Influence of surface finish and composition on the deterioration of building stones exposed to acid atmospheres

Construction and Building Materials 106 (2016) 392–403

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Influence of surface finish and composition on the deterioration of building stones exposed to acid atmospheres Patricia Vazquez a,b,⇑, Lucia Carrizo b, Celine Thomachot-Schneider a, Soizic Gibeaux a, Francisco Javier Alonso b a b

GEGENAA EA3795, Université Reims-Champagne-Ardenne, CREA, 2 esplanade Roland Garros, 51100 Reims, France Dpto de Geologia, Universidad de Oviedo, Jesus Arias de Velasco s/n., 33005 Oviedo, Spain

h i g h l i g h t s  Acid atmospheres may cause damage to all types of stones.  In calcitic stones, the effect of acid atmospheres is related to porosity.  Intensity of colour change depends mainly on the inherent stone colour.  Artificial finish influences the acid atmosphere’s attack.  HNO3/H2SO3 ratio controls whether nitrogen compounds react with the stone or acts only as a catalyser.

a r t i c l e

i n f o

Article history: Received 15 July 2015 Received in revised form 21 November 2015 Accepted 17 December 2015

Keywords: Acid atmosphere Pollutants Stone Colour Roughness Sulphates Nitrates Artificial finish

a b s t r a c t Six stone types with differences in composition and texture were exposed to four strong acid atmospheres formed from different acids: H2SO3, HNO3, and two mixed solution with different proportions of H2SO3 and HNO3. The changes on the surface were assessed by weight, colour, roughness and microscopic observation. Exposure to the atmosphere formed by HNO3 hardly affected the stone, whereas these formed from H2SO3 produced evident alterations. Depending on the HNO3/H2SO3 ratio, the nitrogen compounds may react with the stone and precipitate nitrates or nitrites or may only act as a catalyser of SO2 and enhance the formation of gypsum. Colour and roughness are efficient non-destructive approaches to evaluating the damage produced by acid atmospheres. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Weathering of natural building stones exposed to the environment is a field of study in constant development. The main reason is that the agents of decay, climate and pollution are changing due to technical progress and political choices. One of the remediation procedures for global warming involves the modification of combustibles, producing a different air pollution that leads to different stone weathering processes [1]. In the last century, SO2 concentrations in the atmosphere have decreased, and thus, the proportion of nitrogen oxides and volatile organic compounds increased in ⇑ Corresponding author at: GEGENAA EA3795, Université Reims-ChampagneArdenne, CREA, 2 esplanade Roland Garros, 51100 Reims, France. E-mail address: [email protected] (P. Vazquez). http://dx.doi.org/10.1016/j.conbuildmat.2015.12.125 0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.

the atmosphere. This also led to an increase in the ozone production in urban air and consequently to oxidation products such as nitrogen dioxide or nitric acid [2]. SO2 has been the main pollutant involved in the decay of building stones, giving rise to crust formation, yellowing and carbonate dissolution [1,3–7]. In recent times, there has been a strong decrease from high levels of sulphate deposition. Currently, the deposits are dominated by diesel soot and nitrogen compounds that may act as catalysts for the SO2 oxidation [8]. Stone decay generated by mixed gases has been demonstrated to be faster than decay induced by a single gas [9–11]. Stones exposed to a mixture of NO2, SO2, O3 and H2O show higher degradation rates than those exposed to SO2 and NO2 individually [8,12]. The role of metal oxides in the oxidation reactions is, however, under discussion [8–11,13]. On a dry surface exposed solely to

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

SO2, the gas phase oxidation from SO2 to SO3 is slow, whereas for a surface exposed only to NOx, the conversion from NO2 to NO3 is fast. On a dry surface exposed to the mixed gases SO2 + NO2, the oxidation of SO2 to SO24 is 10 times greater than that for exposure to only SO2, whereas the oxidation of NO2 to NO3 is 0.4 times less than that of exposure to single NO2. This confirms that the reaction with stone is dominated by the oxidation of SO2 compared with the oxidation of NO2. However, NO2 can enhance SO2 absorption without producing nitrates, acting as catalysts and increasing the decay produced by SO2 pollutants [14]. Nitrogen oxides have been mainly studied from the point of view of their action as SO2 oxidants. The interaction between NOx and the stone surface is less known [15,16]. The effect of pollutants on stone decay is related to its nature and the inherent characteristics of the stones such as composition, texture or porosity [8,11,17–20]. Moreover, artificial features such as surface finish lead to different surface topographies and consequently to different decay characteristics [17,21,22]). There are also differences between the SO2 absorption in calcite and dolomite [16,23,24]. Calcareous stone reacts with acid pollutants via dissolution. Sulphur and nitrogen salts crystallise on the surface [17,21,23]. Sandstones with carbonate as fragments, cement or clays are also vulnerable to acid attack. The texture of the stone, especially the grain boundaries, may enhance the crystallisation process [24] or the weight loss when exposed to acid rain [25]. Sandstones are vulnerable to air pollutant attack due to their porosity, even if they have low or negligible carbonate contents. Open porosity allows the airborne particles and gazes (SO2, O3, NOx) to enter within the stone and interact with its components. Pore distribution and specific surface area influence the uptake of moisture from the air. High specific surface area implies high moisture content, which will favour dry deposition. Dry deposition could be important in the case of NO2 deposition, as NO2 is less soluble than SO2 [26]. Acid atmospheres produce chemical and physical changes on the stone surface, potentially reaching depths of a few millimetres. As a result, surface properties such as colour and roughness will provide precious information about the degree of the decay. Colour is one of the most representative features of an ornamental stone. Consequently, the colour variations produced by chemical reactions during pollutant exposure have a priority over the evaluation of the decay [e.g., 27–29]. The contaminants studied to date produce yellowing or blackening of the stone. However, the effect of the new proportion of pollutants is less well known, and consequently, so is the colour change. Another parameter closely related to surface decay is roughness. Surface roughness varies due to erosion, dissolution and/or precipitation, providing information about the intensity and the types of decay [30–34], and this is sometimes also related to colour [35]. In addition, surface roughness increases the exposed surface area, trapping more gas and particles [22], and the deterioration process may be enhanced by the increase in surface roughness [17,21]. Ageing tests are used in research to simulate stone decay. These tests do not allow a real comparison between the decay produced in the laboratory and in the real environment. Nevertheless, they are extremely helpful for comparing the durability between different stones and to determine which stone characteristics influence the type and degree of decay, as well as which types of interactions between the stone and the weathering agent take place. In relation to the study of stone decay by pollutants, the standard UNE-EN 13919 [36] is an optimal test for research aims. This test produces an intensive acid attack on the stone, allowing observation in the short-term of the precipitation of salts [28,34] and the evolution of stone properties [37]. The general aim of this research is to evaluate the impact of mixed acid atmospheres on the surfaces of different stones by

393

colour and roughness measurements. Six porous stones with different compositions (three limestones, two sandstones and one dolostone), commonly used as ornamental stones, were tested. The samples were prepared with two different finishes, one smooth (polished) and another rough (bush hammered). The strong atmospheres were obtained from single H2SO3, single HNO3, and two mixed atmospheres with different proportions of H2SO3 and HNO3. This study is intended to shed light on (i) the impact produced by new mixes of pollutants, (ii) the effect of different acid atmospheres on stones with different mineralogies and textures, and (iii) the influence of surface finish on the deposition of pollutants on the surface.

2. Materials and methods Six stones were selected in relation to their composition and texture. Three of them were pure calcareous stones, showing differences in their textures: Albox travertine (AT), Fraga limestone (FL) and Santa Pudia limestone (SPL). The other three stones, which exhibit differences in composition, were Boñar dolostone (BD), Uncastillo sandstone (US), and Villaviciosa sandstone (VS). For future explanations in terms of composition, the three limestones were named ‘‘pure calcitic stones”, and the group formed by the dolostone and the two sandstones was called ‘‘other sedimentary stones”. Two surface finishes were selected in this study: polished and bush hammered finish.

2.1. Methods for the characterisation The characterisation consisted of a petrographic description of the materials and a surface evaluation by means of colour and roughness measurements. Mineralogy and texture were studied using polarised optical microscopy (Zeiss Jenapol POM). Pore system parameters were obtained using a Hg intrusion porosimeter Micromeritics AutoPore III 9410, which reaches 414 MPa pressure and can measure pore radius sizes from 0.003 to 360 lm. All of the data were extracted from previous studies [38,39]. Colour was measured and quantified with a MINOLTA CR-200 colourimeter using the illuminant D65, a beam of diffuse light of 8-mm diameter, and a 0° viewing angle geometry, with a specular component included and the spectral response closely matching the CIE (1997) standard observer curves. Measurements were expressed following the CIE L⁄ a⁄ b⁄ systems (EN ISO 105-J03: 1997) [40]. L⁄ is the lightness that goes from black (value 0) to white (value 100), a⁄ goes from red (values up to +60) to green (values up to 60), b⁄ goes from yellow (values up to +60) to blue (values up to 60). DE⁄ is introduced as the total colour change, to compare the variations before and after the tests as follows: DE⁄ = [(DL⁄)2 + (Da⁄)2 + (Db⁄)2]1/2. The determination of the colour in a heterogeneous material requires a previous study of the number of measurements needed, which is related to the colour of each mineral, grain size and heterogeneity [33,34]. The number of data points was determined by calculation of the cumulative average until the stabilisation of the values. In this study, 15 data points were the minimum required in each stone group (i.e., 5 measurements  3 slabs). Surface topography and roughness were measured using a stereomicroscope Leica MZ16A and the associated software Leica Stereo Explorer 2.1. The resolution is 840 Lp/mm, which provides enough accuracy to detect small surface variations. The maximum vision diameter is 57.5 mm, so a representative surface of the stone can be measured. This equipment uses visible light, which attenuates the irregular reflections and refractions of lasers in the presence of quartz. Prior studies determined an optimal area of measurement of 25  18 mm. Area roughness parameters were computed from an average of three areas in which twenty 2D profiles, with a spacing of 1.25 mm were measured, following the standard EN 4287 [41]. A cut-off length Gaussian filter was applied to evaluate only the roughness wavelengths. This cutoff allows focusing the analysis on the deterioration and not on the slab shape. The following parameters were selected to define surface roughness. – Ra: Arithmetical mean deviation. Arithmetical mean of the absolute values of the deviations (Zi) from the mean line. – Rp: Maximum peak height. The maximum value of the deviations (Zi) from the main line. – Rv: Maximum valley depth. The absolute value of the minimum value of the deviations (Zi) from the mean line. – Bearing area curve. Cumulative height distribution curve from the mean profile. The curve is plotted with a normalised abscissa from 0 to 100% of the profile length. The ordinate indicates the percentage of area analysed from the highest peak to the deepest valley. A parameter that takes into account the whole surface, without the application of the filter, was also obtained:

394

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

– D: Fractal dimension. A fractal describing a rough geometric shape that can be subdivided into parts, each of which, at least approximately, is a reducedsized copy of the whole. They are generally self-similar and independent of scale [42]. Fractal dimension results are a good indicator of surface weathering [31].

2.2. Material characterisation The selected stones with both finishes are shown in Fig. 1. Mineralogy is compiled in Table 1. For more information, see [38,39]. Table 2 shows the chemical compositions of the six stone types expressed in percentages [39]. Porous system parameters can be found in Table 3 [38,39]. Albox Travertine (AT) is a heterogranular limestone, of type travertine, with a yellow-brownish colour. It has a crystalline texture with bands of different colours due to variations in composition or texture. AT has grains of distinct shapes and sizes (micrite, sparite, fibrous), which appear either elongated in bands or filling pores. There are also areas with dark crystals due to plant precursors and Mg and Fe oxides/hydroxides. Very fine stratification surfaces with silt-sized quartz crystals and muscovite can be observed. The insoluble residue is only 1%, and it is composed mainly of quartz and some clay (illite, paragonite). AT has approximately 15% porosity, mainly as elongated and tubular pores following the bedding, and low specific surface area. This stone has pores of large size that cannot be detected by Hg porosimetry with values of Hg porosity about 8%. Boñar Dolostone (BD) is a crystalline stone with a brownish colour. It has a crystalline texture above a granular relict texture. Its mineral composition is formed by hypidiomorphic dolomite fine crystals that replace the original calcite, and the intergranular spaces may be filled with secondary calcite crystals or may remain empty. The dolomite/calcite ratio is 80/20. The 4% of insoluble residue consists mainly of quartz and clays (illite and chlorite/kaolinite). BD has approximately 10% porosity, with larger pores in the intergranular spaces, many small pores (<0.1 lm) and notable specific surface area. Fraga Limestone (FL) is a micrite to biomicrite limestone showing bioturbation and having a grey-yellowish colour. It has a microcrystalline texture due to the transformation of carbonate mud. FL is mainly composed of microcrystalline calcite (1 lm) with bioclasts (10%) such as ostracods, bivalves and charophytes. The insoluble residue is approximately 3%, and it is composed of quartz, clays (illite and chlorite/kaolinite) and organic matter. FL has approximately 30% porosity with moldic pores, is often dissolution recrystallization, and has many small pores (<0.1 lm) and high specific surface area. Santa Pudia (SPL) is a bioclastic limestone that varies from white to yellow colour. This stone has a clast-supported texture with more than 95% fossil fragments (up to 5 mm). Bioclasts are mainly bryozoans, red algae, crinoids (sometimes altered to peloids), and occasional serpulids, echinoderms and mollusc shells. Most of the grains are formed by micrite, and microsparite can be found in the bryozoans and fibrous calcite in the molluscs. SPL shows syntaxial calcite and microsparite cement in low proportions. The insoluble residue is approximately 3%, and it reveals a higher proportion of silt-sized quartz than clays (illite and chlorite/kaolinite). SPL has intragranular and moldic porosity of approximately 35%, fewer small pores (<0.1 lm) and low specific surface area. Uncastillo Sandstone (US) is a litharenite rich in carbonate fragments (calcareous sandstone) with brown-yellowish colour. It has a grain-supported texture with medium to fine grains in contact, and syntaxial calcite cement. This stone has similar proportions of silicate (quartz and feldspar) and carbonate (peloids and bioclasts) minerals. Micrite and clays can also be found in the calcite cement. US has the highest clay content (illite and chlorite/kaolinite) of the studied stones and the highest iron content due to the presence of biotite, chlorite and goethite. US has approximately 20% intergranular porosity, and notable specific surface area.

Table 1 Mineral compositions of the six types of stone. Stone

Cal (%)

Dol (%)

Q (%)

Fp (%)

RF (%)

C (%)

AT BD FL SPL US VS

99 18 97 97 46 –

– 78 – – – –

0.7 2 1 2 15 68

– – – – 10 15

– – – – 25 15

0.3 2 2 1 4 2

AT: Albox travertine, BD: Boñar dolostone, FL: Fraga limestone, SPL: Santa Pudia limestone, US: Uncastillo sandstone, VS: Villaviciosa sandstone. Cal = calcite, Dol = dolomite, Q = quartz, Fp = feldspars, RF = rock fragments, C = clays [39].

Table 2 Chemical composition of the six types of stone expressed in percent. Stone

SiO2

Al2O3

Fe2O3

MgO

CaO

LOI

AT BD FL SPL US VS

0.00 1.03 1.80 2.93 47.80 94.74

0.00 0.49 0.33 0.06 3.43 2.23

0.05 0.38 0.54 0.37 1.80 0.29

0.31 17.63 0.57 0.31 0.59 0.08

55.33 34.34 53.28 52.53 24.73 0.03

43.46 45.48 43.05 43.20 20.40 0.46

AT: Albox travertine, BD: Boñar dolostone, FL: Fraga limestone, SPL: Santa Pudia limestone, US: Uncastillo sandstone, VS: Villaviciosa sandstone LOI: Loss on ignition [39].

Table 3 Porous system distribution obtained by mercury intrusion porosimetry. Specific surface area (SSA) was obtained from N2 adsorption. Stone

po (%)

>1 lm

1– 0.1 lm

0.1– 0.01 lm

<0.01 lm

SSA (m2/ g)

AT BD FL SPL US VS

7.9 ± 1.5 9.1 ± 2.9 29.1 ± 0.2 33.5 ± 1.0 18.0 ± 1.4 19.2 ± 0.6

2.1 0.6 4.3 18.5 14.0 15.6

1.7 5.0 21.6 12.0 2.3 1.8

2.4 3.1 3.0 1.8 1.5 0.9

1.6 0.4 0.1 1.1 0.2 0.9

0.68 2.41 3.82 0.87 2.65 3.76

AT: Albox travertine, BD: Boñar dolostone, FL: Fraga limestone, SPL: Santa Pudia limestone, US: Uncastillo sandstone, VS: Villaviciosa sandstone. po = open porosity, SSA = specific surface area [38].

Villaviciosa Sandstone (VS) is a sublitharenite to subarkose stone with a grey to brownish colour. It has a grain-supported texture with medium-sized grains. Their mineralogy consists predominantly of quartz, microcline type feldspars with different degrees of weathering, and rock fragments (silex, lutite). Overgrowth cement is mainly quartzitic with a low proportion of clay (illite). VS has approximately 20% intergranular porosity and a high specific surface area.

Fig. 1. Macroscopic aspect of the selected stones with both bush hammered and polished finishes.

395

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403 2.2.1. Colour Table 4 shows the colour parameters for each stone type and for both surface finishes, expressed as the average and standard deviation of these 15 measurements. All of the stones were quite homogeneous in terms of colour. Regarding the L⁄ parameter, all of the stones exhibited high values according to their light colour. Within the bush hammered stones, L⁄ varied approximately from 65 (US) to 85 (SPL), showing AT, FL and VS values near L⁄ = 70. For polished stones, L⁄ did not change from the bush hammered surfaces of FL and SPL, whereas the other stones had L⁄ values of approximately 5 units lower. The measurements of the a⁄ parameter yielded results from 0 to 1 for both finishes because the selected stones have mainly white, cream and brownish colours. The only exception was AT, which showed a yellow colour and a⁄ values of approximately 6 in bush hammered and 8 in polished finish specimens. Values of the b⁄ parameter had the highest differences between stones, with values between 10 and 20 for both finishes. AT is again the exception, with b⁄ = 28 for bush hammered and 33 for polished finishes. The total colour variation DE⁄ between bush hammered and polished finishes is higher than 3 (visible threshold [34,39]) in most of the stones, with maximum values of DE⁄ = 7.5 in AT. 2.2.2. Roughness Table 5 shows the averaged roughness parameters and their standard deviations. With the bush hammered finish, most of the stones exhibited similar values, with average roughness (Ra) values between 20 and 25 lm, and maxima peaks heights (Rp) and valleys (Rv) of approximately 60 lm. Some exceptions could be found in the low values of Rv in FL or in the high values in Rp of US. VS had the lowest values among the studied stones. The bush hammered finish produced a regular roughness pattern on the stone surface, which is shown in Fig. 1. This pattern explained the high roughness values found for this finish. In the polished finish, most of the stones showed Ra values of approximately 5–10 lm and Rp and Rv values of approximately 15–25 lm. Because of their coarse grain size, SPL and US exhibited similar values in both polished and bush hammered finishes. Fractal dimension (D) values differed positively from zero. If a value is equal to zero, the surface is completely flat. The higher the values are, the greater the roughness is. Thus, fractal dimension is always lower for a polished finish due to the flatter surface. In this study, the differences in the fractal dimension between bush hammered and polished stones were more significant than for the other parameters. Fig. 2 shows the difference in the bearing area curve between bush hammered and polished stones. These graphs represent the cumulative height distribution curve from the mean profile. Fig. 2 indicates that the polished finish produced flatter surfaces with small differences in height, whereas bush hammered surfaces showed a higher variation in height. No differences were found between finishes in AT and SPL due to the large pores that interfere with the surface data. 2.3. Experimental setup of the ageing tests Four tests were carried out following the methodology described in the standard UNE-EN 13919:2003 [36], but with some modifications. For each test, the specimens were six 4  4  2 cm slabs of each stone type (three of them with polished finish and the other three with bush hammered finish). Prior to the tests, the slabs were immersed in water at atmospheric pressure for 48 h. Each stone type was placed in a closed container on a methacrylate spot over an acid solution film, so that the surfaces were not in direct contact with the solution. The slabs were exposed to the acid environment for 3 weeks. These atmospheres were obtained by adding 50 ml of acid solution to a container with 5 l capacity to create a concentration intermediate between those of the two solutions as indicated in the standard [36]. The standard was also conceived by using H2SO3 + H2O to obtain a SO2-rich atmosphere. In this study, we chose single acidic (H2SO3 and HNO3) and also mixed atmospheres. The nitric acid was Panreac 65% w/w M = 14.4 g mol 1, and the sulphurous acid was Panreac 6% w/w M = 6 g mol 1. The mixtures were calculated proportionally by volume, as follows: – Test 1: Acid solution: 350 ± 10 ml HNO3 + 150 ± 10 ml H2O. Liquid phase: [HNO3] = 10.1 mol l 1. Ratio HNO3/H2SO3 = 100% v/v HNO3. – Test 2: Acid solution: 245 ± 10 ml HNO3 + 105 ± 10 ml H2SO3 + 150 ± 10 ml H2O. Liquid phase: [HNO3] = 7.1 mol l 1, [H2SO3] = 1.6 10 1 mol l 1. Ratio HNO3/ H2SO3 = 70% v/v HNO3 + 30% v/v H2SO3.

– Test 3: Acid solution: 105 ± 10 ml HNO3 + 245 ± 10 ml H2SO3 + 150 ± 10 ml H2O. Liquid phase: [HNO3] = 3.0 mol l 1, [H2SO3] = 3.7 10 1 mol l 1. Ratio HNO3/ H2SO3 = 30% v/v HNO3 + 70% v/v H2SO3. – Test 4: Acid solution: 350 ± 10 ml H2SO3 + 150 ± 10 ml H2O. Liquid phase: [H2SO3] = 5.3 10 1 mol l 1. Ratio HNO3/H2SO3 = 100% v/v H2SO3. From now on they will be referred to in the text in terms of acid volume proportion. Thus, Test 1 is referred as HNO3 test, Test 2 as HNO3 > H2SO3 test, Test 3 as HNO3 < H2SO3 test and Test 4 as H2SO3 test. Visual changes were observed with a Leica MZ-16A stereoscopic microscope (SM) and a JEOL-6100 scanning electron microscope (SEM). RUNSALT simulations were carried out to compare with the chemical compositions of the formed deposits [43]. RUNSALT is a graphical user interface to the ECOS thermodynamic model for the prediction of the behaviour of salt mixtures under changing climate conditions. Weight was measured before and after each test with a precision balance (Metler Toledo Excellence with an accuracy 0.01 g) after 48 h of drying in an oven at 60 °C. Roughness and colour parameters that were used to characterise the materials before the test, were also measured after weight measurement, following the same procedure as for the characterisation.

3. Results 3.1. Macroscopic evaluation The evolution of the samples with the tests is shown in Fig. 3. Visually, no evident changes were observed in samples tested only with nitric acid (HNO3 test). After the HNO3 > H2SO3 test, a yellowing was observed on BD and FL. SPL with the bush hammered finish showed white spots on the surface that may correspond to salt precipitation. In the HNO3 < H2SO3 test, AT exhibited a whitening that could be uniform as in the bush hammered sample or as white spots on the surface in the polished sample. BD with the polished finish also alters to lighter colours compared with the unaltered stone. FL changed to dark brown colours, with black areas on both surface finishes. SPL and the studied sandstones generally exhibited yellowing with a satin finish, and black spots were also observed. In this test, the black areas were better defined than in the HNO3 > H2SO3 test (Fig. 3). The H2SO3 test yielded the highest colour variation. Yellowing was evident in all of the stones, and gave FL a dark colour. Evident white spots that corresponded to gypsum crystallisation were visible in all of the stones. 3.2. Weight variation Fig. 4 shows the weight variation before and after the tests. In general, weight increased slightly. The results showing a loss of weight are not significant statistically and they can be considered as negligeable. Maximum weight gains were approximately 0.2% in the HNO3 test, 0.38% in the HNO3 > H2SO3 test, 0.2% in the HNO3 < H2SO3 test and 0.45% in the H2SO3 test. The results from the HNO3 > H2SO3 test showed the most noticeable increases for all of the stones except for VS (sandstone without carbonates), in which the H2SO3 test produced the greatest change, which is well differentiated from the other tests. The behaviour of polished and bush hammered stones was comparable. Only the HNO3 test showed differences between finishes for almost all of the stones, with slightly higher values in the polished slabs. In this test, SPL got negligeable variations for both finishes.

Table 4 Colour parameters of the sound stones (D65 illuminant). Stone

Bush hammered ⁄

AT BD FL SPL US VS

⁄

L

a

72.0 ± 2.9 76.2 ± 1.1 71.3 ± 6.9 86.1 ± 1.2 64.7 ± 0.9 69.9 ± 0.8

5.8 ± 1.6 0.67 ± 0.4 1.2 ± 0.4 0.06 ± 0.4 0.7 ± 0.4 0.4 ± 0.3

DE ⁄

Polished ⁄

⁄

⁄

⁄

b

L

a

b

27.8 ± 3.9 15.3 ± 0.2 13.0 ± 3.5 12.7 ± 1.0 20.9 ± 1.5 13.0 ± 1.1

66.8 ± 0.6 71.7 ± 1.8 73.8 ± 1.3 87.9 ± 0.5 59.6 ± 0.5 64.7 ± 1.2

7.9 ± 0.4 1.5 ± 0.3 1.0 ± 0.4 0.0 ± 0.3 0.8 ± 0.2 0.9 ± 0.4

32.8 ± 2.0 16.7 ± 0.9 11.6 ± 0.7 12.6 ± 1.2 19.4 ± 0.6 14.3 ± 0.5

7.50 4.82 2.92 1.81 5.26 5.35

396

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

Table 5 Surface roughness parameters Ra, Rp and Rv expressed in lm. Bush hammered

AT BD FL SPL US VS

Ra

Rp

Rv

24 ± 5 23 ± 4 23 ± 6 25 ± 10 20 ± 5 13 ± 1

60 ± 10 56 ± 9 60 ± 22 63 ± 31 78 ± 49 32 ± 3

61 ± 17 53 ± 11 29 ± 24 62 ± 32 62 ± 24 42 ± 10

Polished D 10

2

2.68 ± 1.30 1.70 ± 0.04 2.08 ± 0.78 2.01 ± 0.36 4.56 ± 4.28 1.21 ± 0.16

Ra

Rp

Rv

9±2 6±2 5±1 24 ± 13 11 ± 6 8±3

24 ± 2 19 ± 5 15 ± 3 58 ± 28 52 ± 40 25 ± 11

27 ± 2 19 ± 4 19 ± 6 66 ± 22 62 ± 39 27 ± 10

D 10

2

0.85 ± 0.36 0.49 ± 0.11 0.98 ± 0.64 1.59 ± 0.41 3.64 ± 4.45 0.50 ± 0.05

Fig. 2. Bearing ratio curve of bush hammered and polished stones. Rmr(z): Material ratio of profile. Ratio of the material length of the profile elements Ml(z) at a given level z to the evaluation length. That is the percentage of the area analysed starting from the highest point to the lowest. Positive values on the vertical axis represent peaks, and negatives are valleys.

3.3. Colour change The stone colour changed depending on the stone and the test. The measured parameters did not show a progressive evolution. However, some trends could be observed. In Fig. 5 is shown the colour parameter variations for HNO3 > H2SO3, HNO3 < H2SO3 and H2SO3 tests. HNO3 test was not represented due to the lack of visible colour variation. (Fig. 5). L⁄ values (Fig. 5.i) indicated a decrease in this parameter that meant darkening stone. In general, the highest change was observed in the H2SO3 test followed by the HNO3 > H2SO3 and HNO3 < H2SO3 tests. Only AT showed its highest darkening for the HNO3 > H2SO3 test. Parameters a⁄ and b⁄ indicated a change in the tone of the stone. Parameter a⁄ increased in a general way for all of the stones and tests, so that stone colours trended to red tones (Fig. 5.ii). AT and VS exhibited a decrease in a⁄ for the H2SO3 test for both finishes. Parameter b⁄ increased, meaning that stones changed to yellow colours, with the exceptional decrease in AT for the H2SO3 and HNO3 < H2SO3 test with both finishes (Fig. 5.iii). Values over 3 were found in the DE⁄ parameter, so the decay can be observed by the human eye. Some trends can be defined with some exceptions, as usual in the assessment of natural materials (Fig. 5.iv). In the HNO3 test, all of the stones showed values under or close to 3, so no evident change occurred. For the three ‘‘other sedimentary stones”, the HNO3 > H2SO3 test yielded higher DE⁄ values than the HNO3 < H2SO3 test. In the ‘‘pure calcitic stones”, the trend was the opposite, with higher values in the

HNO3 < H2SO3 test than in the HNO3 > H2SO3 test. The H2SO3 test showed the greatest total colour change for all of the stones except for the travertine (AT). 3.4. Roughness Fig. 6 shows the differences in roughness parameters before and after the test. Due to the lack of differences found after exposure for the HNO3 test, roughness parameters were presented only for the other tests. In general, the HNO3 > H2SO3 and HNO3 < H2SO3 tests showed the highest variations with changes between 100% and 200% for most of the parameters and stones. As a general trend, it has to be remarked that five of the six stones (AT, BD, FL, US and VS) exhibited a higher increase in the bush hammered finish than in the polished finish for all of the tests. SPL texture and porosity did not allow large differences between the polished and bush hammered finishes. This explains why the results were similar for both finishes. Fractal dimension increased in the HNO3 > H2SO3 and HNO3 < H2SO3 tests (Fig. 6.i), and changes in the H2SO3 test were almost imperceptible. The bush hammered finish showed higher variation in fractal dimension than polished stones for the other tests. Only SPL had the opposite behaviour, which was a higher increase in the relief of polished stones compared with the bush hammered stones. In general, average roughness (Ra) increased in bush hammered stones for the HNO3 > H2SO3 and HNO3 < H2SO3 tests. Ra did not

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

397

Fig. 3. Changes produced in each polished stone after the tests. Black and white spots are clearly observable.

Fig. 4. Weight variation in grams before and after the tests for bush hammered finish (BH) and polished finish (P).

change in the polished samples of AT, BD and FL. In SPL and VS with the polished finish, this parameter increased for the HNO3 > H2SO3 and HNO3 < H2SO3 tests. With the polished finish, US showed the highest variation in all the stones, with a decrease of more than 30 lm. Peaks and valleys (Rp and Rv) also followed a similar trend. Bush hammered stones exhibited an increase if submitted to the HNO3 > H2SO3 and HNO3 < H2SO3 tests, with approximately 50– 100 lm for peaks and 60–200 lm for valleys. In general, polished stones showed lower values than bush hammered samples. Only SPL exhibited higher changes in polished samples. The H2SO3 test

did not cause any variation, or it was featured as a decrease in roughness. The bearing curves of the HNO3 > H2SO3 and HNO3 < H2SO3 tests showing the highest variations were graphed (Fig. 7). The trends observed above in the single parameters were clearly identifiable. Three behaviours could be defined from these graphs. The first behaviour was that no change could be observed, and distributions of deviations around the mean plane remained similar. This was the case of AT and VS for the HNO3 > H2SO3 test and of US for the HNO3 < H2SO3 test. The second behaviour was a general amplifica-

398

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

Fig. 5. Variations in colour parameters (i) L⁄, (ii) a⁄, (iii) b⁄. (iv) Total colour change (DE⁄) after each test for the bush hammered finish (B) and the polished finish (P). Values over 3 are considered as visible to the human eye.

Fig. 6. Variation in roughness parameters before and after the tests for the bush hammered finish (B) and the polished finish (P): (i) fractal dimension D*102; (ii) average roughness (Ra); (iii) peak roughness (Rp); (iv) valley roughness (Rv).

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

tion of the curve slope. This indicated a proportional augmentation of the deviations above and behind the mean line, as occurred in BD and SPL for both tests, and also in VS for the HNO3 > H2SO3 test. The third behaviour was an increase in extreme peaks and valleys while the mean valued remained intact, showing a curve with a sinusoidal shape. This means that a percentage of the area analysed corresponded to new and deeper valleys than in the sound stone. This was the case for AT in the HNO3 > H2SO3 test, US in the HNO3 < H2SO3 test, and FL in both tests. 3.5. SEM analysis and RUNSALT simulations Examination under SEM of the 6 types of stones showed similarity according to the type of tests, but differences in relation to the composition of stones. EDX analyses were carried out on salt crystallisation. Composition was confirmed by RUNSALT simulations at 20 °C and RH = 30–98% [43]. In the HNO3 test, the surface of the samples remained unchanged, almost without any crystallisation. Tiny white spots were observed that corresponded to salt crystallisations. However, due to the small size of the spots, EDX analysis was impossible [28]. For the other tests (HNO3 > H2SO3 and HNO3 < H2SO3) the amount of crystallisation increased with the proportion of H2SO3. For the HNO3 > H2SO3 test, RUNSALT simulations found that nitrate should crystallise in AT, BD and SPL, gypsum in FL and US, and no crystals precipitate in VS. SEM observations confirmed these calculations. Crystallisation of hemispherical-shaped nitrate occurred at the surface of the AT, BD and SPL samples (Fig. 8.a). Their sizes reached approximately 40 lm, and their density ranged from 4 (BD) to 30 (AT) crystals per mm2. According to RUNSALT, the nitrates found in the dolostone (BD) could be MgCa(NO3)4, Mg(NO3)2 Ca(NO3)2, and Ca(NO3)2 appeared in higher proportion in SPL and AT. SEM observations confirmed the predominance of gypsum in FL and US. VS did not show salts on the surface. For the HNO3 < H2SO3 test, RUNSALT simulations yielded results where no salts could crystallise in VS, and gypsum should be the only salt present in the other stones. This was confirmed by SEM observation, where only gypsum crystals were found. Short

399

rhomboidal-shaped gypsum crystals (SPL, US, FL), sometimes with needle-shaped crystals (AT) and forming small rosettes, covered the surface and filled the pores (Fig. 8.b). Sizes ranged from 30 to 100 lm diameter. The concentration of gypsum increased with the percentage of calcite in the stones, from 0.4% in the siliceous sandstone (VS) to 18% in the micritic limestone (FL). Compared with the other tests, the H2SO3 test showed the largest amount of gypsum crystals (Fig. 8.c and d). Like the HNO3 < H2SO3 test, the concentration was related to the calcite concentration in the stones. On stones with the lowest calcite concentrations (VS, BD), gypsum formed rosettes approximately 50 lm in diameter and composed of needle-shaped prisms of up to a few lm in width and 20 lm in length. The rosettes were dispersed and located in the macropores (VS), or they covered the entire surface, forming a microporous layer (BD, Fig. 8.c). On the other stones, gypsum formed a thick layer of cauliflower-shaped crystals that were sometimes truncated (SPL, US, FL, Fig. 8.d). 4. Discussion The appearance of new types of pollutants and their effect on the stone is still a subject under research. In this study, we exposed six stones to four polluted atmospheres, produced by single acid solution with HNO3 or H2SO3 and with two different HNO3/H2SO3 mixes. The stones had different compositions, three of them being ‘‘pure calcitic stone” and the other three being ‘‘other sedimentary stones”. All of the stones were tested with two different finishes: polished and bush hammered. This study showed a high variability of the results, due mainly to the difference of material composition and textural properties as well as initial surface roughness given by the artificial finish. However, some trends have been observed. 4.1. Influence of acid concentrations In this subchapter, the effect of each atmosphere is discussed, taking into account all of the stone samples and making no special distinction between stone composition or finish. First, single gas atmospheres are discussed to highlight the extreme behaviours,

Fig. 7. Bearing area curve before and after the NOx > SO2 and NOx < SO2 tests. Positive values in the vertical axis represent peaks, and negatives are valleys. The abscissa axis indicates the percentage of the area analysed, starting from the highest point to the lowest. Only the bush hammered finish (BH) was represented due to its higher variations compared with the polished finish.

400

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

Fig. 8. Characteristic SEM images of the surfaces of the stones subjected to (a) HNO3 < H2SO3 atmosphere (hemispherical-shaped nitrate crystal in AT); (b) HNO3 < H2SO3 atmosphere (needle-shaped crystals of gypsum, forming rosettesin AT); and H2SO3 test atmosphere, exhibiting (c) microporous layer of gypsum crystals and needle-shaped crystals in AT, and (d) cauliflower-shaped crystals, sometimes truncated in SPL.

and then, the analysis of mixed atmospheres is assessed. The hermetic recipient produced a saturated acid atmosphere for 21 days. Stones were already soaked in water, so most of the pores were filled, and a water film layer was developed on the stone surface. Acid atmospheres reacted with the stone surface, and dissolution or mineral precipitation may have taken place. – HNO3 test: After exposure, no visual changes were observed on the stones. Colour measurement revealed that DE⁄ was lower than 3 for all types of stone, confirming that colour variation was not perceptible by the human eye. Nevertheless an increase in weight was measured. That agrees with previous researches from Haneef et al. (1992) and Vazquez et al. (2010) [9,28]. This is explained by the fact that NO2 enters into the pores and attacks the stone from the interior [15,16]. Statistical analysis revealed that when all of the samples were compared before and after this test (N = 36), the weight variation was significant. This indicated that a punctual surface attack leading to salt crystallization occurred even if they were scarce under SEM. No variation in roughness was obtained in this test. These results match those of Camaiti et al. (2007) [16], who did not observe any change in colour or roughness after the exposure to nitrogen-rich atmospheres. – H2SO3 test: SO2 exposure lead to gypsum crystallization as observed in previous studies [7,9,11,24,27,28,34]. In this research, stones turned to yellow as observed with the naked eye. Changes in colour were evident and also statistically significant. DE⁄ showed values higher than 3 and up to almost 20. There was a general decrease in lightness (L⁄), and generally an important increase in yellowing (b⁄). The brownish-yellow

travertine (AT) behaved oppositely maybe due to the initial colour and heterogeneity and the presence of gypsum as white spots that induced an increase in lightness and a decrease in the yellow parameter. The white crystals observed on the surface of the samples were found mainly at grain boundaries and on and around iron oxides. These crystals formed a crust on the surface, hardly conditioned by the stone texture. A statistically significant increase in weight of all the samples was measured, explained by the large amount of gypsum crystallisation. The roughness showed two behaviours: no variation or a decrease. The decrease in roughness was explained by the filling of the pre-existing pores and irregularities by the gypsum crystals confirmed by SEM observations and also constated by Vazquez and Alonso (2010) and Vazquez et al. (2015) [27,34]. – Mixed atmosphere tests: They produce more damage than single acidic atmospheres and revealed in general more important changes in most of the properties than in single acidic tests. Yellowing and whitening could be observed in some samples, but without a defined trend. Colour measurements indicated that, in general, L⁄ decreased very slightly and a⁄ increased to red, whereas the behaviour of b⁄ varied randomly. Statistically, in this research the weight variations were significant. The pure sandstone (VS) was the only one that did not show high weight gain, because scarce chemical reactions occurred. The other stones showed the highest increase in the HNO3 > H2SO3 test and yielded salt crystals with a varied composition, including calcium and magnesium nitrates and sulphates such as Ca,Mg (NO3)2 that may influence the highest weight gain. In contrast, only gypsum was found in the HNO3 < H2SO3 test. This means

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

401

Fig. 9. Surface differences between BD and SPL with both finishes. DRa indicated the variation in average roughness before and after the test, expressed as a percentage. Roughness changes are evident for polished surfaces only in SPL.

that the proportion of HNO3 and H2SO3 may control whether HNO3 reacts with the stone to form nitrites and nitrates, or acts as catalyser of SO2 reactions. Both tests yielded high and similar roughness variations on the stone surface. All parameters increased, which may implicate dissolution [30] and salt precipitation in a non-homogeneous way, indicating stone weathering [33,34]. 4.2. Influence of composition and stone properties Stone attack by acid atmospheres occurred on reactive surfaces such as calcite. If no calcium is present in the stone composition, a gypsum crust may be formed by a sufficient contribution of SO2 or other pollutants and an external calcium source [6]. If no external calcium source is present, SO2 reaction can be produced by ion exchange from clays or the hydrolysis of aluminium feldspars or micas in the acid atmosphere [11]. VS, the siliceous sandstone with only 0.3% Ca, exhibited the highest gain of weight for the H2SO3 test. These values were two to three times higher than in the other

tests (Fig. 4). This indicates that even if HNO3 accelerates the SO2– Ca interaction, the catalytic effect does not influence the mass gain. The highest specific surface area of VS (3.76 m2/g), together with an average porosity among the studied stones and the fact that no dissolution took place, explains the highest weight gain. When a stone with calcium in its composition is in contact with a SO2 atmosphere, a gypsum crust is formed, although crystallisation patterns and substrate reactivity depends on the texture of the stones [24,37]. To crystallise gypsum, a previous dissolution or a replacement process must take place [30,37]. BD, the dolomitic limestone with only 18% calcite, is the stone with the second highest weight gain after the H2SO3 test. The dissolution process concentrated more in calcite than in dolomite crystals due to their different solubilities. Nitrogen compounds react preferentially with Mg, resulting in both calcium and/or magnesium compounds [19,23,26,29], which may explain the high weight gain of BD after the HNO3 test. According to RUNSALT, magnesium salts as Ca,Mg(NO3)2 could form on the stone, favouring a weight increase similar to the calcitic stones. BD showed the

402

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

lowest porosity (approximately 10%) but also the highest amount of access to the smallest pore radii, with 45% of the pores below 1 lm. In addition, according to Vazquez et al. (2013) [39], BD has the lowest vapour permeability among the studied stones favouring an accumulation of salts for the HNO3 > H2SO3 test. US, with half of its composition of a carbonate nature, exhibited high weight gain due to surface reaction for the HNO3 > H2SO3 test, and less weight gain due to possible dissolution when SO2 was the main reactant. FL is a calcareous stone with the highest specific surface area and almost 30% porosity, which is coherent with the highest weight gain results for the HNO3 > H2SO3 test among all of the stones, and high values in the other tests. SPL has large pores that avoid HNO3 reaction, whereas the reaction with the HNO3 > H2SO3 test lead to an increase in weight by salt crystallisation. AT has large pores that allowed the water to enter through the stone and smaller pores in which salts precipitate. Thus, the acid dissolution and the salt precipitation is more equilibrated than in other calcareous stones with less porosity or smaller sized pores. Statistically, when comparing weight variation before and after the tests of the two groups (‘‘pure calcitic stone” and ‘‘other sedimentary stones”), the changes were always significant. Colour changes are related to the reactive surface and also to the original colour of the stone and the colour of the deposits [27]. In this research the carbonate content did not influence the final colour change. Parameter a⁄ increased in all of the stones. This reaction may be due to the iron oxidation in the acid atmosphere that may lead to redness of the stone [9–11]. Regarding the iron content of the studied stones, US (a sandstone with 45% carbonate content) showed the highest content (above 3%). However, the total colour changes were among the lowest. The lack of noticeable colour changes may be due to the fact that the distribution of iron molecules through the minerals was not optimal to induce a significative colour change. Statistically, when comparing ‘‘pure calcitic stones” and ‘‘other sedimentary stones” (2 groups, N = 90) before and after the tests, none of the parameter variations were significant in the HNO3 test. Parameters a⁄ and b⁄ varied significantly in ‘‘other sedimentary stones” in the HNO3 > H2SO3 test, and only b⁄ varied significantly in the H2SO3 test. No significant differences in roughness were observed between ‘‘pure calcitic stones” and ‘‘other sedimentary stones”. Statistically, only D variation was significant for the HNO3 > H2SO3 test for pure calcitic stones. However, an increase in peaks and valleys for these three stones was observed in the bearing ratio curve. 4.3. Influence of surface finish Lipfert (1989) [17] already wondered about the influence of the surface relief on the pollutants’ attack. It was determined that carbonate deterioration is enhanced by increased roughness [21] and that the roughness of already formed black crusts increases the development of the crust, trapping new particles [22]. In our research, weight gain in relation to finish varied depending on the acid atmosphere. Thus, polished surfaces yielded higher weight variation than bush hammered surfaces for the HNO3 test. The bush hammered finish may retain the particles on the surface due to the deterioration of the artificial processing avoiding partially the migration into the stone and the salt crystallization. For the other tests, weight gain was similar for both finishes or high for bush hammered. Colour changes did not show any general trend in relation to finish, for any parameter. In general, roughness increased in bush hammered AT, BD, FL and US, whereas polished slabs did not exhibit any change (Fig. 9, BD). This was evident in mixed atmosphere tests more than in the single acidic atmosphere tests. In SPL and VS, no large differences in roughness were found between both finishes. SPL showed pores of millimetric size, even in the polish surface, that explain the higher attack of the acid

atmospheres to this polished stone compared with the other calcareous stones (Fig. 9, SPL). Rp (maximum peak height) was the parameter with the highest variation, which indicates a predominance of crystal formation, slightly higher in bush hammered than in polished specimens. Statistically, only fractal dimension showed a significant change in the HNO3 > H2SO3 test for both finishes and in the HNO3 < H2SO3 test for bush hammered samples. That means an increase of relief of the stone surface but also that D is a meaningful parameter in stone weathering. 5. Conclusions Acid atmospheres may cause damage to all types of stones, even those with extremely low Ca content, as in the case of VS, the siliceous sandstone with only 0.3% Ca. The damage is lower in nonpure calcitic stones, and only scarce reaction occurs. In carbonate-containing stones, the variations are more related to pore size, porosity distribution and specific surface area than to calcium proportion. Colour changes for mixed atmospheres and for H2SO3 attack. However, the intensity of the variation depends more on the inherent stone colour than on the composition of the stone or the atmosphere that the stone is exposed to. The influence of finish is evident in terms of roughness, with higher variations in bush hammered samples than in polished samples. Fractal dimension (D) and bearing ratio area curve were good parameters for evaluating the surface weathering, with a general overview of the increase or decrease in roughness (D) and also of the new distribution of peaks and valleys (bearing). In mixed atmospheres, the HNO3/H2SO3 ratio plays an important role in the reactions with the stone and salt products. With a low proportion of H2SO3 in the mixed atmosphere, nitrates and sulphates were found, indicating an interaction of HNO3 with the stone. With a high proportion of H2SO3 in the mixed atmosphere, only gypsum precipitated on the stone surface. HNO3 acted as a catalyser of the H2SO3 reaction without interacting with the stone. These results indicate the importance of the proportion of pollutants and their different effects on building stones. Futures works will be focused on the deepening on the chemical reactions that took place on the surface and their reaction kinetics. Conflict of interest The authors confirm that there is no conflict of interest in the publication of this article Acknowledgements This research was funded mainly by the project MEC (PN I+D+i) MAT2008-06799-C03-01 and the Project IFEPAR (Reims Metropole). The authors thank the reviewers for their helpful comments. References [1] P. Brimblecombe, C.M. Grossi, Millennium-long damage to building materials in London, Sci. Total Environ. 407 (4) (2009) 1354–1361, http://dx.doi.org/ 10.1016/j.scitotenv.2008.09.037. [2] M. Ferm, F. De Santis, C. Varotsos, Nitric acid measurements in connection with corrosion studies, Atmos. Environ. 39 (35) (2005) 6664–6672, http://dx.doi. org/10.1016/j.atmosenv.2005.07.044. [3] Á. Török, N. Rozgonyi, Morphology and mineralogy of weathering crusts on highly porous oolitic limestones, a case study from Budapest, Environ. Geol. 46 (3–4) (2004) 333–349, http://dx.doi.org/10.1007/s00254-004-1036-x. [4] C.M. Grossi, P. Brimblecombe, R.M. Esbert, F.J. Alonso, Colour changes in architectural limestones from pollution and cleaning, Colour Res. Appl. 32 (4) (2007) 320–331, http://dx.doi.org/10.1002/col.20322. [5] F. Monna, A. Puertas, F. Lévêque, R. Losno, G. Fronteau, B. Marin, J. Domonik, C. Chateau, Geochemical records of limestone façades exposed to urban

P. Vazquez et al. / Construction and Building Materials 106 (2016) 392–403

[6]

[7]

[8] [9]

[10]

[11] [12]

[13] [14]

[15] [16]

[17]

[18]

[19] [20]

[21]

[22]

[23]

[24]

[25]

atmospheric contamination as monitoring tools?, Atmos Environ. 42 (5) (2008) 999–1011, http://dx.doi.org/10.1016/j.atmosenv.2007.10.021. B. Graue, S. Siegesmund, P. Oyhantcabal, R. Naumann, T. Licha, K. Simon, The effect of air pollution on stone decay: the decay of the Drachenfels trachyte in industrial, urban, and rural environments—a case study of the Cologne, Altenberg and Xanten cathedrals, Environ. Earth Sci. 69 (4) (2013) 1095– 1124, http://dx.doi.org/10.1007/s12665-012-2161-6. S. Janvier-Badosa, K. Beck, X. Brunetaud, M. Al-Mukhtar, The occurrence of gypsum in the scaling of stones at the Castle of Chambord (France), Environ. Earth Sci. 71 (11) (2014) 4751–4759, http://dx.doi.org/10.1007/s12665-0132865-2. S.W. Massey, The effects of ozone and NOx on the deterioration of calcareous stone, Sci. Total Environ. 227 (2–3) (1999) 109–121. S.J. Haneef, J.B. Johnson, C. Dickinson, G.E. Thompson, G.C. Wood, Effect of dry deposition of NO(x) and SO2 gaseous pollutants on the degradation of calcareous building stones, Atmos. Environ. 26 (16) (1992) 2963–2974. S.J. Haneef, J.B. Johnson, G.E. Thompson, G.C. Wood, Effects of dry deposition of pollutant gases on the degradation of pentelic marble, Corros. Sci. 35 (1–4) (1993) 743–747. U. Müller, The mineralogical composition of sandstone and its effect on sulphur dioxide deposition, Mater. Constr. 58 (2008) 81–95. Y. Bai, G.E. Thompson, S. Martinez-Ramirez, S. Brüeggerhoff, Mineralogical study of salt crusts formed on historic building stones, Sci. Total Environ. 302 (1–3) (2003) 247–251. H. Böke, E.H. Gökturk, E.N. Caner-Saltik, S. Demirci, Effect of airborne particle on SO2–calcite reaction, Appl. Surf. Sci. 140 (1999) 70–82. Y. Bai, G.E. Thompson, S. Martinez-Ramirez, Effects of NO2 on oxidation mechanisms of atmospheric pollutant SO2 over Baumberger sandstone, Build. Environ. 41 (4) (2006) 486–491. D. Sikiotis, P. Kirkitsos, The adverse effects of nitrates on stone monuments, Sci. Total Environ. 171 (1–3) (1995) 173–182. M. Camaiti, S. Bugani, E. Bernardi, L. Morselli, M. Matteini, Effects of atmospheric NOx on biocalcarenite coated with different conservation products, Appl. Geochem. 22 (6) (2007) 1248–1254, http://dx.doi.org/ 10.1016/j.apgeochem.2007.03.035. F.W. Lipfert, Atmospheric damage to calcareous stones: comparison and reconciliation of recent experimental findings, Atmos. Environ. (1967) 23 (2) (1989) 415–429. H. Sweevers, R. Van Grieken, Analytical study of the deterioration of sandstone, marble and granite, Atmos. Environ. Part B. Urban Atmos. 26 (2) (1992) 159–163. C.M. Grossi, M. Murray, R.N. Butlin, Response of porous building stones to acid deposition, Water Air Soil Pollut. 85 (4) (1995) 2713–2718. A. Bonazza, P. Messina, C. Sabbioni, C.M. Grossi, P. Brimblecombe, Mapping the impact of climate change on surface recession of carbonate buildings in Europe, Sci. Total Environ. 407 (6) (2009) 2039–2050. D.A. Dolske, Deposition of atmospheric pollutants to monuments, statues, and buildings, Sci. Total Environ. 167 (1–3) (1995) 15–31, http://dx.doi.org/ 10.1016/0048-9697(95)04566-J. P. Ausset, M. Del Monte, R.A. Lefèvre, Embryonic sulphated black crusts on carbonate rocks in atmospheric simulation chamber and in the field: role of carbonaceous fly-ash, Atmos. Environ. 33 (1999) 1525–1534. W.G. Cobourn, K.L. Gauri, S. Tambe, S. Li, E. Saltik, Laboratory measurements of sulfur dioxide deposition velocity on marble and dolomite stone surfaces, Atmos. Environ. Part B. Urban Atmos. 27 (2) (1993) 193–201. K. Malaga-Starzec, I. Panas, J.E. Lindqvist, O. Lindqvist, Efflorescence on thin sections of calcareous stones, J. Cult. Heritage 4 (4) (2003) 313–318, http://dx. doi.org/10.1016/j.culher.2003.09.002. E. Franzoni, E. Sassoni, Correlation between microstructural characteristics and weight loss of natural stones exposed to simulated acid rain, Sci. Total Environ. 412 (2011) 278–285.

403

[26] C.M. Grossi, M. Murray, Characteristics of carbonate building stones that influence the dry deposition of acidic gases, Constr. Build. Mater. 13 (3) (1999) 101–108, http://dx.doi.org/10.1016/S0950-0618(99)00019-7. [27] P. Vazquez, F.J. Alonso, SO2 exposure effect in monumental granitic stones. Proceeding, in: IMA2010 Abstracts Volume, 6th Volume of the Abstract Series of the journal Acta Mineralogica-Petrographica. International Mineralogical Association (IMA) Conference: Bonds & Bridges, 21–27 August 2010, Budapest, 2010. [28] P. Vázquez, L. Carrizo, M. Fernandez, F.J. Alonso, Cambios en la superficie de rocas sedimentarias tras su exposicion a HNO3, in: Macla, 13, XXVII. Reunion Sociedad Espanola de Mineralogıa (SEM), 16–18 September, Madrid, Spain, 2010, pp. 219–220. [29] A. Luque, M.V.M. de Yuso, G. Cultrone, E. Sebastián, Analysis of the surface of different marbles by X-ray photoelectron spectroscopy (XPS) to evaluate decay by SO2 attack, Environ. Earth Sci. 68 (3) (2013) 833–845. [30] L. Tecer, Laboratory experiments on the investigation of the effects of sulphuric acid on the deterioration of carbonate stones and surface corrosion, Water Air Soil Pollut. 114 (1–2) (1999) 1–12. [31] J.L.P. Bernal, M.A.B. Lopez, The fractal dimension of stone pore surface as weathering descriptor, Appl. Surf. Sci. 161 (1) (2000) 47–53. [32] F.J. Alonso, P. Vázquez, R.M. Esbert, J. Ordaz, Durabilidad de granitos ornamentales: valoración de los daños inducidos por el ensayo de cristalización de sales Ornamental granite durability: evaluation of damage caused by salt crystallisation test, Mat. Const. 58 (2008) 191–201. [33] P. Vázquez, A. Luque, F.J. Alonso, C.M. Grossi, Surface changes on crystalline stones due to salt crystallisation, Environ. Earth Sci. 69 (4) (2013) 1237–1248, http://dx.doi.org/10.1007/s12665-012-2003-6. [34] P. Vázquez, B. Menéndez, M.F.C. Denecker, C. Thomachot-Schneider, Comparison between petrophysical properties, durability and use of two limestones of the Paris region, Geol. Soc. Lond., Spec. Publ. 416 (2015). SP41615. [35] D. Benavente, F. Martinez-Verdi, A. Bernabeu, V. Viqueira, R. Fort, M.A. Garcia del Cura, S. Ordonez, Influence of surface roughness on colour changes in building stones, Colour Res. Appl. 28 (5) (2003) 343–351, http://dx.doi.org/ 10.1002/col.10178. [36] CEN (EUROPEAN COMMITTEE FOR STANDARDISATION), EN 13919: Natural Stone Test Methods. DETERMINATION of Resistance to Ageing by SO2 Action in the Presence of Humidity, CEN, Brussels, 2002. [37] J. Dewanckele, T. De Kock, M.A. Boone, V. Cnudde, L. Brabant, M.N. Boone P. Jacobs, 4D imaging and quantification of pore structure modifications inside natural building stones by means of high resolution X-ray CT, Sci. Total Environ. 416 (2012) 436–448, http://dx.doi.org/10.1016/j.scitotenv.2011. 11.018. [38] E. Molina, G. Cultrone, E. Sebastián, F.J. Alonso, L. Carrizo, J. Gisbert, O. Buj, The pore system of sedimentary rocks as a key factor in the durability of building materials, Eng. Geol. 118 (3–4) (2011) 110–121, http://dx.doi.org/10.1016/j. enggeo.2011.01.008. [39] P. Vázquez, F.J. Alonso, L. Carrizo, E. Molina, G. Cultrone, M. Blanco, I. Zamora, Evaluation of the petrophysical properties of sedimentary building stones in order to establish quality criteria, Constr. Build. Mater. 41 (2013) 868–878, http://dx.doi.org/10.1016/j.conbuildmat.2012.12.026. [40] CEN (EUROPEAN COMMITTEE FOR STANDARDISATION), EN ISO 105-J03: Textiles – Tests for Colour Fastness – Part J03 Calculation of Colour Differences, CEN, Brussels, 2009. [41] CEN (EUROPEAN COMMITTEE FOR STANDARDISATION), Geometrical Product Specifications (GPS) – Surface Texture: Profile Method – Terms, Definitions and Surface Texture Parameters, CEN, Brussels, 1997. [42] B.B. Mandelbrot, The fractal geometry of nature, vol. 173, Macmillan, 1983. [43] http://science.sdf-eu.org/runsalt/.