INTRODUCTION

Calcium-carbonate materials are widely used in the construction and production of works of art. Under natural conditions, limestone and marble are destroyed due to weathering, temperature fluctuations, the formation of salt and ice crystals in the pore space, exposure to acids of bacterial and fungal origin, etc. [1]. Additional factors should be considered for cultural heritage sites, namely, microclimate change due to deforestation, the construction or destruction of buildings near the site; direct impact from visiting tourists; the use of abrasive or corrosive agents when cleaning stones; and the use of materials incompatible with limestone, such as cement mixtures, for restoration [2].

After a while, additional problems can appear when conventional approaches based on the use of cement mixtures or acrylic, vinyl, and silicone polymers (since the 1960s) are used to reinforce limestone objects. Due to a higher strength and lower water permeability, cement promotes the destruction of neighboring limestone blocks, which can accelerate the destruction of limestone by 38 times compared with the state before restoration [2]. After 5–40 years, a polymer coating causes an increase in surface hydrophobicity, violation of the water permeability of the surface layer of limestone, and the crystallization of salts under the polymer film, which leads to its peeling and often irreversible destruction of the restored object [3]. At the same time, the coatings can biodegrade leading to destruction of the protective coating [4].

The formation of salt or ice crystals in a pore space is the most dangerous mechanism of limestone destruction. Such salts as sodium sulfate and chloride are the most dangerous for limestone [5]. For example, upon the dissolution of thenardite (Na2SO4) and subsequent formation of heptahydrate (Na2SO4· 7H2O) and mirabilite (Na2SO4·10H2O) crystals, the crystallization pressure in the pore space can exceed 10 MPa [6]. The tensile strength of limestone is about 2.25 MPa [7]. Water freezing and ice forming in the pore space can increase the pressure to 200 MPa which is also higher than the tensile strength of limestone [8].

Thus, it is necessary to develop approaches to the preservation of objects made of calcium carbonate which would increase the salt resistance and the strength characteristics of limestone retaining its properties such as moisture permeability, color, etc. The use of nanoscale and microscale inorganic coatings is the most promising way to protect cultural heritage sites [9]. Hypothetically, such coatings should not prevent the process of moisture transfer inside limestone and, at the same time, should increase the strength and resistance of limestone to the effects of destructive factors. Chemical and biological approaches are used to obtain such coatings. Chemical approaches are the following: the use of a suspension of calcium-hydroxide nanoparticles which promotes the formation of calcium-carbonate particles binding the surface layer of limestone under the influence of atmospheric factors [10, 11]; the use of a diammonium-phosphate solution which forms a hydroxyapatite film on the surface of limestone [12] or silicon-dioxide nanoparticles [13].

Biological approaches are based on the use of microorganisms capable of forming calcium carbonate during growth [14]. In this case, specially selected and grown cultures of microorganisms with carbonate-forming activity can be used [15] or the stimulation of microorganisms existing in the surface layers of limestone can be activated using solutions that encourage the formation of calcium carbonate by the microorganisms [16]. The first approach has such advances as a high number and high activity of microorganisms used in the treatment, but a significant amount of biomass of microorganisms needs to be prepared and transferred to the place of treatment in compliance with sterility and temperature conditions. The benefit of the second approach is the possibility of applying solutions prepared on site which reduces sterility and temperature requirements. However, the species composition of the microorganism community depends on the treated surface which requires a preliminary verification of the carbonate-formation potential.

This work is aimed at studying the possibility of biogenic mineral formation to reinforce the ruins of a medieval town on the Eski-Kermen plateau (last quarter of the 6th–14th centuries) located on the Crimean Peninsula, Russia. The town whose historical name was lost was severely damaged by fire during the raid of Nogai’s troops in 1299. The ruins of the town are a site of cultural heritage of federal significance. During the experiments, limestone samples were collected from the masonry walls of the main basilica located on the territory of the fortress, the community of carbonate-forming microorganisms was stimulated under laboratory conditions, and the limestone samples were studied using electron microscopy, microtomography, and methods for determining the specific surface area, strength characteristics, water permeability, and salt resistances.

EXPERIMENTAL

Description of the Town on the Eski-Kermen Plateau

So-called cave towns located on the Inner Ridge of the Crimean Mountains are of particular importance in studying the medieval history of Crimea. Most of them emerged due to the military and political activity of Byzantium in the second half of the 6th century: Byzantine engineers built several fortresses for the allies of the Goth–Alan empire. These fortresses simultaneously protected the approaches to Chersonesus, the main Byzantine center on the peninsula. Over time, the fortresses were transformed into small Byzantine towns which became the administrative, economic, and religious centers of the surrounding territories (archontia).

A special place among the named monuments is occupied by the town, the remains of which have been preserved on the Eski-Kermen plateau (“old fortress” if translated from the Crimean Tatar language), located in the Bakhchysarai district, 6 km south of the village Krasnyi Mak (Fig. 1). The historical name of the city has not been preserved. The fortress which later turned into a town was built on a flat top surrounded by deep ravines of a mesa limestone mountain. An important role in the defensive system was played by cave towers and casemates, carved into the rocky massif of promontories protruding from the eastern and western faces of the mountain. Through the loopholes cut in the towers, the space between them and the foot of the plateau was controlled. Powerful double-shell defensive walls were constructed between many cave towers along the edge of the plateau [17].

Fig. 1.
figure 1

Eski-Kermen plateau on top of which the ruins of a medieval town have been preserved. Aerial photography, 2007. View from the north. Photo by E.A. Khairedinova.

Full size image

Simultaneously with the defensive fortifications, two large longitudinal and several transverse streets were laid on the plateau, and a three-aisled basilica (the main temple of the city) was built in the center. In large-scale construction work, a large amount of local limestone was used since it was the most affordable and highest quality building material. Blocks and rubble for buildings were broken loose in the vicinity of the town, in the ravines to the south, and directly on the plateau at the foot of the western defensive wall.

Based upon the building remains preserved on the surface, in the late period of the existence of the town (841–the end of the 13th century), the part of the plateau inhabited in the preceding time was densely built up. Almost the entire territory of the southern half of the plateau was occupied by rectangular blocks built on both sides of the main and parallel streets. Narrower lanes were laid between the blocks at right angles to the main street. After the destruction of most of the city in a fire during the raid of Nogai’s troops in 1299, its ruins remained untouched almost until modern times. Thanks to this, it is possible to trace the general layout of urban development on the plateau, identify topographic features, and study the structure of residential areas and temple complexes.

Conservation is of great importance for monument preservation. Nummulitic limestone used for building is slowly destroyed under the influence of the environment. The remains of buildings damaged with fire are particularly susceptible to destruction. The process of destruction of limestone masonry can be clearly seen in the example of the main basilica of the town discovered in 1930. The temple walls were preserved to a height of 1.0–1.5 m. Stone blocks from which the apses were built had a well-polished smooth surface. Based upon photographs of the monument taken in 1933 (3 years after excavation), the stone began to crack and delaminate (Fig. 2) [18, 19]. The monument was significantly damaged by restoration work in the 1980s when modern cement-based mortars were used to strengthen the masonry. To date, the masonry of the basilica has been almost completely destroyed (Fig. 3). For the experiments, limestone samples were collected from the broken masonry of the basilica.

Fig. 2.
figure 2

Photo of the basilica in the 1930s: (a, b) the basilica immediately after excavations in 1930 [18], (c) the basilica in 1933 [19].

Full size image
Fig. 3.
figure 3

Current state of the basilica: (a) general view from the west (2021), (b) general view from above (2022). Photo by E.A. Khairedinova.

Full size image

Preparation of the Limestone Samples

Small cube-shaped pieces of limestone were sawn from the masonry fragments and placed into media to stimulate the growth of microorganisms in distilled water. The limestone samples were not sterilized except for those used to determine the strength characteristics. Reference samples were sterilized at 140°C for 60 min using a dry heat oven. All manipulations were performed on a sterile surface using instruments sterilized with ethanol. All experiments were carried out in at least three replicates.

X-ray Tomography of the Internal Structure

The internal structure of the limestone samples without media treatment was visualized using an X5000 (North Star Imaging, USA) industrial CT X‑ray inspection system with an open tube. Shadow projections were recorded using a position-sensitive X-ray detector (Perkin Elmer, USA) with a matrix size of 2048 × 2048 pixels, a pixel size of 200 × 200 μm, and a dynamic range of 16 bits. A scintillator based on CsI : Tl was used. The measurement parameters for sample 1 were the following: tube voltage of 120 kV, current of 150 μA, a focal-spot size of 18 μm, an exposure time per frame of 0.17 s, a gain of 0.5 pF, and an angular step of rotation around the vertical axis of 0.18°. Measurements of samples 2 and 3 were carried out under the following conditions: tube voltage of 130 kV, current of 150 μA, a focal spot size of 19.5 μm, an exposure time per frame of 0.25 s, a gain of 0.5 pF, and an angular step of 0.18°. Sample 4 was measured with the parameters: tube voltage of 130 kV, current of 150 μA, a focal spot size of 19.5 μm, an exposure time per frame of 0.17 s, a gain of 0.5 pF, and an angular step of 0.18°. The pixel size of the tomograms for all samples was 9 × 9 μm. The efX-CT software package was used to reconstruct the tomographic sections. Data visualization and calculation of the volume, surface area, and porosity were carried out using Volume Graphics studio 3.5.1.

Influence of Treatment on the Salt Resistance of Limestone

To determine the effect of limestone treatment on the resistance to high salt concentrations, cube-shaped limestone samples with a side-edge length of about 3 cm were made. Then, the samples were kept in the following media for 10 days:

(i) 0.46 g/L yeast extract, 1.38 g/L peptone, 5.63 g/L CaCl2, 10 g/L urea, 1 mL/L trace elements solution [20], 1.6 mg/L NiCl2·2H2O (hereinafter medium 1) [21];

(ii) 1 g/L sodium acetate, 1 g/L yeast extract, 6 g/L CaCl2, 1 mL/L trace elements solution [20].

Limestone samples incubated in distilled water were used as reference samples. After incubation, all samples were rinsed with distilled water and dried at 40°C to constant weight. Then, all samples were subjected to salt-attack cycles. Each cycle consisted of immersing into 1 M Na2SO4 solution for 2 h and drying at 70°C for 24 h; then, each sample was weighed. The experiment included 13 cycles.

Effect of Treatment on the Water Absorption of Limestone

To determine the effect of limestone treatment on water absorption, Crimean limestone samples in the form of a cube with a side edge length of about 3 cm were used. Only one face of the samples was treated. The sample was immersed into a container with the medium to a depth of 5 mm. Treatment was carried out for 7 days with different types of media:

(i) 0.46 g/L yeast extract, 1.38 g/L peptone, 5.63 g/L CaCl2, 10 g/L urea, 1 mL/L trace elements solution [20], 1.6 mg/L NiCl2 2H2O (medium 1);

(ii) 3.75 g/L glycine, 0.75 g/L NaOH, 0.5 g/L yeast extract, 1 g/L sodium acetate, 2.7 g/L CaCl2;

(iii) 1 g/L sodium acetate, 1 g/L yeast extract, 6 g/L CaCl2, 1 mL/L trace elements solution [20];

(iv) 1 g/L calcium acetate, 1 g/L yeast extract, 1 mL/L trace elements solution [20];

(v) 5 g/L glucose, 4 g/L yeast extract, 2.5 g/L calcium acetate, 1 mL/L trace elements solution [20];

(vi) 0.46 g/L yeast extract, 1.38 g/L peptone, 5.63 g/L CaCl2, 10 g/L urea, 1 mL/L trace elements solution [20];

(vii) deionized water (MilliQ).

The water absorption of each sample was measured before and after treatment. The treated face of the sample was placed on a layer of paper towels saturated with water, after which the change in the weight of the sample was measured every minute for 15 min, and then after 30, 60, 90, 120, 180, 240, and 300 min. The value of water absorption per unit area was calculated by the formula:

$$Q{\text{ }} = \frac{{{{m}_{t}} - {{m}_{0}}}}{s},$$

where mt is the sample weight at time t, m0 is the weight of the dry sample, s is the sample face area, and t is the time (min). The plot of dependence of Q on t1/2 was built using the obtained data [22].

Effect of Treatment on the Strength Characteristics of Limestone

To determine the effect of treatment on the strength characteristics, limestone parallelepiped samples were used; the length of the lateral ribs was no more than 1 cm. Two opposite upper and lower faces were parallel for correct strength measurement. At least five repetitions were used for each type of treatment. The samples were completely immersed in a medium 1 for 7 days; the media were replaced with fresh ones on the fourth day of incubation. Three variants of processing were used: a sterile sample with the medium and microorganisms obtained by incubating Crimean limestone in medium 1 for a day at an average stirring speed of 140 rpm; a reference sterile sample with sterile medium; and a reference sterile sample with distilled water.

The mechanical properties were studied using an Instron 5965 (Instron, United Kingdom) universal testing system equipped with load cell ±5 kN in the uniaxial compression mode at a constant rate of 0.05 mm/min. Samples of air-dry limestone with a height of 7.62 mm and a cross-sectional area of 1 cm2 were used for the tests. The statistical sampling was five specimens.

Effect of Treatment on the Specific Surface Area of Limestone

To determine the specific surface area of limestone, spherical limestone samples with a diameter of about 1 cm were prepared. The samples were processed for 7 days using medium 1. The samples were inoculated using enrichment culture obtained from scrapings from the limestone surface on the medium of the abovementioned composition and grown for one day at room temperature.

The specific surface area of the limestone surface was determined by the Brunauer–Emmett–Teller (BET) method before and after treatment with the culture medium. The measurements were performed using an Autosorb iQ (Quantachrome Instruments, USA) specific surface area and porosity analyzer by BET processing of the nitrogen vapor adsorption isotherm at liquid-nitrogen temperature 77.35 K in the range of relative pressures p/p0 from 0.05 to 0.3. The range of pore-diameter indication of the analyzer is from 0.35 to 400 nm. Before each measurement, the samples were degassed at 150°C for 12 h under high vacuum.

SEM Study of the Surface of the Limestone Samples

The morphology of the samples before and after processing was studied using a Versa 3D (Thermo Fisher Scientific, USA) scanning focused-ion-beam microscope under low vacuum conditions (30 Pa) at an accelerating voltage of 5 kV using a circular backscatter detector (CBS). The samples were treated for 7 days using medium 1 and distilled water.

RESULTS

Visualization of the Internal Structure of the Limestone Samples

X-ray tomography was used to visualize the internal structure of samples 1–4 before processing with solutions. The samples were inhomogeneous and contain many coarse organic inclusions (Fig. 4). Table 1 lists data on the surface area, volume, and specific surface area.

Fig. 4.
figure 4

Tomographic sections of limestone samples 1–4 (a–d).

Full size image
Table 1. Parameters of samples 1–4 determined by X-ray tomography
Full size table

Salt Resistance

Treating the limestone samples with media stimulating the natural limestone community led to an increase in limestone resistance to the impact of high salt concentrations. The reference samples of untreated limestone withstood only 7 cycles of treatment with 1 M sodium sulfate solution and were completely destroyed during the eighth cycle. The samples treated with a medium containing urea or sodium acetate withstood 13 cycles of the experiment despite losing 80% of its mass (Fig. 5). Thus, treatment with media increased salt resistance by 86%.

Fig. 5.
figure 5

Limestone resistance to high concentrations of sodium sulfate. When the sample without treatment reaches a value of 100% it means complete destruction of the sample; (1) sample without treatment, (2) treated with sodium-acetate medium, (3) treated with urea-based medium with the addition of nickel.

Full size image

Limestone Water Absorption

The capillary water absorption of limestone insignificantly decreased after treatment of the limestone samples with different media which indicates that the parameters of limestone moisture permeability were retained (Figs. 6, 7).

Fig. 6.
figure 6

Capillary water absorption of Eski-Kermen limestone: (1) limestone sample treated with a urea-nickel medium and (2) limestone sample before treatment with urea-nickel media.

Full size image
Fig. 7.
figure 7

Capillary water absorption of Eski-Kermen limestone treated with media containing the following components: (1) urea with nickel; (2) sterile MilliQ; (3) glucose, sodium acetate, glycine; (4) calcium acetate; (5) urea without nickel.

Full size image

Limestone Strength

Limestone treatment improved its strength characteristics. The strength of the sample treated with distilled water was 12.3 ± 2.8 MPa, while the strength of the sample treated with a microorganism-free medium was 13.6 ± 0.6 MPa. The samples treated with a medium based on urea possessed the highest strength of 15.8 ± 2.6 MPa. Thus, stimulation of the limestone microbial community increased the average strength value by 28% compared with the reference sample treated with distilled water (Fig. 8). When sterilized samples were treated with a sterile medium, their strength increased by 11% compared with the reference sample treated with distilled water. The process of changing the compressive stress for samples treated with medium 1 with microorganisms for five replicates is shown in Fig. 9.

Fig. 8.
figure 8

Strength of distilled water, limestone treated with urea-based medium and urea-based medium with microorganisms.

Full size image
Fig. 9.
figure 9

Deformation curves obtained under the uniaxial compression of samples 1–5 treated with a urea-based medium containing nickel and microorganisms.

Full size image

Specific Surface Area

The specific surface area of the limestone sample before treatment was in the range from 2.08 to 3.55 m2/g, and the average value was 2.93 ± 0.55 m2/g. After treatment with microorganisms, the specific surface area increased by 42% to 4.16 ± 0.53 m2/g. This could be caused by an increase in the roughness of the inner pore surface due to newly formed calcite that appeared during treatment with microorganisms, which leads to an increase in the specific surface area of the sample.

Scanning Electron Microscopy of the Limestone Surface

According to the results of scanning electron microscopy, there was a uniformly distributed layer of small crystallites with a size from 0.4 to 1.3 µm (Figs. 10a–10d) on the surface of the treated samples. No significant changes in the surface morphology were observed for the samples treated with distilled water (Figs. 10e, 10f).

Fig. 10.
figure 10

SEM image of the sample surface before (a, c) and after (b, d) treatment with medium 1 containing bacteria; before (e) and after (f) treatment with distilled water.

Full size image

DISCUSSION

One of the main principles of the scientific restoration and conservation of monuments of historical and cultural heritage is minimum interference in the historical material of the monument with its maximum preservation [23]. To achieve these goals, the modern approach to conservation and restoration reflected in the Venice Charter allows the use of “modern conservation and construction technologies, the effectiveness of which is confirmed by scientific data and guaranteed by experience,” which emphasizes the importance of developing scientifically-based methods for the conservation of monuments [24].

The application of means of preservation and restoration without sufficient scientific justification results in damage to many heritage sites. The medieval fortress Eski-Kermen located in Crimea, Russia was no exception. The use of cement mortars in the 1980s to strengthen the masonry of the main basilica of the town led to its almost complete destruction. The accumulated experience made it clear that the use of concrete mixtures negatively affects the state of limestone structures: “there is an inherent and fatal incompatibility between cement and lime-mortared construction” [25]. For example, in 1991, the destruction of the limestone of the Church of Our Lady (Breda, the Netherlands) was found to be caused by the use of cement during restoration in the period from 1910 to 1921 [26]. In 2012, during the study of House of the Turtles (Uxmal, Yucatan, Mexico), it was found that the replacement of wooden door lintels with concrete ones between 1969 and 1972 led to acceleration of the destruction of limestone blocks in contact with concrete by 38 times compared with nonrepaired areas [2].

The application of modern methods based on the creation of nanoscale and microscale coatings mostly corresponds to the principle of minimal interference with historical material. Biogenic mineral formation using microorganisms capable of forming calcium carbonate is being actively tested for the conservation and restoration of heritage sites. For example, in [27], stones of the monastery in Saint Jeronimo (Granada, Spain) were first covered with a solution with microorganisms; then, these stones were covered with a nutrient solution for the growth of microorganisms for 6 days twice a day. Finally, a protective coating was obtained with a crystal size from 30 to 100 nm, a thickness up to 0.5 mm, and a penetration depth to 3–5 mm. In work [28], the water absorption of facade stones of the Angers Cathedral (Angera, Italy) was reduced by 16.7% by treatment with a fraction of the bacterial cell wall. The application of carbonate-forming microorganisms allows one to increase the compression strength of building materials up to 25% [29], reduce their water absorption up to 90%, and enhance degradation resistance during freeze-thaw cycles [30].

According to the results of laboratory tests of stimulating biogenic mineral formation on limestone masonry samples of the medieval fortress Eski-Kermen, a layer of crystallites with a size from 0.4 to 1.3 μm is formed on the visible surfaces of limestone. At the same time, the limestone strength increased by 28% from 12.3 ± 2.8 to 15.8 ± 2.6 MPa. This value is comparable with results previously obtained on samples from a limestone quarry in Crimea, the compression strength of which in the dry state was increased by 23% after treatment with stone-reinforcing material based on silicic acids [31]. It should be noted that the studied samples have a nonuniform internal structure and contain a large amount of coarse organic inclusions. It might affect the strength of limestone which, according to the classification, corresponds to the “very low strength” class (strength range from 0 to 25 MPa) [32]. Another important indicator of the efficiency of treatment is that the resistance of limestone to the effects of high salt concentrations increased by almost two times; it increased from seven exposure cycles for untreated limestone to at least 13 exposure cycles for treated limestone. Moreover, the specific surface area increased by 42% from 2.93 ± 0.55 to 4.16 ± 0.53 m2/g which could be explained by the growth of surface roughness of the pores during processing [33]. The capillary water absorption of the limestone remained almost the same during treatment which indicates potential preservation of the intensity of moisture transport in limestone.

CONCLUSIONS

The preservation of cultural heritage sites requires detailed investigation of the material, environmental factors, existing damage, and microbiological effects on the structure. The combination of all these factors determines the success and longevity of measures for the conservation and restoration of a site. We carried out laboratory testing of the stimulation of biogenic carbonate formation on samples of limestone masonry from a medieval town on the Eski-Kermen plateau (Crimea, Russia). As a result, a layer of crystallites with a size of 0.4 to 1.3 μm was formed on the surface of limestone, the strength and resistance to salt attack was increased by 28 and 86%, respectively, and the level of capillary water absorption was retained. The obtained results show the potential of using biogenic mineral formation for the conservation and restoration of medieval limestone masonry of the city on the Eski-Kermen plateau. In the future, it is necessary to experimentally test the method under field conditions and evaluate the possibility of combining the biological approach with other types of processing.