the coating application on all sides/all surfaces, the sandstone cubes were stored at room temperature for 48 h (coating’s curing time).

The following physical properties of untreated natural sandstones were determined according to [17,18,19,20,21,22]: bulk density ρ_d [g.cm⁻³], particle density ρ_s [g.cm⁻³], total porosity n [%], open porosity n_o [%], water absorption at atmospheric pressure by complete immersion WAI [%] into the water until reaching a constant mass, degree of saturation after total immersion S_r (%), uniaxial compressive strength UCS [MPa], and ultrasound propagation speed V_p [km.s⁻¹].

The efficiency of the coatings was assessed by three types of tests on treated sandstone samples: water absorption by capillarity, water absorption by complete immersion, and frost resistance tests followed by the UCS testing [17]. UCS testing was used to determine uniaxial compressive strength.

The tests of the water absorption by capillarity (WAC) were carried out according to [23,24]. The samples dried to constant mass were immersed in water to a height of 3 ± 1 mm in a plastic tank on non-absorbent and non-oxidizing pads (Figure 2).

Water was sucked into the unsaturated part of the sample by the capillary rise. The time intervals used for measuring the mass increase due to absorbed water were given: 1, 3, 5, 10, 15, 30, 60, 120, 180, 240, 360, 480, 1440, and 2880 min (48 h). The results were expressed through the WAC curves where WAC = f(t). WAC was calculated as the amount of water absorbed by the specimen per unit area at time t_i as follows:

$$WAC=\frac{m_{ti}-m_d}{A} [\mathrm{g}/{\mathrm{m}}^2]$$

(1)

where m_ti [g] is the mass of the sample determined at the time t_i (s) since the first contact with water, m_d [g] is mass of the dried sample, A [m²] is the area of the sample base in contact with water.

The first part of the curve is dominated by the capillary force when the water front has not reached the top of the sample, while the second stage is mainly governed by the diffusing of the entrapped air out of the sample after the water front reached the top of the sample [25].

At the end of the WAC tests, samples were fully immersed into water with the water level 1 cm above the samples and were left there until they reached a constant weight. Under these conditions, water can also fill bigger open pores, not only those with high capillary forces. Constant weight m_sat is a signal of full saturation. After reaching the full saturation, WAI (%) was determined as follows:

$$WAI=\frac{m_{sat}-m_d}{m_d}·100 [\%]$$

(2)

where m_sat [g] is the mass of the full saturated sample, and m_d [g] is the mass of the dried sample.

The natural variability and heterogeneity of stone cannot be avoided and must be considered during the evaluation of the results. Therefore, mathematically normalized values of WAI seem to be a better criterion for the comparison of the impact of different coatings by “bringing the different stone cubes to the same starting open porosity“. Values of WAI were related to the open porosity, n_o, which controls the water absorption by capillarity and saturation degree. Normalised values were calculated as follows:

$$WA{I}_n=WAI:n_o$$

(3)

In the next research step, saturated samples underwent freeze–thaw tests to determine the frost resistance. Saturated samples were taken out from the water and frozen at −12 °C for 4 h. Thawing in water at room temperature followed. After 2 h a new freeze–thaw cycle started, and each sample underwent 25 such cycles. Tests of the cyclic freeze–thaw impact upon the UCS were carried out only on HS, both on untreated and coated samples but with the coating A and coating F only.

The penetration depth of the coating is very important for its efficiency and sustainability. The quantity of the coating agents must be controlled to reach the desired penetration depths, which leads to higher resistance and lower costs [8]. There is no standard method on how to determine the penetration depth of the impregnation. Most authors prefer a simple test [7,26,27] applied in this research. The test of the coating’s penetration depth was done on the prismatic samples of HS as well. Two layers of the coating were applied on the whole dry surface of the prism. After 48 h of curing time, samples were cut into two halves perpendicularly to the coated surface, and the surface of the fresh cut was put in contact with blue-coloured water for 1 s. Ink was used to color the water and to make the coating’s penetration depth better visible. The hydrophobic zone does not suck so much coloured water than the rest of the sample, and it should create a visible light rim around the dark blue centre of the cut. The thickness of this rim was measured by a calliper.

3. Results

3.1. Basic Physical Properties

Though the stability of both sandstones in water is high, they are classified as soft rocks due to their low UCS [28]. High porosity makes them a relatively light building material that is easy to work with (manipulation, forming, etc.), but still stable enough. Different mineral compositions and intergranular cements are responsible for different physical properties, especially for porosity and UCS. Siliceous sandstones, such as HS, show higher UCS and higher durability than calcareous ones (MS).

The basic physical properties of both sandstones in their natural states can be compared in

Table 2. Basic physical properties of studied sandstones—mean values.

Property/Sandstone Type	HS	MS
Bulk density ρd [kg.m−3]	1931	2196
Particle density ρs [kg.m−3]	2612	2733
Total porosity n [%]	26.07	19.65
Open porosity no [%]	18.30	15.82
Water absorption by total immersion WAI [%]	9.48	7.21
Degree of saturation after total immersion Sr [%]	70.21	80.54
Uniaxial compressive strength UCS [MPa]	29.9	20.6
Ultrasonic P-waves velocity Vp [km.s−1]	2.23	2.74

.

3.2. Assessment of Impregnation Coatings

3.2.1. Hořice Sandstone

Deviations of the mean physical properties of untreated samples and of those with selected coatings are negligible before the frost resistance tests (

Table 3. Impact of the freeze–thaw cycling on the properties of untreated samples and samples with selected coatings—HS (n = not tested).

Parameters	Types of Samples	Before Test	After Test
Bulk density ρd [kg.m−3]	untreated	1931	1928
	with coating F	1937	1924
	with coating A	1926	1921
	with coating H	1919	n
Total porosity n [%]	untreated	26.07	26.20
	with coating F	25.86	26.34
	with coating A	26.25	26.46
	with coating H	26.54	n
Open porosity no [%]	untreated	18.30	21.37
	with coating F	17.71	22.04
	with coating A	18.43	21.15
	with coating H	18.02	n
Water absorption by total immersion WAI [%]	untreated	9.48	11.08
	with coating F	9.15	11.46
	with coating A	9.57	11.01
	with coating H	9.47	n
Degree of saturation Sr [%]	untreated	70.21	81.56
	with coating F	68.48	83.67
	with coating A	70.19	79.94
	with coating H	70.15	n
Uniaxial compression strength UCS [MPa]	untreated	29.9	24.9
	with coating F	n	28.3
	with coating A	n	28.8
	with coating H	n	n

), as they only reflect the natural heterogeneity of the sandstone.

The impregnation did not improve the water absorption considerably, which was quite unexpected. Results are more interesting after the freeze–thaw tests (samples with coating H were not involved in these tests). The increased n_o is especially evident from Table 3 in all sample groups, both untreated and impregnated ones. This led to increased WAI and saturation degree, S_r, though the increase of n is very small. The apparent water-repellent effect of coating H (see WAI in Table 3) was in fact only due to the lower n_o of those samples, as shown by normalized values WAI_n (

Table 4. Normalized values of WAI_n before and after freeze–thaw cycling tests—HS.

Sample	WAIn
	Before (A)	After (B)	A:B [-]
untreated	0.5178	0.6056	0.86
with coating F	0.5164	0.6478	0.80
with coating A	0.5191	0.5976	0.87
with coating H	0.5212	n	n

).

After 25 freeze–thaw cycles, UCS of untreated samples dropped from 29.9 MPa to 24.9 MPa, whereas impregnated samples showed values close to the sound untreated sandstone; the frost impact was negligible.

Time-dependent values of WAC can be compared in Figure 3. The first four graphs in Figure 3 show the number of samples tested with one type of coating, i.e., one curve for each sample. There were different numbers of samples in each group. The last graph compares the mean values, i.e., representative curves from those four sample groups, meaning it compares the different coatings.

Untreated samples reached the WAC_max of 8600 g.m⁻² within ca. 1 h. The coating A started to lose its impregnation effect after 8 h of immersion in the water, and at the end of the test, these samples were saturated to WAC_max close to 8000 g.m⁻². After 48 h from the immersion moment, WAC increased only to one half of the value measured for both untreated samples and those with coating A or coating H. The last mentioned showed very low efficiency in this test, and the retardation of the adsorption curve resembles the curve of the untreated samples. As the mean value of WAC_max is slightly below that of the untreated sample, a certain hydrophobic effect is indisputable, but it is insufficient.

In samples with the coating F, the absorption rate is half the value of the untreated ones after two days of permanent contact with the water. It demonstrates that this hydrophobic impregnation is very efficient, and it slows down the water penetration in the stone.

However, coatings, even coating F, were immersed in the water during and after the freeze–thaw cycles. The coatings were dissolved or diluted, and WAI reached similar values in all sample sets, regardless of the treatment (Table 3). Normalized values, WAI_n, of samples with coating F were even higher than those of the unsaturated ones.

The experiment with the blue ink needs further development. However, it is evident from Figure 4 that the expected light rim (zone 1) really occurred near the edges impregnated with coating F. The rim is very fine, a fully hydrophobic zone is only 1–2 mm thin and a partly hydrophobic zone (zone 2) follows, dissipating towards the sample centre. The dark blue parts (zone 3) were not impregnated at all. When looking at the samples coated with coating A, only zones 2 and 3 can be recognized. No penetration zone was observed on the samples with the coating H.

3.2.2. Malé Skalky Sandstone

Table 5. Impact of the freeze–thaw cycling on the properties of untreated samples and samples with selected coatings–MS.

Parameters	Types of Samples	Before Test	After Test
Bulk density ρd [kg.m−3]	untreated	2196	2190
	with coating F	2192	2189
	with coating A	2162	2161
	with coating H	2187	2191
Total porosity n [%]	untreated	19.65	19.86
	with coating F	19.79	19.90
	with coating A	20.88	20.91
	with coating H	20.00	19.83
Open porosity no [%]	untreated	15.82	17.19
	with coating F	15.76	15.47
	with coating A	17.16	16.92
	with coating H	15.73	16.32
Water absorption by total immersion WAI [%]	untreated	7.21	7.85
	with coating F	7.18	7.07
	with coating A	7.92	7.83
	with coating H	7.19	7.45
Degree of saturation Sr [%]	untreated	80.54	86.59
	with coating F	79.35	77.76
	with coating A	81.98	80.96
	with coating H	78.60	82.29

summarizes the physical properties of the untreated MS and the values after the treatment by three preventive coatings. The right part of the table shows how they changed after 25 freeze–thaw cycles.

Here, the long-term resistance indicated by WAI after the freeze–thaw cycling increased by the following rank: coating A < coating H < coating F. All give slightly lower values than the untreated samples. However, the rank changed with WAI_n to coating H < coating A < coating F, pointing out the impact of n_o and showing what would be the rank if all samples had the same n_o at the beginning (

Table 6. Comparison of the mean normalized values of WAI_n before and after freeze–thaw cycling tests—MS.

Sample	WAIn
	Before (A)	After (B)	A:B [-]
untreated	0.4554	0.4961	0.92
with coating F	0.4551	0.4495	1.01
with coating A	0.4615	0.4569	1.01
with coating H	0.4569	0.4733	0.97

).

Curves in Figure 5 represent WAC increasing with the time. The explanation for the lines in the graphs is the same as for Figure 3. The untreated samples reached their WAC_max (6800 g.m⁻²) very quickly, within ca. 30 min from the test start. Samples with both coating F and coating H behaved quite identically. The small difference arose from the different number of tested samples where two samples with coating F of four in that group were as nearly identical with both curves of the coating H sample. Their hydrophobic effect was extremely short, and the time delay is only 1.5 h compared to the time of achieving WAC_max by untreated samples. However, the value of WAC_max is slightly lower with these coatings, reaching 6775 g.m⁻². The set of three samples with coating A is the most different one. The first sample copies the behaviour of the samples with other coatings with a slightly postponed saturation start, while the other two samples are more and more retarded and evidently not fully saturated even after 48 h. The mean value of WAC_max calculated from these three tests reached 6990 g.m⁻² in samples with coating A, and this level of saturation was not achieved before 24 h from the test start.

4. Discussion

The different behaviours of water drops on an untreated surface versus an impregnated one is manifested in Figure 6. Here, the coating A was applied on HS making its surface water repellent. Water drops remain on the coated sample surface and do not infiltrate into the stone. However, this simple experiment does not tell anything about the durability of this effect. It can neither tell which type of coatings is better and more efficient. This phenomenon is also described in the work of [29], which deals with Janus membranes. Janus wettability membranes feature asymmetric surface wettability.

Determination of WAC is the best method to test the immediate efficiency of the hydrophobization treatment [30,31]. Its popularity is rising. [32,33] recommend using it more often for the evaluation of the natural building stones in the practice. Applied on HS cubes, the efficiency of the coatings was increasing in the rank coating, H < coating A < coating F, according to WAC. The best of them, coating F, was resistant to water at least for 30 min before the water infiltration started. This is quite a long time as the structures are not considered to be permanently covered with water during rain. The test with the ink illustrated that coating F also yielded the deepest infiltration, i.e., the thickest water-repellent layer within the stone which can be the reason for the best WAC. This is in accordance with the statement in the information leaflet of coating F that good absorption of the impregnation is necessary for the optimum efficiency. And, of course, the water-repellent effect may lead to lower values of the measured n_o and n when tested by weighing under water. However, these results must be interpreted very carefully because samples must be fully saturated with water for the underwater weighing, which takes much longer than the times of efficient function of the coatings shown by WAC curves. Within that time, the coating may be more or less diluted or dissolved.

The untreated sandstone MS behaved as expected. Due to the freeze–thaw cycling, the total porosity, n, slightly increased, and the open porosity n_o was even more evident. This caused an increase of the sample volume and a decrease of the dry bulk density, ρ_d. The opening of the pores led to higher WAI and S_r of the samples after the frost resistance test.

Higher WAI_n indicates the less efficient long-term performance of the coating considering an under permanent water level, i.e., under extreme conditions not typical for houses. The long-time efficiency/quality of the coating on HS increased as follows, coating H < coating A < coating F. However, only coating F showed a real long-time hydrophobic effect before the freeze–thaw tests as WAI_n of the samples with the other two coatings was even higher than that of the untreated sample. Determination of WAI_n was also used to assess the impact of the different impregnation coatings upon the frost resistance of the sandstone. Here, the situation changed. While all samples showed higher WAI_n, only coating A caused slighly lower values than the untreated samples. Coating F lost its hydrophobic effect during this relatively long test. Coating H was not tested. According to WAI_n (Table 6), the only efficient coating before the freeze–thaw cycling applied on MS might be the coating F. Coating A and coating H yielded higher normalized values than the untreated sample, though coating H caused a slightly lower absolute mean value of WAI than those measured in the untreated samples. In this case, the significance of the normalized values is outlined. However, after the frost resistance test, all coated samples were better than the untreated ones. The frost resistance efficiency decreased in the rank coating, F > coating A > coating H, whereby the last mentioned showed higher values after the test (ratio before:after < 1), but was still not as high as the untreated samples. Unlike HS, samples of MS did not confirm that tested hydrophobic coatings reasonably reduce WAC.

Though the hydrophobic effect of coatings was most evident at the saturation start and slowly faded during tests of frost resistance and WAI, the drop of UCS after the freeze–thaw cycles was negligible compared to untreated samples of HS. The strengthening effect of both coating F and coating A was the same. It is in good compliance with findings of [34] where coated tuff samples withstood all 25 standard freeze–thaw cycles, while untreated samples disintegrated only after 17 cycles. The presence of the coating in the stone pores controls the amount of both absorbed and free water in the pores and thus, the destructive ice-crystallization pressures. Thus, the preventive impact of the impregnation is best evident after simulation of winter seasons; increased frost resistance due to coatings was indicated. There was no study on if any chemical bonds between the coating and stone could contribute to the increase of the frost resistance. Based on test results, coating F can be recommended for impregnation of stone facades of public buildings, sculptures, and monuments made of HS. A hardening time of 48 h seems to be a sufficient, but hardening must not be disturbed by rain.

Hydrophobization serves as protection of those parts of buildings which are often exposed to heavy rains. If this is not that case, treatment is not necessary. Facades protected from rain, e.g., by other buildings, do not need it [9]. The amount of the applied coating agent and the porosity of the material determine the penetration depth of the impregnation. Good absorption and sufficient hardening time are crucial for the efficiency and longevity of the coating. None of the producers give strict advice about the necessary hardening time. The producer of coating F recommends 5 h of hardening with rain disclosure to reach the optimum impregnation. The experience of the paper authors proved this is too short. On the other hand, [7] reported two weeks are necessary for hardening. In practice, it is impossible to control the occurrence of rain, and such a long dry period is a question of luck. Two days of hardening were chosen for the laboratory tests as a compromise. However, deducing from the high-water absorption by the impregnated MS, even 48 h of hardening is insufficient. Though laboratory samples can be much better impregnated than real buildings and laboratory-based applications produce generally better, more reproducible results than field-based studies [35], laboratory experiments with two days of hardening did not show high efficiency.

This research pointed out that different stones may interact with different hydrophobization coatings in different ways. It does not mean that some of the coatings are not recommended at all. Perhaps more than two layers are necessary to reach the desired efficiency. For example, the producer of coating A recommends applying 0.1 to 0.3 kg.m⁻² on the surface of a natural stone, which represents one to five sheets. Only two sheets were tested here. More sheets may bring better results for all coatings. Coating F and coating A can be recommended only as protection against occasional wetting by moderate rain and fog. On the other hand, the efficiency of the coating H is critically low and insufficient.

There are many different state of the art hydrophobic products in this field that are available in the market that can be applied to natural stones and are used in research reports under very different test conditions and with different materials for the treatment [3,7,36,37]. As curing times are different, the results are very difficult to compare to published papers due to the number of coats as well as the diversity of stones. Our current research shows that the curing times of the coatings are the main impact factor of the coatings’ effectiveness. In the presented research, a curing time of two days was chosen to find out how the coating behaves after such a short time and if it is sufficient. Thus, the results cannot be compared to works where authors used curing times of two weeks or longer.

5. Conclusions

The efficiency of three commercial hydrophobization/impregnation coatings applied on two sandstone types was tested. Both were commonly used as building stones or materials for sculptures either in the Czech Republic or Slovak Republic. Thus, the findings presented in the article are useful for the practice, as it is assumed that Hořice sandstone will be mined and frequently used in the future as well. The findings of this pilot research are necessary to optimize more extensive research in the future. It should be conducted on more samples in particular sets but also on other types of stones because the structures and compositions of the rocks also have significant influence on the test results. The presented data are applicable only for similar sandstones under similar test conditions.

According to the presented results, the determination of WAC is the most reliable value for the coatings assessment because tests of WAC resemble the situation of real buildings exposed to rain. WAI_n is of secondary relevance, but its application in the analysis was a novel attempt to minimize the impact of the different initial porosities of the samples.

The partly different interactions of the two sandstones with these two coatings may be caused by their different mineral/chemical compositions, which must be next studied. However, both coating F and coating A helped to preserve the frost resistance of the Hořice sandstone when looking at the uniaxial compressive strength, UCS, after 25 freeze–thaw cycles, regardless of their fading hydrophobic effect. This is a very interesting issue for further research. Based on our results, a targeted and responsive expert assessment is necessary in every single application, especially on objects of cultural heritage. A total prevention of the degradation processes is impossible, but they can be slowed down by suitable protective coatings.