Monument Future

Influence of zeolites and swellable clay minerals

Scanning electron microscopy (SEM) images reveal the type and location of clay minerals in MG and CV_O (Fig. 3a–d). Swellable smectite, muscovite and kaolinite are located in both pore space (Fig. 3a) and in the matrix (Fig. 3b,d) as well as on grain boundaries (Fig. 3c). Smectite has the ability to cause volume increase by swelling, especially when it is located on grain contacts. The alteration of feldspar to clay minerals often leads to the formation of intracrystalline porosity and therefore increases the effective porosity (Fig. 3c).

The scanning electron microscopy (SEM) images (Fig. 3e–h) reveal very small mordenite needles and clinotilolite laths in CV_E, which cause a high specific surface area (Fig. 3e–h).

The zeolite-rich samples CV_O, CV_E and CA_E show the highest hygroscopic water sorption and hydric expansion of all samples (see Kück et al. 2020a in this issue). During thermal expansion, the zeolite-rich samples (CV_O, CV_E and CA_E) show a large difference between the first and the second heating cycle, with high residual strain (Fig. 4). After the first wet cycle the samples show a decrease in residual strain. CA_O and CA_E do not recover from contraction after the second drying cycle. The Mitla samples and QB are rather unaffected by both thermal and thermohyric dilatation (see Kück et al. 2020a).

Figure 3: a–d: Different appearances of swellable clay minerals. a): smectite appears as a ‘spiderweb’ in the pore space as relict of weathered pumice clast in MG, b): kaolinite booklet and smectite in densely packed matrix, c): dissolved feldspar with clay minerals on the grain boundary reveals intracrystalline porosity, d): smectite in the matrix of CVO; e–h): Zeolites in CVE cause a high specific surface area. e): flaky zeolite in the matrix, f): zeolites grow into pore space, g): clinoptilolite laths in the matrix, h): 100 nm thin mordenite needles in a pore.

123

Figure 4: Residual strain [mm/m] of CVE in the X- and Z-direction for two heating cycles from 20–90 °C. Noticeable are intense shrinking and non-reversible thermal dilatation.

Discussion and conclusions

The cation exchange capacity (CEC) value depends on both clay minerals and zeolites, however, the CEC method used is not suitable for CEC analysis of zeolites because the large colored Cu-trien cations cannot enter the cavities of zeolites (Meier and Kahr 1999). For the analysed samples a clear trend between CEC and hydric expansion could not be observed (Fig. 5a). On the other hand, a linear dependency of hydric expansion and hygroscopic sorption on the specific surface area is observable (Fig. 5b, Fig. 6b). The specific surface area increases when clay minerals and zeolites are present in a volcanic tuff, because they yield a high share of micropores (often in the nanometer-scale) within the rock fabric. The higher the specific surface area is, the more water can interact with the rock and the expansion is resultingly larger. Zeolites can adsorb and release substantial amounts of water and therefore passively lead to hydric expansion, whenever water is released during dehydration.

The results of this study show, that the presence of both zeolites and swellable clay minerals has a combined effect on the water-related weathering behavior of tuff building stones. The porespace properties (e. g. pore radii, specific surface area) and the water transport and storage properties 124of volcanic tuff rocks are significantly influenced, by either of them. Volcanic tuff rocks with a significant amount of swellable clay minerals as well as zeolites show extremely high hygroscopic water sorption and hydric swelling. Thermal effects like shrinkage and fracturing during drying are particularly high. This study showed that the identification of the clay mineral and zeolite content, as well as their location and shape within the rock fabric is an important measure for the prediction of the weathering behavior of tuff building stones. For the planning of effective conservational treatment of tuff building stones a clay mineral and zeolite analysis should be considered substantial.

Figure 5: Correlation between a: hydric expansion [mm/m] and CEC [meq/100 g], b: hydric expansion [mm/m] and BET [m²/g].

Figure 6: Correlation between a) hygroscopic water sorption at 95 % rh [wt.-%] and CEC [meq/100 g], b) hygroscopic water sorption at 95 % rh [wt.-%] and BET [m²/g].

Acknowledgements

This work was supported by the German Research Foundation (Si-438/52-1) and the German Federal Environmental Foundation (AZ20017/481). We would like to thank J. L. Noria Sánchez, A. E. Andrade Cuautle and M. Á. González for their support of the onsite work. For the laboratory support and helpful comments we thank K. Wemmer, J. Menningen, W. Wedekind and C. Gross.

References

Colella, C., Gennaro, M. d.’ and Aiello, R. (2001), Use of zeolitic tuff in the building industry, Reviews in Mineralogy and Geochemistry, Vol. 45 No. 1, pp. 551–587.

Di Tchernev (1978), Solar energy application of natural zeolites, Natural zeolites: Occurrence, properties, use, Vol. 474, p. 485.

Gonzalez, I. J. and Scherer, G. W. (2004), Effect of swelling inhibitors on the swelling and stress relaxation of clay bearing stones, Environmental Geology, Vol. 46 No. 3–4, pp. 364–377.

Korkuna, O., Leboda, R., Skubiszewska-Zie, J., Vrublevska, T., Gunko, V. M. and Ryczkowski, J. (2006), Structural and physicochemical properties of natural zeolites: clinoptilolite and mordenite, Microporous and Mesoporous Materials, Vol. 87 No. 3, pp. 243–254.

Kück, A., Pötzl, C., López-Doncel, R. A., Dohrmann, R. and Siegesmund, S. (2020a), Tuffs in pre-Columbian and colonial architecture of Oaxaca, Mexico, in Siegesmund, S. and Middendorf, B. (Eds.), Monument future: Decay and conservation of stone, Göttingen, Kassel, Mitteldeutscher Verlag, Halle.

Pablo-Galán, L. de (1986), Geochemical trends in the alteration of Miocene vitric tuffs to economic zeolite deposits, Oaxaca, Mexico, Applied geochemistry, Vol. 1 No. 2, pp. 273–285.

Pötzl, C., Dohrmann, R. and Siegesmund, S. (2018a), Clay swelling mechanism in tuff stones: an example of the Hilbersdorf Tuff from Chemnitz, Germany, Environmental earth sciences, Vol. 77 No. 5, p. 188.

Pötzl, C., Siegesmund, S., Dohrmann, R., Koning, J. M. and Wedekind, W. (2018b), Deterioration of volcanic tuff rocks from Armenia. Constraints on salt crystallization and hydric expansion, Environmental earth sciences, Vol. 77 No. 19, p. 660.

Ruedrich, J., Bartelsen, T., Dohrmann, R. and Siegesmund, S. (2011), Moisture expansion as a deterioration factor for sandstone used in buildings, Environmental earth sciences, Vol. 63 No. 7–8, pp. 1545–1564.

Siegesmund, S. and Dürrast, H. (2011), Physical and mechanical properties of rocks, in Siegesmund, S. and Snethlage, R. (Eds.), Stone in architecture, Springer, pp. 97–225.

Snethlage, R., Wendler, E. and Klemm, D. D. (1995), Tenside im Gesteinsschutz-bisherige Resultate mit einem neuen Konzept zur Erhaltung von Denkmälern aus Naturstein, Denkmalpflege und Naturwissenschaft-Natursteinkonservierung I. Verlag Ernst & Sohn, Berlin, pp. 127–146.

Wedekind, W., López-Doncel, R., Dohrmann, R., Kocher, M. and Siegesmund, S. (2013), Weathering of volcanic tuff rocks caused by moisture expansion, Environmental earth sciences, Vol. 69 No. 4, pp. 1203–1224.

125

PHYSICO-CHEMICAL CHARACTERIZATION OF THE CARTAGENA WALL AND QUARRY MATERIAL STONE USED FOR ITS RESTORATION

Manuel Saba¹, Juan Lizarazo-Marriaga², Nicole Hernández-Romero², Cristina Tedeschi³, Edgar Quiñones-Bolaños¹

IN: SIEGESMUND, S. & MIDDENDORF, B. (EDS.): MONUMENT FUTURE: DECAY AND CONSERVATION OF STONE.

– PROCEEDINGS OF THE 14TH INTERNATIONAL CONGRESS ON THE DETERIORATION AND CONSERVATION OF STONE –

VOLUME I AND VOLUME II. MITTELDEUTSCHER VERLAG 2020.

1 University of Cartagena, Research Group Esconpat, Faculty of Engineering, Cartagena, Colombia

2 Universidad Nacional de Colombia, Department of Civil and Agricultural Engineering, Bogotá, Colombia

3 Department of Civil and Environmental Engineering, Politecnico di Milano, Italy

4 University of Cartagena, Research Group Environmental Modelling, Faculty of Engineering, Cartagena, Colombia

Abstract

The deterioration of historical structures is a topic of primary importance due to their historical, cultural and economic importance. Aggressive environments around structures can depend on anthropogenic factors such as NO₂, SO₂ gases among others, and environmental factors such as differential dilation, differential hydric dilation and crystallization. In the present study is proposed a physical-chemical characterization of the structure and quarry stone used for the replacement of deteriorated blocks of the defensive Wall of Cartagena, Colombia, UNESCO Cultural Heritage since 1984. X-ray and ionic chromatography identification were done in the structure stone. Furthermore, is proposed a Petrographical comparison between structure and quarry material as well as mechanical characterization of the quarry material.

The structure is composed mainly by soft limestone. Similar physico-chemical characteristics were found through X-Ray and Petrographical comparison, while salt crystallization is found playing a secondary role in the structure deterioration.

Introduction

According to the United Nations’ Educational, Scientific and Cultural Organization (UNESCO), there are around 869 sites of cultural interest around the world (UNESCO 2019). They were selected for their historical relevance and uniqueness. All of them struggle with surrounding aggressive climate conditions that accelerate their deterioration. Currently, 53 of those sites are rated as being in imminent danger due to anthropogenic and natural factors (UNESCO 2019).

For example, the Fortifications on the Caribbean Side of Panama (Portobelo-San Lorenzo), part of the defensive colonial system built by the Spanish Crown to protect transatlantic trade, constitute a valuable example of 17th and 18th-century military architecture, whose integrity has been compromised by environmental factors, uncontrolled urban extension, development and the lack of maintenance and management. Dozens of structures have been destroyed in the past in the name of Civilization, or because of high deterioration and/or earthquakes. An example of this is the Noto Cathedral, Italy, a historic masonry building that suddenly collapsed in 1996 (Binda et al. 1999) 126producing during centuries a redistribution of stresses from the core of lime mortar concrete to the external cladding of stiff masonry. This is likely one of the causes of long-time damage of some ancient masonry towers. With these motivations, coupled processes of moisture diffusion, carbon dioxide diffusion and carbonation reaction are analyzed numerically. Due to the absence of models and data for lime mortar, one of the simplest models proposed for Portland cement concrete is adapted for this purpose. The results reveal the time scales of the processes involved and their dependence on wall thickness (size. Factors responsible for over-stressing and damaging the monuments and thus inhibiting their conservation are high levels of air pollution in the atmosphere, daily and seasonal cycles of temperature and humidity, sea spray (or salt spray), rising damp, atmospheric conditions (rainwater, wind and sunlight exposure (Kameni & Orosa 2016)).

During the process of restoration of cultural heritage buildings, it is usual to replace highly deteriorated stones and mortars with new ones. Unfortunately, often the choice of replacement materials is done without sufficient preliminary investigation of the properties of the existing materials. In order to come to a selection of replacement stones compatible with the existent ones, several material properties need to be taken into account, such as petrographic properties, mechanical strength, moisture transport behaviour, colour, texture etc., (Baronio G., et al. 2003), (Binda L. et al. 2003).

In the present research the study case of the Fortresses of Cartagena is reported. They are UNESCO Cultural heritage since 1984, nevertheless, UNESCO inspectors identified and high level of deterioration and lack of long-term maintenance plans, (UNESCO, 1984). The structure stands in front of the Caribbean Sea, in a tropical area, therefore, salt crystallization process was studied on the structure surface to define its role in the structure deterioration.

Quarry samples were taken from an ancient quarry in Cartagena (Tierrabomba Island) in the geological Formation called La Popa. La Popa Formation rests on the Bayunca Formation (of the Pliocene) and it was formed during the Upper Pleistocene. It is conformed by coral reefs formed on an underwater platform in an area with little contribution of terrigenous sediments, clear waters and temperatures between 21 °C and 25 °C. High porous limestones (> 30 %) with bulk density < 1,500 kg m^–3 are common in the area. Similar physical-mechanical characteristics were identified in the Cartagena’s Fortification from previous analysis, (Saba et al. 2019).

Therefore, structure and quarry samples were physical-chemical compared to assess the reliability to use them as a replacement of the deteriorated structure stones.

Materials and Methods

5 structure samples were collected from the stone surface (5 mm depth). Additionally, thin sections were done on those samples following the ASTM C1721 – 15, (2015) standard using blue epoxydic resin. Thin sections were petrographical analysed with an Olympus CX 31 microscope with magnifications ranging from x5 to x100 for assessing the presence of bioclasts, type of cement, terrigenous, distribution and quantification of primary and secondary porosity. Each thin section has a dimension of 4.5 cm × 2.6 cm. Point counting technique was used in a mesh of 300 equidistant points.

On these samples, ion analyses were carried out using Ion Chromatograph (IC). Powdered samples were dried at 60 °C until constant weight. Saline solubilisation was achieved by shaking 1 g of each dried sample in 100 ml of ultra-pure water. The 10 ml of obtained supernatants were filtered through a 0.2 µm PTFE membrane. The separation of cations Na+, Mg+, K+ was achieved by using a stationary-phase featuring a CS12A 250*4 mm column with a 10*4 mm guard (Dionex). As for anions Cl–, SO₄–, NO₃–, the stationary phase featured a AS9-HC 250 *4 mm column with a 10*4 mm guard (Dionex), (Nasraoui M. et al. 2009)the standard analytical equipment as ion chromatography (IC).

80 cubes of 5.0 × 5.0 × 5.0 cm were selected in the original quarry of the structure for the physical-mechanical 127characterization. Specimens were Characterized following the Natural Stone Test Methods (UNI EN 1936:2007 Natural stone test methods – Determination of real density and apparent density, and of total and open porosity, 2007). Real Volume V_R (m³), Open Porosity P_o (%) and Apparent Density ρ_b (kg m^–3) were calculated, (1–3).

Where m_d (g) is the Dry mass, m_s (g) Saturated mass, m_h (g) water immersion mass, ρ_rh water density at 20 °C, 998 kg m^-3.

Stone Uniaxial Compressive Strength (SUCS) measurements were done on the quarry samples.

X-Ray analysis were done in the 5 quarry samples and in 2 structure samples.

Results and Discussion

Thin section analysis results showed in Table 1 highlights that structure and quarry sample are both classified as Packestone according to Dunham (1962). They have similar ranges of Bioclasts and Sparry cement, while Primary porosity is significantly higher in the quarry samples (see Figure 1). Increasing of stone porosity often is related to decreasing of mechanical properties, which is probably the reason why this specific quarry in Tierrabomba Island was abandoned for new ones in the same area at the middle of the XVII century (Álvarez-Carrascal 2018; Cabrera et al. 1995).

Figure 1: Thin Section structure and Quarry Stones, (Saba et al. 2019).

Table 1: Petrographycal analysis results.

From the quarry stone physical-mechanical characterization, can be highlighted that open porosity differs about 10 % from thin section analysis, which means that thin section measurements despite their low representativeness provide an acceptable approximation to porosity values compared with results coming from a large number of samples analysed whit gravimetrical measurements.

Structure and quarry stone are found as pure limestone from the X-ray analysis, with CaCO₃ higher than 98 %, and a presence of Quartz and Halite lower than 1 %, (Figure 2).

Figure 2: Structure samples X-Ray results.

Ion chromatography analysis on structure samples show a total range of salts between 0.4 % and 2.4 % of Mass on average 1.0 % in all 5 structure samples (Table 2). From the literature review it is difficult to assess if this is a high or low salt content because it should be compared with samples taken at the same depth.

Salts are usually found with higher concentration 128on the first millimetres of the material surface, while lower concentrations correspond to deeper sampling. In our case sampling depth has been 5 mm from the surface. Nevertheless, sampling depth in literature works usually it is not adequately reported (Ahmad & Haris Fadzilah 2010; C, Lopez-Arce et al. 2016). Hence, a comparison is risky and not appropriate (Figure 3).

Table 2: Stone Physical-mechanical Characteristics.

V_R is Real Volume P_o Open Porosity, ρ_b Apparent Density, m_d is the Dry mass, m_s Saturated mass, m_h water immersion mass, SUCS is Stone Uniaxial Compressive strength

Figure 3: Structure samples Ion Chromatography results elaborated with Runsalt^®.

Salt crystallization process is not directly observed in the structure surface of the area of study, nevertheless, a careful revision of the whole structure should be done to eventually detect this type of deterioration and compare salt concentrations with generic sections of the structure and define if ion chromatograph results are effectively reliable to correctly detect the deterioration process in this case of study.

Results were further analysed with Runsalt software (V. 1.9), (Bionda 2005). It is the graphical user interface to the ECOS thermodynamic model (Price 2000). It allows to define which salts are crystallizing from a mixture depending on temperature and RH conditions. The ion contents, as determined by IC, are input for the model. As an example, sample 5 ion and cation values were introduced as a input, as well as the average temperature in Cartagena, 27.7 °C according to DIMAR (2018).

Table 3: Ion Chromatography results (%), input data of Runsalt. N.= Number of Sample.

Results are reported in Figure 3 and are similar for the whole structure samples. The mix of salts on the samples is identified and Relative Humidity of equilibrium (RH_eq) is reported for each salt in the solution. It is worth to mention that Cartagena stands in a tropical area, with a Relative Humidity range between 72 % to 90 % during the year, considering minimum and maximum values of daily and seasonal variations. From these preliminary studies, Thenardite and Bloedite could crystalize having a RH_eq of 75 %, although their role on the material deterioration should be further studied, for example pondering whether the number of crystallization cycles per year is relevant.

Regarding mechanical properties, SUCS varies between 1.5 MPa to 2.2 MPa which is typical of a soft pure limestone (Table 3).