Archaeological chert artifacts from Atapuerca sites (Burgos, Spain): characterization, causes of decay and selection of compatible consolidating products

20 ARTÍCULO / ARTIGO / ARTICLE CONSERVAR PATRIMÓNIO 36 (2021) 20-35 · https://doi.org/10.14568/cp2019037 · ISSN 2182-9942 ARP · Associação Profissional de Conservadores-Restauradores de Portugal · https://conservarpatrimonio.pt AINARA ZORNOZA-INDART 1 * PAULA LOPEZ-ARCE 2,3 LUCÍA LÓPEZ-POLÍN 4 1. Department of Painting, Faculty of Fine Arts, University of the of the Basque Country (UPV/EHU), Lejona, Spain 2. University College London (UCL), Institute for Environmental Design and Engineering (IEDE), The Bartlett, School of Environment, Energy and Resources, Faculty of the Built Environment, London, United Kingdom 3. Property Care Association, Cambridgeshire, United Kingdom 4. Institut Català de Paleoecologia Humana i Evolució Social (IPHES), Zona Educacional, Tarragona, Spain * [email protected] Archaeological chert artifacts from Atapuerca sites (Burgos, Spain): characterization, causes of decay and selection of compatible consolidating products Artefatos arqueológicos de cherte de Atapuerca (Burgos, Espanha): caracterização, causas de deterioração e seleção de produtos de consolidação compatíveis Abstract Chert tools from Galería and Gran Dolina Caves, located in the Sierra de Atapuerca site complex (Burgos, Spain), were characterized (macro-visual inspection, mineralogical phases, degree of crystallinity, soluble salts, surface morphology and optical surface roughness) and compared to chert samples collected from the surrounding Atapuerca mountain range. The chert tools were studied to determine their causes of decay and for selecting the most compatible consolidation treatments. It was found that samples solely containing quartz were not significantly altered and required little conservation treatment compared to those that contained quartz and moganite, which were more weathered and powdery, requiring consolidation. The efficacy of the consolidating products traditionally used by conservators (acrylic resin and ethyl silicate) to preserve these chert remains, together with novel nanoparticle-based products (SiO2 and a mixture of SiO2 and Ca(OH)2 nanoparticles) were assessed in this study. Changes produced by these consolidating products in the physical (surface morphology and cohesion) and aesthetic properties of the chert tools were evaluated using non-destructive techniques (peeling test, spectrophotometry and optical surface roughness), followed by destructive techniques, such as SEM and XRD. Resumo Os utensílios de cherte das grutas Galería e Gran Dolina, localizadas no complexo da Sierra de Atapuerca (Burgos, Espanha), foram caracterizados (inspeção macro-visual, fases mineralógicas, grau de cristalinidade, sais solúveis, morfologia da superfície e rugosidade óptica da superfície) e comparadas com amostras de cherte recolhidas na cordilheira de Atapuerca. Os utensílios foram estudados para determinar as causas de deterioração e selecionar os tratamentos de consolidação mais compatíveis. Verificou-se que as amostras contendo exclusivamente quartzo não foram significativamente alteradas e exigiram pouco tratamento de conservação em comparação com as que continham quartzo e moganite, que estavam mais alteradas e pulverulentas, exigindo consolidação. Neste estudo foi avaliada a eficácia dos produtos de consolidação tradicionalmente usados pelos conservadores (resina acrílica e silicato de etila) para preservar os restos de cherte, juntamente com novos produtos baseados em nanopartículas (SiO2 e uma mistura de nanopartículas de SiO2 e Ca(OH)2). As alterações produzidas por estes produtos consolidantes nas propriedades físicas (morfologia e coesão da superfície) e estéticas foram avaliadas por técnicas não destrutivas (teste de descamação, espectrofotometria e rugosidade da superfície óptica), e destrutivas, como SEM e DRX. PALAVRAS-CHAVE Artefatos arqueológicos de cherte Consolidação Nanossílica Nanopartículas de Ca(OH)2 Resina acrílica Silicato de etilo KEYWORDS Archaeological chert artifacts Consolidation Nanosilica Nanolime Acrylic resin Ethyl silicate ARTICLE / ARTIGO 21 CONSERVAR PATRIMÓNIO 36 (2021) A. Zornoza-Indart, P. Lopez-Arce, L. López-Polín Introduction Chert is a sedimentary cryptocrystalline siliceous rock primarily composed of quartz (SiO2) and other silicate minerals like moganite, opal, or clay or carbonates and organic material, which are considered impurities [1]. The intrinsic properties of chert, such as hardness and conchoidal fracture, had made it widely used as a raw material for knapping throughout the history of mankind. Although it is a durable material, chert artifacts are often found in poor conservation conditions in archaeological sites such as Sierra de Atapuerca, limiting their study and requiring consolidation treatments to enable examination and conservation. Sierra de Atapuerca (Burgos, Spain) is a limestone karst complex with many cavities in which archaeological fieldwork has been systematically carried out since 1978 [2]. The archaeological site has provided crucial data on human evolution in Eurasia from 1.2 Myr to the end of the Middle Pleistocene [4]. Therefore, it was included in the UNESCO’s World Heritage list in 2000. Fieldwork extended throughout the cave system and around the range, including open-air and rock shelter sites, has increased the time period up to Holocene occupations [6]. The chert samples included in this study come from Gran Dolina and Galería (Figure 1a), two of the cave sites located in the Trinchera del Ferrocarril. The stone tools recorded at Galería and Gran Dolina Caves comprise various raw materials such as sandstone, quartzite, and other less wellrepresented materials, but the majority are chert artifacts, particularly Neogene chert that is usually highly altered. It is formed after silica precipitation in a hypersaline environment rich in gypsum and carbonates. It appears in the Neogene border of the Duero River and belongs to the late Miocene [10]. Almost all the Neogene chert artifacts from the Atapuerca sites have been whitened and lost their original luster. They also present visibly increased surface porosity, with an aspect that corresponds to the commonly described “white patina” [11]. The alteration, however, does not concentrate on the surface; in fact, it is more pronounced in the inner part, where usually the chert becomes powder [14]. In some cases, this issue impedes the recovery of intact artifacts from the site, as they often break during the excavation works and, on many occasions, require consolidation before being lifted or subsequently handled (Figure 1b-d). Up to now, the chert artifacts of the Atapuerca sites have predominantly been consolidated with the acrylic resin Paraloid B72 and, occasionally, with ethyl silicate-based products [14]. Figure 1. Chert samples collected from Sierra de Atapuerca (Burgos, Spain) archaeological site: a) Upper part of Gran Dolina site; b) Prior consolidation with acrylic resin Paraloid B72 of an altered chert artifact found in the TD10.2 unit; c) Lifting of the treated artifact; d) Chert stone tool after conservation treatments. a bcd A. Zornoza-Indart, P. Lopez-Arce, L. López-Polín 22 CONSERVAR PATRIMÓNIO 36 (2021) Archaeological chert artifacts from Atapuerca sites (Burgos, Spain) The aim of this research is to characterize chert samples collected from the archaeological Atapuerca site to determine their composition, texture and conservation state, as well as the causes of their decay in order to select the most compatible consolidation treatments. In addition, we assessed the efficacy of the most widely-used consolidation treatments by comparing changes on the aesthetic and physical properties of the treated substrates with those produced by two novel consolidation products based on SiO2 and Ca(OH)2 nanoparticles. These products are still not widely used by conservators for conservation purposes and unreported for archaeological prehistoric stone tool preservation thus far. Materials and methods Chert samples In the Atapuerca archaeological site, the chert is divided into two petrological groups based on its origin and geological period of formation, i.e., Cretaceous and Neogene [10, 16]. Neogene chert comes from Late Miocene formations and is found as large blocks outcropping in marls and marly limestone, and was formed due to silica precipitation in a hypersaline environment rich in gypsum and carbonates [10]. Previous studies have determined that this type of chert is composed of quartz and moganite; it also contains a certain amount of gypsum, calcite (filling some pores), and occasional impurities. Microscopically it appears highly heterogeneous and particularly porous in crystalline zones [17]. Six chert samples classified as having different degrees of decay based on visual inspection were selected to determine their composition, texture and state of conservation (Table 1). The causes of their decay were also studied with the purpose of being able to select and evaluate the necessary conservation treatments and the most suitable consolidating products for preserving the samples. Four samples of chert lithic remains came from the Middle Pleistocene units of the site. Two of them (samples SX3 and SX9) were collected from unit GIII of Galería site, dated at around 300 ky [18] and another two (samples SX1 and SX17) were collected from the TD10.2 unit in the upper part of Gran Dolina and dated at around 400 ky [10, 20]. Two further, fresh chert samples, were collected from the surrounding outcrops in the Sierra de Atapuerca (samples SX18 and SX19) with the aim of comparing these with the decayed samples collected from the archaeological sites (Figure 2). The soil remains that covered the samples from the dig were carefully removed from their surfaces with a brush, in order to carry out the compositional and textural analyses of the clean cherts. Study of consolidation treatments Sixteen chert samples from Middle Pleistocene units were Nomenclature Nomenclature on site Year Level Square Nº Middle Pleistocene lithic remains SX1 2008 TD10 M18 341 SX3 2004 GIII N03 71 SX9 2004 GIII N03 70 SX17 2011 TD10 L13 66 Fresh chert samples collected from the surrounding outcrops SX18 SX19 Table 1. Studied chert samples from Atapuerca archaeological site. Figure 2. Studied chert samples: a) Sample SX17; b) Sample SX9; c) Sample SX3; d) Sample SX1, from the archaeological site of Atapuerca; e) Sample SX19, and f) Sample SX18 from the surrounding outcrops. a c b f e d 23 CONSERVAR PATRIMÓNIO 36 (2021) collected in two different sites to evaluate the different consolidating products after their treatment. Six samples were selected from the TD10.2 unit in the upper part of Gran Dolina and ten samples were collected from unit GIII at the Galería site. The samples were large enough to be analyzed using various analytical techniques and test methods (Table 2). The 16 selected samples (four control samples and 12 samples to be treated, three samples for each consolidating product) initially displayed a similar degradation state, i.e. milky white in color, loss of density and mass, superficial decohesion, large roughness, friable surface and loss of external areas (apart from samples SX2 and SX8 which preserved some parts of the most superficial area). As explained above, the soil covering the samples was brushed off to enable their characterization and further application of the consolidating products. As it was described in a previous work by ZornozaIndart et al. [21], four types of consolidating products were applied to consolidate chert artifacts (Table 2). The first one comprises the most widely-used conventional alkoxysilane consolidant (ethyl silicate: Tegovakon V100, Evonik Industries AG). This product was chosen because it is highly compatible with siliceous substrates and can form chemical bonds with substrates containing hydroxyl groups during the sol-gel process [22]. The second product was a colloidal dispersion of silica nanoparticles in water (Nano Estel, C.T.S.) with a particle size of 10-20 nm and a concentration of 150 g/L. This product was chosen because it is a novel product and is also expected to be chemically compatible with siliceous substrates due to the formation of a silica gel with silanol groups [23]. The third product was an acrylic resin, the most widelyused by conservators (Paraloid B72 by Dow Chemical), in a concentration of 5 % in xylene. The choice of solvent is based on its volatility. When Paraloid is used as a consolidant, a low volatile solvent is used to allow maximum penetration of the product (conservators use to apply acetone as a solvent when the product is used as an adhesive because the rapid evaporation rate of the acetone decreases the penetration of the product in the porous network remaining only on the surface). Although Paraloid is the most used consolidant, silicate-based products with a lesser penetration depth than ethyl has been reported [24]. Finally, a mixture of two inorganic products based on colloidal dispersions of nanoparticles was used (the aforementioned Nano Estel, and Nanorestore C.T.S., i.e. Ca(OH)2 colloidal dispersion in isopropyl alcohol with a particle size of ≤ 100 nm and a concentration of 5 g/L), in a ratio of 1:1. The products were applied to reproduce the application method used nowadays by most restorers, which usually follow the recommendations stated by the manufacturers. The application was performed through a capillary tube and deposited drop-by-drop (the most frequently used application method in conservation of archaeological artifacts) onto the surface of the samples. Then, the samples were exposed to a humid environment (19 ± 1 oC temperature (T) and 93 ± 3 % relative humidity (RH)) for one month, even though it should be noted that the time suggested by the manufacturer (2-4 weeks) might not be time enough to complete the polymerization process. This high RH was selected because the carbonation rate of Ca(OH)2 nanoparticles is faster under higher RH [25]. In addition, new silica nanoparticles can be used in conditions where ethyl silicate or acrylic resin are not appropriate, such as in the presence of high levels of humidity, free water or wet Table 2. Chert samples and applied consolidating products by dripping. Nomenclature Nomenclature on site Applied consolidating product Year Level Square Nº Consolidating product Commercial product Applied amount of product (g) MC SX1 2008 TD10 J19 85 Control samples MC SX2 2004 GIII N03 71 MC SX3 2004 GIII N03 71 MC SX4 2011 TD10 K14 151 SX1 2008 TD10 M18 341 Nano SiO2Nano Estel 1.09 SX2 2008 TD10 J19 85 2.29 SX3 2004 GIII N03 71 1.2 SX4 2004 GIII N03 71 Ethyl silicate Tegovakon V100 2.23 SX6 2004 GIII N03 71 0.59 SX7 2004 GIII N03 71 1.41 SX8 1998 TD10 N14 34 Acrylic resin Paraloid B72 2.11 SX9 2004 GIII N03 70 1.11 SX13 1998 TD10 N14 34 1.53 SX10 2004 GIII N03 70 Nano SiO2 + Ca(OH)2Nano Estel + Nanorestore 1.54 SX11 2004 GIII N03 70 1.33 SX12 2004 GIII N03 70 1.22 A. Zornoza-Indart, P. Lopez-Arce, L. López-Polín 24 CONSERVAR PATRIMÓNIO 36 (2021) surfaces. Note that moisture is common in archaeological excavated objects, which can generate compatibility problems with conservation products. The temporary hydrophobicity of ethyl silicates, already investigated by other authors, can preclude future treatments, as the consolidation treatment is usually followed by cleaning of the samples. In this deeper cleaning procedure, hydric solvents are used, but, the cleaning is not possible if there is a hydrophobic layer. In addition, if a volumetric reintegration is made and the surface is hydrophobic, the adhesion of the reintegration stuccos is hinder. Also, the hydrophobicity of some consolidants preclude desalination treatments (very common in buried archaeological pieces). Although the concept of reversibility has been widely studied and discussed, the concept of retreatability is less studied. Research is usually focused on the application of these treatments, especially consolidation products to freshly excavated or deteriorated materials but not to samples that have been already treated with other products. Therefore and despite being one of the most important criteria in the selection of conservation products and treatments, retreatability is less considered in conservation studies. The time of this hydrophobic behavior varies according to the authors and the research, as well as with commercial products applied and substrates treated. The hydrophobicity has been observed until 28 days of curing [26], five weeks [27], three weeks [28], one month [29], two months [30] and several months [22, 31]. Environmental data loggers, ibuttons model DS1923-F5, were introduced into the container and also placed in the laboratory to register T and RH during the test, using the software OneWireViewer version 3.04. Further details on the use of these consolidating products, application method and environmental conditions are described in ZornozaIndart et al. [21]. Analytical techniques and test methods Chert characterization The decay and conservation state of the samples was studied following the protocol and terms established by Font et al. [15] based on macro-visual inspection with the naked eye. In addition, X-ray diffraction (XRD) was used to determine the main mineralogical phases and degree of crystallinity of the samples. A Phillips PW-1710 diffractometer was used with CuKα radiation to study the total powder fraction of samples. The measurements were conducted by step scanning 2θ from 2 o and 68 o, scan step size 0.02 o, scanning rate 2 o/min, with a continuous mode and beam intensity of 40 kV and 40 mA. Three chert samples (SX3, SX17 and SX18) were analyzed with higher precision to distinguish the different silica polymorphs by means of a multi-purpose PANalytical X´Pert MPD with CuKα radiation. Analysis conditions were 2θ between 2 o and 90 o, scan step size 0.02 o, count time of 3 seconds per step, with continuous mode and beam intensity of 45 kV and 40 mA. Furthermore, sample SX18 was preserved and analyzed, without grinding, using micro-diffraction analyses to obtain several diffractogram patterns on various spots of interest. The mineralogical phases were identified by comparing the sample with the Joint Committee on Powder Diffraction (JCPDS) database and Bruker AXS DiffracPlus EVA software. Ion chromatography (IC) analyses were performed to identify soluble salts in all the samples, i.e. type and quantity of some anions (Cl−, NO3− and SO42−). Soil samples collected from the same dig areas as the chert samples were also analyzed to ascertain whether or not the presence of salts. Approximately 0.1 g of sample was dissolved in 10 ml of Milli-Q ultrapure water and placed for 45 min in an ultrasonic bath at room temperature. Afterward it was centrifuged for 5 min at 3500 rpm and 3400 rfc centrifugal force. The soluble salts in the extracted sample were quantified using a Metrohm 761 Compact IC ion chromatograph. The surface morphology and texture of the specimens were examined by means of environmental electron scanning microscopy (ESEM) using an Inspect FEI microscope coupled with energy dispersive X-ray spectroscopy (EDS) (model 7509 Oxford Instrument Analytical, UK). Optical surface roughness (OSR) analyses were also performed on the surface of samples SX18 and SX19 (the other samples were too small to take these measurements on). This was carried out to evaluate changes in the surface roughness caused by decay processes. The equipment used was a contact-free surface profilometer (white light), TRACEiT, Innowep GmbH. The OSR analyses included the generation of 3D-topography maps (25 mm2) using Gyddion 2.44 software displaying the average roughness parameters according to ISO 4287 (1998) standards [33], i.e., Ra (arithmetic mean of the absolute values of profile deviations from the mean line) and Rz (sum of the vertical distances between the five highest peaks and the five deepest valleys within the sampling length). The cutoff (λc) used for the calculations was 0.80 mm. Consolidation The sixteen specimens were also analyzed using various non-destructive techniques and test methods both before application of the consolidating products and one month later. The surface morphology, texture of the specimens, and the distribution and morphology of consolidating products was examined by ESEM-EDS. A peeling test was carried out on the surface of the samples to assess the detachment degree of the material using transparent double-sided adhesion tape (Tesa), with 1.5 cm wide × 5 cm long, on 1 zone per sample (ten sequences) and 90 seconds per each sequence. This method is commonly used for evaluating the consolidation effect of the products on stone surfaces by determining the detached material after applying and removing pressure sensitive tape over the surface [34]. Archaeological chert artifacts from Atapuerca sites (Burgos, Spain) 25 CONSERVAR PATRIMÓNIO 36 (2021) Spectrophotometry was performed to determine the color parameters, with a spectrophotometer MINOLTA CM-700d using the CieLab color space; the measuring area was 1-3 mm. Three measurements were conducted for each sample, the standard illuminant was D65 and the observer angle, 10 o. The measured parameters were L*, which accounts for luminosity, a* and b* coordinates (a* being the red-green parameter and b* the blue-yellow), total chrome difference ΔC* (from the formula ΔC* = (Δa*)2 + (Δb*)2)1/2), and total color difference ΔE* (from the formula ΔE* = ((ΔL*)2 + (Δa*)2 + (Δb*)2)1/2). The white (WI) and yellow (YI) indices were measured according to ASTM E313-73 [35] and brightness was measured according to ISO 2470-2 [36]. Results and discussion Characterization of chert samples Macroscopic examination The main results from the macroscopic examination are shown in Table 3. All the weathered samples from the archeological sites displayed a milky-white color and lost the characteristic densely packed cohesion of chert. No fissures or fractures were observed with the naked eye in any of the samples. However, all of them showed a coarse surface with different degrees of decohesion (Figure 2a-d). Sample SX17 (Figure 2a) was the most damaged chert. Part of this sample was broken down into a white powder and the rest was very dusty, friable and disaggregated on the surface. Sample SX9 (Figure 2b) displayed a certain degree of compacted structure, slightly higher than the previous sample. It also had a dusty, friable and disaggregated surface that powdered to the touch. In spite of some lost material, samples SX3 and SX1 (Figure 2c-d) seemed to be better preserved compared to the other samples. These were more compacted and did not present the same degree of disaggregation as the previous samples, even though they also had coarse, grainy and quite dusty surfaces. Among the chert samples from the outcrops surrounding the archaeological sites, sample SX19 (Figure 2a) was muddy-white in color with a coarse but compacted surface, similar to samples SX1 and SX3. Two different zones could be distinguished on the surface of this sample, a more crystalline, compacted and smoother core zone, and another whiter, coarser and dustier external zone that appeared more weathered. Sample SX18, also from the surrounding outcrops, was grayish-brown in color with some faceted core areas that were fine grained and crystalline (Figure 2f). Only this latter zone of this particular sample displayed the characteristic cutting edge morphology caused by the distinctive conchoidal fracture of the chert. This core zone was surrounded by a 1 mm-thick crust, grayish-white in color and earthy but with a compacted appearance. This was circled by another 5 mm-thick, white external crust that in spite of its compacted structure displayed a similar aspect to the more weathered samples previously described (milky-white in color, rough, dusty and with a disaggregated surface). X-ray diffraction (XRD) All the XRD results obtained from the total powder samples showed the presence of quartz (SiO2) and moganite (SiO2 polymorph), with the exception of SX9 and SX17 which were solely composed of quartz and calcite (CaCO3); SX17 had a large amount of calcite. These results are in agreement with those obtained by other authors for similar materials [17, 37]. Sample SX19 had more moganite in the whiter external zone compared to the more crystalline core zone. The amount of moganite was even higher in sample SX3, which also displayed some calcite, followed by sample SX18 (Figure 3a). The four diffractogram patterns obtained in different spots on this unground sample indicate certain mineralogical differences (Figure 3b). The external zone (surrounding crust; crust 1) was mainly composed of quartz and calcite with a lower proportion of dolomite, and had poorly-defined broad peaks indicating less crystallinity. The most external white crust (crust 2) showed only quartz and very well-defined moganite peaks. The quartz peak at 41 o 2θ only appeared well-defined in the crystalline core zones of the sample (Chert 1; Figure 3b). As inside the same nodule or quartz level textures and impurities can change, that’s the main difference between the faceted and colored areas in the crystalline core zone of the sample [38]. This is also the main difference between the faceted and colored areas in the crystalline core zone of this sample, since changes in texture and impurities can occur inside the same nodule or quartz level [38]. The chert samples from the archaeological site and the most weathered samples from the surrounding outcrops (with a totally disaggregated surface), i.e., samples SX17, SX9 and the most external crust of sample SX18, displayed quartz and calcite minerals. The less-weathered samples, Nomenclature Macroscopic examination Middle Pleistocene lithic remains SX1 Better preserved. More compacted, without disaggregation but with coarse, grainy and quite dusty surfaces. SX3 SX9 Certain degree of compacted structure but, dusty, friable and disaggregated surface that powdered to the touch. SX17 Most damaged. Broken down into a white powder, very dusty, friable and disaggregated on the surface. Fresh chert samples SX18 Grayish-brown in color with some faceted core areas that were fine grained and crystalline. SX19 Muddy-white in color with a coarse but compacted surface. Two different zones: a more crystalline, compacted and smoother core zone, and another whiter, coarser and dustier external zone that appeared more weathered. Table 3. Macroscopic examination of archaeological chert samples. A. Zornoza-Indart, P. Lopez-Arce, L. López-Polín 26 CONSERVAR PATRIMÓNIO 36 (2021) with coarse, grainy but more compacted surfaces (samples SX1, SX3, SX19 and middle crust of sample SX18), contained quartz, moganite and small amounts of calcite. Finally, the best preserved crystalline cores of samples SX18 and SX19 had only quartz. High alkalinity, sulfates and ferric ion activity play a certain role in the precipitation of moganite [38]. In the research work on Mogan rocks from Gran Canaria, Spain, carried out by García-Guinea et al. [39], it was found that their core area and edges (more porous, whiter and with additional ions) resulted in a different cathodoluminescence spectrum. This was interpreted as late hydrothermal weathering mechanisms with alkaline ions, metals and volatiles to form moganite. In silica weathering, moganite represents an intermediate position between opaline phases and quartz [38]. It is considered a metastable phase which can be transformed into quartz if there is enough time or there are changes in the surrounding environmental conditions [40]. The abundance of moganite in arid environments has been partially explained by the lack of water for mediating the dissolution of this mineral and the simultaneous precipitation of quartz, as between 10 % and 80 % of the silica present in different varieties of fine-grained quartz is in fact moganite [41]. In this case, the quantity of moganite is a key indicator in the decay of chert samples. The crystalline core zone of sample SX18 from the outcrops, which only has quartz, is in a good state of preservation. The chert samples from the archeological site and the outer crust of sample SX18 (with mixtures of quartz-moganite) are weathered, with coarse, porous and disaggregated surfaces. In quartzmoganite mixtures, fast weathering may occur, as moganite is more soluble than quartz. This means that those parts of the chert containing moganite could be more weathered and more porous, favoring fluid penetration leading to disaggregation processes. In the research conducted by Navazo et al. [17], poorly-preserved Neogene chert, and massive, well-preserved Cretaceous chert samples with low porosity, both from outcrops close to the Atapuerca mountain range, were compared. They also found that moganite played a significant role in chert decay and preservation state, since the quantity of this mineral was the main difference between the two types of chert. Increased amounts of calcite and decreased moganite in the samples, relating to a greater degree of decay and weathering, may be due to the fact that the areas originally comprising moganite could have been weathered or dissolved. This favors fluid penetration and pore filling by external compounds [39], in this case by calcite since these cherts were within limestones rock settlements. The absence of moganite in the most weathered chert samples could be in line with the data provided by Heaney & Post [41] and Rodgers & Cressey [40], who report that moganite does not appear in weathered or eroded cherts, since it completely disappears due to dissolution processes. Ion chromatography (IC) Ion chromatography data indicate negligible amounts of soluble salts (chlorides, nitrates and sulfates), with total weight percentages of salts below 0.08 % in all samples as it was described in Zornoza-Indart et al. [21]. The soil samples collected from the same dig areas where the chert samples were buried also contain insignificant amounts of salts. Therefore, the decay of the chert samples by salt crystallization processes is discounted. Optical surface roughness (OSR) The surface roughness analysis results are shown in Table 4. Figure 3. X-ray diffraction (XRD) patterns of the archaeological chert samples: a) XRD obtained from the total powder samples; b) Diffractograms from micro XRD analysis (no powdered samples) obtained in SX18 sample, in the external (crust) and in the core zones (chert). Q: quartz; Mog: moganite; C: calcite; D: dolomite. a b Sample Area Ra (µm) Rz (µm) SX18 Smooth 2.57 ± 0.40 9.35 ± 1.37 Crust 7.43 ± 0.34 40.82 ± 1.84 SX19 Smooth 3.67 ± 0.29 17.52 ± 1.12 Weathered 10.30 ± 0.42 41.62 ± 1.94 Table 4. Optical surface roughness (OSR) parameters (Ra and Rz) of fresh chert samples SX18 and SX19, measured on the external crust, weathered surfaces and smooth areas. Archaeological chert artifacts from Atapuerca sites (Burgos, Spain) 27 CONSERVAR PATRIMÓNIO 36 (2021) The micro-detailed surface images can be seen in Figure 4, and 3D OSR maps are presented in Figure 5. By comparing the results obtained on the smooth, crystalline core zones and less weathered parts (that are more similar to fresh chert) of the two studied samples (SX18 and SX19), it can be seen that the Ra roughness values for sample SX19 are 43 % higher than for sample SX18 (3.67 ± 0.29 µm vs. 2.57 ± 0.40 µm, respectively). The same occurs with the Rz values that are 87 % higher in sample SX19 (17.52 ± 1.12 µm vs. 9.35 ± 1. 37 µm). These results show that sample SX19 is much coarser than sample SX18 due to surface degradation processes, as was also observed in the macroscopic study. The average surface roughness values are greater in both the smoother and better-preserved areas and across the entire surface of sample SX19, which is denoted by the longer distances between the highest and deepest parts of the measurement areas indicated by the Rz parameter. In the coarse and decayed areas, the rough and dusty external zone of sample SX19 and the external crust of SX18 (crust 1, which according to XRD comprises quartz and calcite), roughness values are similar in the two samples, although slightly higher in sample SX19 (7.43 ± 0.34 µm vs. 10.30 ± 0.42 µm in the Ra parameter and 40.82 ± 1.84 µm vs. 41.62 ± 1.94 µm in the Rz parameter respectively). The micro-detailed images obtained with the profilometer (Figure 4) and the surface roughness 3D height maps Figure 4. Micro-detailed images obtained by optical surface roughness (OSR) measurements: a) Even area of sample SX18; b) External crust of the sample SX18; c) Coarse surface texture of sample SX19. Figure 5. Surface roughness 3D height maps obtained under optical surface roughness (OSR) on the different surface areas of chert SX18 and SX19 samples: a) Even area of sample SX18; b) External crust of the sample SX18; c) Even area of sample SX19; d) Coarse surface of sample SX19. a b c b a c d A. Zornoza-Indart, P. Lopez-Arce, L. López-Polín 28 CONSERVAR PATRIMÓNIO 36 (2021) (Figure 5) show the surface differences on the texture morphology, and hence the surface roughness contrast among several zones of both samples. The image of an even area of sample SX18 shows a smooth, compacted and homogeneous surface (Figure 4a and Figure 5a), whereas the image of sample SX19 reveals a whiter color and dustier, poorly-compacted surface, where mineral grains are clearly observed, as well as discontinuous areas with pores (Figure 4c and Figure 5c-d). In the most decayed areas, some further differences can be seen (Figure 5b-d). The transition from the smooth area towards the external rough zone of sample SX19 can be observed as a progressive loss of homogeneity and compaction of the surface, showing up as a more white and porous, disaggregated grainy zone where loss of material is observed. In sample SX18, the image shows the previous step in the total decay process observed in sample SX19. The color is still darker than the surface coloring of sample SX19, although white areas not detected in the smooth zone can be observed. The surface of SX18 sample maintains its continuity and homogeneity, but cracks and fissures appear and, eventually, if decay processes continue, there could be loss of material and disaggregation of the surface, in a similar way to that observed in sample SX19. Environmental SEM-EDS Sample SX3 from the archaeological site exhibits silica with radial-fibrous textures in some areas and nodular textures in others (Figure 6). Even though sample SX17 is more homogenous, the fibrous texture is less rigidly oriented (Figure 6d). The external part of sample SX18, from the outcrop, shows signs of dissolution, retraction cracks and has a higher calcium content (Figure 7a). A different zone (corresponding to the whitish crust) displays a similar texture to the crystalline core zone. However, in this case it is weathered due to dissolution-recrystallization processes (Figure 7b). In the crystalline core zone, a crypto-crystalline, fairly fibrous texture can be seen (Figure 7c), together with some areas containing pores and fissures from 50 to 100 µm in size, almost completely cemented by micro-crystalline quartz crystals (Figure 7d). These pore and fissured areas might correspond to moganite, leading to a certain degree of micro-porosity. These results are in line with those obtained by XRD, where differences between the two crusts were distinguished (the outer crust had a higher degree of alteration and comprised quartz and calcite, while the other crust was composed of calcite and moganite). Variability in moganite content has been linked to macroscopic color and texture variations within a sample, indicating that variable moganite content may be linked to changing environmental or depositional conditions [42]. The outer zone of sample SX19 presents crystals with radial-fibrous morphologies (which could be moganite) together with nodules of agglomerated Figure 6. Environmental SEM (ESEM) images from the archaeological site chert samples: a) Sample SX3; b) Detail of radial-fibrous texture area in former sample; c) Detail of nodular textures in the same sample; d) Detail of disoriented fibrous texture in sample SX17. a b c d Archaeological chert artifacts from Atapuerca sites (Burgos, Spain) 35 CONSERVAR PATRIMÓNIO 36 (2021) Stone’, Key Engineering Materials 391 (2009) 1-25, https://doi. org/10.4028/www.scientific.net/KEM.391.1. 32. Naidu, S.; Liu, C.; Scherer, G. W., ‘Hydroxyapatite-based consolidant and the acceleration of hydrolysis of silicatebased consolidants’, Journal of Cultural Heritage 16(1) (2015) 94101, https://doi.org/10.1016/j.culher.2014.01.001. 33. ISO. 4287. Geometrical product specifications (GPS)-surface texture: Profile method-terms, definitions and surface texture parameters 4287 Germany (1998). 34. Drdacky, M.; Lesak, J.; Rescic, S.; Slızkova, Z.; Tiano, P.; Valach, J., ‘Standardization of peeling tests for assessing the cohesion and consolidation characteristics of historic stone surfaces’, Materials and Structures 45 (2012) 505-520, https://doi. org/10.1617/s11527-011-9778-x. 35. ASTM. E313-73. Practice for calculating yellowness and whiteness indices from instrumentally measured color coordinates E313-73 (1993), Conshohocken, PA.: ASTM International. 36. ISO. 2470-2. Paper, board and pulps–Measurement of diffuse blue reflectance factor (ISO brightness) 2470-2. Geneva, Switzerland (1999). 37. Sala, R., Introducció d'un model reòlogic de les deformacions microscòpiques per ús en objectes lítics. Primers resultats al Complex del Plistocè Mig de la Sierra de Atapuerca, Universitat de Barcelona, Barcelona (1993). 38. Bustillo, M. A., ‘Aparicion y significado de la moganita en las rocas de la silice: Una revisión’, Journal of Iberian Geology 28 (2002) 157-166. 39. García-Guinea, J.; Bustillo, M. A.; Crespo, E.; Finch, A;, Townsend, P. D.; Hole, D. E.; Correcher, V., ‘Spectrallyresolved luminescence of moganite from Mogan (Gran Canaria)’, in Conference on Micro-Raman Spectroscopy and Luminescence Studies in the Earth and Planetary Sciences, Mainz, Germany (2009). 40. Rodgers, K. A.; Cressey, G., ‘The occurrence, detection and significance of moganite among some silica sinters’, Mineralogical Magazine 65 (2001) 157-167, https://doi. org/10.1180/002646101550181. 41. Heaney, P.; Post, J. E., ‘The widespread distribution of a novel silica polymorph in microcrystalline quartz varieties’, Science 255 (1992) 441-443, https://doi.org/10.1126/science.255.5043.441. 42. Kingma, K.J.; Hemley, R. J., ‘Raman spectroscopic study of microcrystalline silica’, American Mineralogist 79 (1994) 269-273. 43. CNR-ICR. NORMAL 20/85. Interventi conservativi: progettazione esecuzione e valutazione preventive (1996), Milan, Italy. 44. Benavente, D.; Martınez-Verdu, F.; Bernabeu, A.; Viqueira, V.; Fort, R.; García del Cura, M. A.; Ordoñez, S., ‘Influence of surface roughness on color changes in building stones’, Color Research & Application 28(5) (2003) 343-351, https://doi. org/10.1002/col.10178. 45. Rodrigues, J. D.; Grossi, A., ‘Indicators and ratings for the compatibility assessment of conservation actions’, Journal of Cultural Heritage 8(1) (2007) 32-43, https://doi.org/10.1016/j. culher.2006.04.007. RECEIVED: 2019.10.8 REVISED: 2020.4.1 ACCEPTED: 2020.4.9 ONLINE: 2020.6.29 This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/deed.en. A. Zornoza-Indart, P. Lopez-Arce, L. López-Polín