Developing a Provenance Framework for Ancient Stone Materials: A Subduction-Related Serpentinite Case Study from Tinos, Cyclades, Greece

Abstract

Serpentinite is a common lithology in Greece with significant geotectonic and petrogenetic implications, yet its archaeometric study remains limited. This also applies to the Ras serpentinite outcrop on Tinos Island—an enigmatic ancient quarry with no prior documentation of its exploitation. This study aims to provide a holistic provenance analysis and full material characterization to establish a unique fingerprint for future comparative research. A multidisciplinary methodology was applied, including petrography, XRD, FTIR, BET, LA-ICP-MS, electron microprobe analyses, and lichenometry—the latter used to estimate the quarry’s period of activity. The Ras serpentinite consists of a dark-green, brecciated lizardite matrix cut by a pale-green chrysotile network, and notably hosts the first recorded occurrence of minnesotaite in Greece. Relict antigorite within the matrix indicates high-pressure, high-temperature metamorphism preceding a greenschist facies overprint. Geochemical data classify the Ras occurrence as a subduction-related serpentinite derived from a depleted ultrabasic protolith. The aforementioned traits distinguish it from other Greek examples. Lichenometric analysis suggests the quarry was active in the 13th century A.D., with evidence of deliberate extraction techniques. These findings offer valuable insights into archaeometric studies and the provenance of serpentinites in the Aegean.

1. Introduction

Serpentinite is a common lithology of the Earth, usually related to ancient oceanic realms and has been proven to represent different modes of formation. The general classification of [1] will be adapted in this work, in which serpentinites are divided into (a) subduction zone, (b) mantle wedge, and (c) abyssal. This classification is based on geochemical and mineralogical criteria. Serpentinites comprise more than 90% of the serpentine group minerals. This group entails antigorite, lizardite, and chrysotile, each one representing successively lower degrees of metamorphosis/metasomatism. The existence of antigorite points to metamorphic degrees above blueschist facies metamorphism, whereas lizardite and chrysotile are indicative of greenschist facies, with chrysotile being formed in lower P-T conditions [2,3,4,5]. These minerals crystallize at the expense of primary minerals (olivine and pyroxenes) of a pre-existing ultrabasic rock (and/or mantle rocks). These original rocks, known as protoliths, are exposed to various geotectonic environments that drive their transformation through metamorphic processes. Depending on the prevailing pressure–temperature (P-T) conditions, these processes alter the protoliths’ mineralogical composition, geochemical signatures, and textural features. In most cases, serpentinites are subjected to retrograde metamorphosis after reaching their peak metamorphic event, resulting in complex mineralogical, geochemical, and textural modifications.

In Greece, serpentinites outcrop along three distinct areas with related mafic and ultramafic rocks, such as the Pindos ophiolites [2,6,7,8] and Vardar ophiolites [9,10,11] and occurrences of Attico-cycladic massifs and Rhodope Massifs [12,13]. Due to serpentinite’s accessibility, as well as its mechanical properties such as softness, it is a widely used material, especially during prehistoric and early historic eras. Serpentinite and related rocks, such as ophicalcite (serpentinite enriched in carbonate minerals, usually forming a network) and ophicalcite breccia, display a wide range of uses. Serpentinites were used as tools (Neolithic Avgi; [14]) and low-grade architectural features (Knossos; [15]), whereas the rather mechanically stronger ophicalcite was used in prestigious architectural features, even being used in UNESCO Heritage sites such as the Hagia Sophia (Green Thessalian stone; [16]). This seems to be the norm for polychrome ‘marbles’ (commercial term) such as the green Lapis Lacedaemonius (altered volcanic rock; [17]) used in St. Peter’s Basilica.

The current case study focuses on the Ras serpentinite outcrop on Tinos Island (Greece), a region also known for the modern and extensive extraction of ophicalcite—commercially known as ‘Tinos Green’—a brecciated serpentinite cemented with carbonate minerals. Despite the prominence of this material in contemporary quarrying, archaeological investigations on Tinos remain limited, and there is no documented evidence of “Tinos Green” being used in antiquity. Similarly, for the adjacent Ras serpentinite occurrence, no information exists regarding the period of its exploitation or its potential historical applications. In addition to these distinctive green stones, Tinos has long been known for its extraction of white marble, both in ancient times and in present-day industry [18]. This underscores Tinos’s significance as a longstanding center of marble/stone production.

While there have been previous attempts at provenance analysis of serpentinite-related rocks, these studies have often been limited in scope, typically relying on X-ray techniques, lacking comprehensive elemental data, and with low detection limits. This study addresses those gaps by presenting an integrated analytical framework that combines mineralogical, geochemical, and physical characterization—as for the analytical approach, see below. It also compiles and contextualizes major published data on Greek serpentinites and offers new insights into the potential uses and chronology of exploitation of the Ras occurrence.

2. Materials and Methods

The geological samples examined in this study were collected from the debris piles adjacent to the ancient serpentinite quarries. To capture the full range of textural variability and assess its impact on the material’s physical properties, samples with distinct textures were systematically selected (Table S1). Macroscopic images of the specimens were analyzed using ImageJ.JS software. Representative areas were chosen, photographed, and converted into 8-bit grayscale images. The images were scaled to actual size and contrast and brightness were optimized manually for each case to better distinguish the two serpentine polymorphs. Next, a manual threshold was applied and the image was binarized in black and white colors (Figure 1), isolating and quantifying the lizardite and chrysotile phases. The binary images were analyzed and the results were summarized, calculating the percentage of the white-colored element of the image. The finalized image can later be pseudo-colorized using a threshold tool.

2.1. Electron Probe Microanalysis (EPMA)

Mineral chemistry and backscattered electron (BSE) images were obtained using an electron probe microanalyzer (JEOL JXA-8530F Plus EPMA, JEOL Ltd., Tokyo, Japan) equipped with five wavelength-dispersive X-ray spectrometers (WDS) and an energy-dispersive spectrometer (EDS). The operating conditions were 15 kV accelerating voltage, 10 nA beam current, and ~1–5 μm beam diameter, with counting times for peak and total background of 20 s. The EPMA analyses were conducted at the Institute of Earth Sciences—NAWI Graz Geocenter, University of Graz, Austria.

2.2. Trace Element Analysis (LA-ICP-MS)

The trace and rare earth element geochemistry of the serpentine group minerals were determined on a polished thin section by the LA-ICP-MS system at the NAWI Graz Central Lab for Water, Minerals, and Rocks (the University of Graz and the Graz University of Technology, Graz, Austria). The procedure is described in [10].

2.3. X-Ray Diffraction (XRD)

Two grams of a composite whole-rock sample were pulverized along with a selection of pulverized lizardite-enriched matrix to be analyzed via X-Ray diffraction, according to the procedure in [10]. The average crystal size of serpentine group minerals was deduced via the Scherrer equation:

B(2θ) = K*λ/L*cosθ

(1)

where B(2θ) is the peak width; L is the crystallite size; the K constant is generally taken as being about 1.0 for spherical nanoparticles; θ is the diffraction angle; and λ is the X-ray wavelength.

2.4. Raman Spectroscopy

Raman spectroscopy was performed on the samples T2, T3, T4, T5, and T6 (Table S1) using a BRAVO™ handheld Raman spectrometer (by Bruker Optics GmbH & Co., Ettlingen, Germany) at the Laboratory of Archaeometry, University of Peloponnese. The spectrometer is equipped with two near-infrared excitation lasers (DUO LASER™, wavelengths at 785 nm and 853 nm) and a CCD detector, allowing for a spectral range of 300–3200 cm⁻¹ and a spectral resolution of 10–12 cm⁻¹. The two lasers operate in a patented sequentially shifted mode (SSE™, Sequentially Shifted Excitation), allowing for the mitigation of fluorescence from samples. The duration of each measurement was typically around 1 to 2 min. Data acquisition and processing were carried out using Origin 2023 software. The identification of the Raman bands in the acquired spectra was carried out by comparing them to the RUFF database.

2.5. Fourier Transform Infrared Spectroscopy (FTIR) and Brunauer–Emmett–Teller (BET) Analysis

Both techniques were applied in a composite whole-rock sample used for the X-ray analysis. Fourier-Transform Infrared Spectroscopy (FTIR) was carried out in pelleted samples at the Department of Materials Science, University of Patras. The FTIR study was performed with the Spectrum 100 by Perkin Elmer FTIR spectrophotometer. The instrument has a range of 7800 to 250 cm⁻¹ with 0.5 cm⁻¹ resolution, although a range of 4000 to 360 cm⁻¹ was used. The specific surface area (SSA), the pore volume (VP), and the pore size distribution of serpentine were determined from the adsorption and desorption isotherms of N₂ at −196 °C using a NOVATOUCH gas sorption system by Quantachrome instruments. Prior to the measurements, the samples were outgassed at 200 °C for 1 h, under a vacuum. The Origin software package was used for data handling.

Brunauer–Emmett–Teller (BET) analysis was conducted to measure the specific surface area and pore characteristics of the serpentinite samples, based on nitrogen adsorption–desorption isotherms at 77 K. The specific surface area (SSA), pore volume (V_p), and pore size distribution were determined from the N₂ adsorption–desorption isotherms at −196 °C using a NOVA TOUCH gas sorption system (Quantachrome Instruments) at the Department of Materials Science, University of Patras.

2.6. Mechanical Properties Testing

The mechanical properties of the serpentinite rocks were determined by a point load test (PLT) performed on samples T2, T3, T4, T5, and T6 (Table S1). The PLT is an effective and rapid way to predict the Uniaxial Compressive Strength (UCS) of a rock sample, which is widely recognized as the most important component in any engineering geology project. UCS analysis involves a time-consuming and expensive laboratory test. Hence, many researchers employ indirect methodologies to calculate this parameter. Among the indirect tests, PLT is the most applied test, utilizing the Point Load Strength Index (Is(50)) as an indirect measure of the compressive strength of rock. For this specific study, the laboratory tests of PLT were conducted in the Laboratory of Engineering Geology, Department of Geology, University of Patras, using point load test apparatus.

2.7. Lichenometric Data

Lichenometry is a method for dating the exposure of rock surfaces by measuring radial growth rates of lichens growing on the surface of interest. Fundamental to this method is the hypothesis that the largest lichens (LLs) on a rock are potentially the oldest individuals and are the same age as the exposure of the rock surface they grow on. Key to this method is the establishment of a growth curve using measurements of lichens on substrates of known age (bridges, gravestones, and abandoned farms). With these data, the growth rate of a species can be calculated and used to estimate the age of a lichen on a surface of unknown age [19]. Austrian botanist Ronald Ernst Beschel established this dating method in the 1950s with the aim of estimating the age of glacial rock deposits (moraines) [20]. Since then, lichenometry has been successfully used in many other disciplines, like dating earthquake-generated rock falls [21], landslides [22], and sea-level changes [23], and in archaeology for dating stone walls [24], other stone-made monuments, and the time of the abandonment of quarries [25].

3. Geology of Tinos Island

Tinos Island is part of the Attic-Cycladic Crystalline Belt (ACCB), consisting of the Upper and Lower tectonic Units. The ACCB emerged as the outcome of the Tertiary HT–HP convergence and the subsequent collision of the Eurasian plate and Apulian microplate. During the following subduction event, the Cycladic Blueschist Unit (CBU) was formed [26,27,28]. During the Oligocene–Miocene exhumation, the HP–HT units were overprinted by greenschist facies [29].

The three main tectonic Units on Tinos include the following [13,30,31,32]: (a) the Upper Tectonic Unit (UTU) with a metamorphosed dismembered ophiolite complex that is believed to have been metamorphosed under greenschist facies [30,33] along with the metamorphic Akrotiri Unit [34]; (b) the most voluminous unit, the Lower Tectonic Unit (LTU), includes three marble horizon occurrences [32] reaching HP–LT metamorphism (CBU) and being overprinted by greenschist facies [35]; and (c) the Basal Tectonic Unit (BTU) outcrops at the NW part of Tinos, and it is dominated by dolomitic marble and phyllites metamorphosed in greenschist facies [33]. Both the UTU and the LTU have been intruded into by the Miocene Tinos Pluton [36]. The case study of the Ras serpentinite (Figure 2a) lies within the UTU, with the serpentinite overlaying, with sharp contact, the greenschist formation (Figure 2b), and confined talc zones may appear at the contact areas of both lithologies.

4. Results

4.1. Field Work and Dating of the Quarry

The abandoned serpentinite quarry of Ras was mined towards the south, covering ~300 m² with a perimeter of ~71 m. This orientation is consistent with the current quarrying activity in the region. The lithology is homogeneous and comprised of serpentinite with an absence of calcite stockwork, as noted in the neighboring ophicalcite quarries within a 2 km range of the study area (Figure 2b). Two distinct open pits are found in the study area, labelled Pit A and Pit B, at Ras Hill, providing insights into ancient extraction practices and potential chronological and functional distinctions between the two (Figure 2c). Pit A (Figure 3a–d), the larger of the two, displays well-preserved tool marks indicative of organized extraction methods, including the use of quarry picks, point chisels, and cylindrical wedges [38]. The material was quarried in blocks (Figure 3c). The dimensions of the surviving blocks in Pit A, as seen in a partially extracted block, are 45 × 45 × 50 cm. Additional evidence of scaffolding mechanisms is suggested by the presence of aligned holes on the NE wall, while exploratory trenches at the hilltop reveal further quarrying attempts. Lichens are developing on the Ras façades (Figure 3d).

Two quarry levels are evident: a main level with waste material scattered across the surface and a smaller section on the northeast (NE) wall, which stands approximately 12.5 m long and 5.3 m high. The north façade of Pit A also preserves tool marks, aligned with geological discontinuities, revealing systematic, directional cuts—primarily at 55° NE–SW and 40° NW–SE—alongside vertical markings that reflect a shear zone (Figure 3e). The exposed serpentinite surfaces are highly cataclastic and mylonitized and, in places, shear zones have developed. Serpentinite is composed of massive dark-green color serpentine, which is cross-cut by a pale-green colored serpentine network, filling the fractured serpentinite (Figure 4). The texture differs in terms of (a) pale-green to dark-green serpentinite ratios and (b) network texture, classified as either oriented pseudo-augen development (Figure 4a), randomly oriented (Figure 4b), or enlarged dense network (Table S1). Pit B is smaller and lacks discernible tool marks.

Artefacts from the Tinian serpentinite quarry “Ras” are not known. Therefore, archaeological evidence cannot establish a chronology for the use and abandonment of the quarry. In order to identify the most recent activity in the serpentinite quarry, lichenometry is the only possible method for age estimation. When a quarry is abandoned, and the freshly exposed rock surfaces are not disturbed anymore, they will immediately be colonized by various organisms. Lichens are one of the first colonizers. Therefore, measurements of the largest lichen (LL) within the façade of the quarry were taken, which might indicate the oldest individuals. The crustose lichen species Aspicilia intermutans (Nyl.) Arnold, which is abundant on the serpentinite substrate, was chosen for measurements. The longevity and slow growth rate of this species make it suitable for the study.

A local growth curve was established based on six dated reference surfaces in Tinos as well as in Naxos (which is in the same climatic region; Figure 5; Table S4), showing a moderately positive correlation between thallus diameter and age (R² = 0.58) caused by the initial rapid growth in young lichens, which stagnates with age and becomes almost linear. The resulting regression equation (y = 0.294x + 29.005) provides an empirical basis to estimate exposure durations in the study area on a linear trend line. Extrapolating from this model, the 239 mm thallus corresponds to an estimated exposure age of approximately 800 years, suggesting that the Ras serpentinite quarry may have been abandoned as early as the 13th century A.D. Given the extrapolated nature of the estimate and the moderate fit of the regression, this value should be treated as approximate, yet in the absence of archeaological dating it provides a useful terminus ante quem for the most recent quarrying activity.

4.2. Petrographic and Mineralogical Results

Ras serpentinite is composed of mesh and hour-glass texture lizardite ranging between 45.98 and 61.40 vol%. Second generation fibrous chrysotile veins (up to 2 mm wide) crosscut the lizardite mass (Figure 6a). Primary spinel grains, found within the lizardite mass, have been completely pseudomorphed into magnetite and clinochlore coronas (Figure 6b,c). Pseudomorphosed spinel or disseminated newly formed magnetite are exclusively included in the lizardite matrix (Figure 6a). Minnesotaite appears orange under polarized light (Figure 6b) and it forms fine-grained aggregates (Figure 6c) crystalized between magnetite grain interstices. Flaky relics of antigorite (<300 μm) were located dispersed within the lizardite matrix. The classification of the serpentine polymorphs was conducted through XRD and Raman spectroscopy (Figure 7a,b; Section 4.3).

4.3. Physical Characteristics of Ras Serpentinite

The physical study of Ras serpentinite was based on X-ray, Raman, BET, FTIR, and strength analysis. The X-ray analysis revealed the participation of the following mineral phases: serpentine (var. lizardite, chrysotile, and antigorite), clinochlore, and magnetite (Figure 7a). The XRD diffractogram of the lizardite-enriched sample from Ras serpentinite indicates its trioctahedral structure (e.g., 2θ~12°, 24.5° and 36° that correspond to (001), (111), and (002) crystalline planes) and belongs to space group P31 m. Lizardite is characterized by strong and sharp diffraction peaks, indicating advanced crystallinity and homogeneity. The average crystal size of its crystallites regarding their length and width, estimated in terms of the Scherrer equation, is ~73 ± 1 and ~36 ± 1 nm, respectively. Raman spectra (Figure 7b) yielded the same results, with antigorite being detectable in two samples (T2 and T5). A sample of Tinos Green was also analyzed, displaying a distinctive peak representing calcite, while lacking significant magnetite mineralization.

Regarding the BET analysis (Figure 8a–c), Tinos’s serpentine shows a typical type II isotherm. This type of isotherm is representative of capillary condensation in its micro-to-meso pores, with a diameter of ~2.8 nm. The adsorption isotherm of serpentine rises gradually up to ~0.45 P/P0, implying that mono-molecule layer adsorption occurs on its surface. Above ~0.45 P/P0, saturation of mono-molecule layer adsorption is followed by multi-molecule layer adsorption. In the range of ~0.45 to ~0.9 P/P0, the desorption branch is noticeably higher than the adsorption one, resulting in the formation of an H3 hysteresis loop. This indicates that Tino’s serpentine is a mesoporous substrate of well-developed slit-like pores formed by the loose stacking of aggregates of platy-to-flaky particles [40]. Also, the presence of type H3 hysteresis loops illustrates that the interactions between the absorbed substances and the substrate are relatively weak, while the interactions between the adsorbed substances are strong. The pore volume of serpentine is low (i.e., 0.00773 ± 0.0002 cm³/g), whereas its specific surface area is large (i.e., 25.93 ± 4.2 m²/g). The pore size patterns suggest that Tinos’s serpentine displays mostly microporous and subsequent mesoporous distribution and reduced crystallinity.

The possible assignments of the Fourier Transform Infrared (FTIR) spectra (Figure 8d) of the analyzed serpentine suggest that the absorption bands at ~460 and ~540 cm⁻¹ represent the Si-O group asymmetrical bending, whereas the ones at ~620, ~650, ~780, and ~1100 cm⁻¹ represent the Si-O group symmetrical bending vibration. The bands between ~500 and ~620, and ~500 and ~620 cm⁻¹ indicate the Si-O4 group tetrahedral bending vibration, whilst the bands at ~480 and ~560 cm⁻¹ represent the octahedral and out-of-plain Mg-O6 group vibrations. The bands between ~530 and ~660 cm⁻¹ are attributed to the Mg-spinel assignments (e.g., magnesia spinel vibration).

The absorption bands between ~3650 and ~3690 cm⁻¹ represent the water-bending and inner-hydroxy-stretching vibrations, with a peak at ~3680 cm⁻¹, representing the 2nd hydroxyl. Due to the interlaminar structure of serpentine, the vibration peak at ~3440 cm⁻¹ corresponds to absorbed water, namely its interlayer water. Peaks at ~1600 and ~1410 cm⁻¹ most likely represent carboxyl vibrations and C-H bending. Likewise, the vibration peaks at ~1070 and ~940 cm⁻¹ are ascribed to the stretching absorption of the Si-O group. Furthermore, the FTIR serpentine spectra confirm its classification as lizardite (see XRD patterns).

According to Table S1, the Point Load Strength Index (Is(50)) values for samples T2–T5 (Ras serpentinite), range from 2.09 to 3.22 MPa and are characterized as medium strength, whereas T6 (‘Tinos ophicalcite’–‘Tinos green’) is calculated with a value (Is(50)) equal to 5.85 MPa and is characterized as high strength. Accordingly, the values of indirect determination of uniaxial compressive strength for samples T2–T5 range between 45.98 and 70.84 MPa and T6 has a value equal to 128.7 MPa, similar to those reported by the Hellenic Survey of Geology and Mineral Exploration (HSGME) at 100 MPa (‘Tinos Green’).

4.4. Chemical Make-Up of the Material

Serpentine is the main mineral phase, and although it is encountered in two distinct textures and polymorphs (lizardite and chrysotile), there are no significant chemical differences between them in major (microprobe analyses; Table S2) and trace elements (LA-ICP-MS analyses; Table S3). MgO contents are almost homogeneous, ranging from 39.05 to 39.65 wt%, FeO and NiO contents range between 3.03–4.12 and 0.23–0.37 wt%, respectively. Al₂O₃ contents range between 0.85 and 1.02 wt% (Figure 9a). Minnesotaite is a member of the pyrophyllite–talc group. In Ras, it is encountered as a Ni-enriched variety (NiO: 1.92–2.02 wt%), whereas MgO and FeO contents range between 2.41 and 2.64 and 37.01 and 39.45 wt%, respectively (Table S2).

Serpentine analyzed via LA-ICP-MS demonstrates the following ranges for each element and ratio: Cr/Ni 0.61–0.74, Ba 0.14–0.28, V 28.06–31.58, and Yb 0.05–0.08 (Figure 9b,c). Ras serpentine is LREE-depleted (Light Rare Earth elements La–Nd < 0.01 ppm; Table S3), and its Primitive Mantle (PM)-normalized patterns are described by positive Dy–Lu patterns (Figure 9d). Since most of the analyzed elements in a multi-elemental PM-normalized diagram are below the detection limit, we present a modified figure, taking into account both Li and B. Positive B, Pb, and P anomalies that are higher than PM values are distinctive and Cs values are also more enriched when compared to PM. Fluid Mobile Elements (FMEs: P, Li, B, Sr, Rb, Cs, Ba, and Pb) range between 52.71 and 55.34 ppm. Negative Rb–La scales are observed along with pronounced negative anomalies of Sr and Zr.

5. Discussion

5.1. Geotectonic Settings and Conditions of Ras Serpentinite Formation

REE abundances are an indicator of the serpentinite protolith, with rocks relatively enriched in MREE and HREE classified as lherzolites, while those poor in MREE and HREE are classified as harzburgites (depleted rocks). Increasing MREE to HREE patterns are mostly indicative of harzburgites and dunites [1]. The Ras serpentinite accumulates the lowest amounts of REE among the Greek serpentinites, pointing to a rather depleted source rock (Figure 9d). This degree of depletion is typical of depleted mantle peridotites, i.e., harzburgites, especially when MREE–HREE patterns are flat and not concave [45]. The levels of such depletion are related to partial melting of the mantle, whereas REE-enriched phases are removed from the mantle. In the case of Ras, this melting ranges between 15 and 20% (Figure 9c), which is relatively higher than the rest of the cases examined in the literature. Such high melting levels are usually the effect of hydrous phases within subduction zones in forearc settings. These environments are often linked with Cr mineralization [9,46,47,48] and this is also evident in the west Tinos area [43]. Thessaly, Othris, and a portion of Tinos Green serpentinites and serpentinized lithologies have also experienced high partial melting levels, whereas E. Chalkidiki, Neolithic Avgi, E. Othris, and the rest of the Tinos Green samples experience lower partial melting levels (<15%), pointing to fertile protoliths or geotectonic environments away from hydrous forearc regions.

Moreover, Ras serpentinite plots between the FMQ and FMQ + 1 curves (Fayalite–Magnetite–Quartz buffer; oxygen fugacity), making it the most oxidized case after ‘Tinos Green’ and E. Othris. These levels of redox state are the effect of slab-derived fluids introduced in a subduction zone [1,2,49,50,51], an event that was not very potent in the cases of Neolithic Avgi, Othris and E. Chalkidiki. The effects of such fluids are evident from the positive anomalies in FME (Figure 9d; e.g., B, Pb, and P). Their concentrations, and especially those of B, Li, Pb, and Cs, are typical for subducted serpentinites (classification after [1]), in agreement with the Cyclades geology which was highly controlled by the HP–HT post-Tertiary subduction event [26,27,28]. Considering the Al contents of serpentine group minerals (Figure 9a) in Ras, the majority of analyses fall within the intersection of the lizardite–antigorite and lizardite fields, with one analysis within the antigorite field. This means that Ras serpentine analyses point to relatively lower metamorphic degrees, yet, as also seen by the spectrometric study (Figure 7), there is still a low yet significant antigorite component. Antigorite is expected to be encountered on Tinos due to the local geology, and indeed was also found in ‘Tinos Green’ [43]. The reason why antigorite represents a relict phase in Ras is due to the pervasive greenschist facies overprint during Oligocene–Miocene exhumation. During this event, antigorite was almost completely transformed into lizardite and disseminated magnetite [52,53]. Late-stage fracturing of the rock due to brittle deformation [43,54] led to the formation of discontinuities that were later filled with fibrous magnetite-free chrysotile. During the latter stage, no carbonate mineral incorporation was noted; hence, no significant Sr and Ba (Figure 9b) contents were introduced in the Ras serpentinite. This was not the case for the neighboring ‘Tinos Green’ and other Greek occurrences [2,6,7,12,14,16,43], which are carbonate enriched due to the percolation of metasomatic carbonate-enriched fluids.

These findings support the classification of Ras as a highly depleted subduction-related serpentinite. The geochemical and mineralogical traits—particularly the low REE content, high melting levels, and fluid-mobile element anomalies—highlight the distinct nature of this occurrence within the broader context of Greek and Aegean serpentinites.

5.2. Geochemical and Physical Fingerprint of the Ras Serpentinite

This section summarizes the combined mineralogical, geochemical, and physical properties of the Ras serpentinite, aiming to define its distinct material fingerprint for provenance and comparative analysis. Serpentinite is a common lithology globally and a key component of Greek ophiolites. However, each occurrence exhibits distinct characteristics when evaluated across multiple parameters, including petrography, physical properties, structural geology, mineralogy, mineral chemistry, geochemistry, metamorphic conditions, and local geological evolution.

Petrographic analysis is a fundamental step in both provenance and petrogenetic investigations, as it provides critical insights into the geological history and transformation processes of a rock. In the case of the Ras serpentinite, the rock is dominantly composed of lizardite, crosscut by fibrous chrysotile veins measuring consistently between 1.5 and 2 mm in width. The lizardite matrix preserves relict phases of antigorite, minnesotaite—here reported for the first time in Greece—and magnetite, which forms both as pseudomorphs after chromite and as disseminated grains. Notably, these phases are entirely absent from the chrysotile veins, clearly identifying chrysotile as a later-stage, post-lizardite mineral phase.

Studies of the currently quarried ophicalcite describe a serpentinite with different mineralogical and textural traits despite being genetically and geotectonically linked to the Ras serpentinite [43]. In this case, serpentinite retains significant amounts of flaky antigorite, transforming into lizardite, as well as primary chromite altered into magnetite; other minerals in amounts <2% include talc and chlorite. Serpentinites from the neolithic Avgi still retain relict primary minerals such as olivine, pyroxene and plagioclase within a serpentine matrix [14]. These minerals are an indication of a relatively fertile protolith enriched in modal clinopyroxene (cpx; Figure 9c). Magnetite, in the case of Avgi, was also noted filling veins. Regarding the Thessalian stone, all three serpentine varieties were found, with antigorite being the main component, which mostly presents random crystallization [16]. Primary chromite is retained and altered into magnetite, chrysotile has a similar occurrence to that of Ras, i.e., as late-stage veins; other common minerals are epidote, tremolite, and talc. The voluminous serpentinites of Thessaly and Othris demonstrate mineralization of antigorite and lizardite, respectively. In the first occurrence, primary features of the protolith have been completely obliterated, whereas in the second case, one can notice relics of texture and minerals.

The crystallization of antigorite is linked with prevalent conditions of metamorphism during subduction. In the case of the Ras, the transition from subduction to exhumation formed thin thrusts of highly deformed serpentinite bodies with extensive retrograde metamorphism, providing uniqueness since, based on the literature data, no such rocks were in significant portions with these textural features [52]. These highly deformed bodies developing within shear zones (Figure 3e) were also described further NW in ‘Tinos Green’ [43], with characteristicly poor levels in carbonates ophicalcites when compared to the ‘Tinos Green’. Interestingly, the limited tectonic zones of the Ras serpentinite became quarry targets due to the unique texture produced by the aforementioned processes. Regarding the physical properties, to our knowledge there have been no efforts in applying FTIR, BET, or crystallinity analyses to serpentinites from Greece or the Aegean region.

Together, these parameters form a comprehensive fingerprint of the Ras serpentinite. Applying the same methodological framework to other serpentinites will enable effective comparison and support future provenance investigations.

5.3. Ras Serpentinite: Extraction, Function, and Chronology

Stone extraction techniques have long been a testament to human ingenuity and adaptability in utilizing natural resources. The technology of a quarry pick for making vertical extraction trenches was already in use during classical antiquity [55]. The precise placement of wedges and consideration of rock faults and crack lengths highlight the craftsmens’ expertise in managing the stone’s natural features to optimize extraction [56,57,58,59,60]. Regarding Pit B, this was probably exploited in a previous or later period or for a different purpose, since there are no organized extraction marks.

It is plausible that the Ras quarry functioned as a trial exploitation site, later abandoned due to the serpentinite’s moderate mechanical strength [61]. This is supported by the Point Load Strength Index (Is(50)) values, which classify the material as medium strength, and the small block dimensions observed (45 × 45 × 50 cm). Similar exploratory quarrying fronts are known from other historical contexts, where extraction ceased once unfavorable material properties were recognized. A notable example is the failed attempt at Baba Dağ, where challenging terrain and logistical difficulties ultimately led to the abandonment of the site despite initial extraction efforts [62]. The combination of organized extraction marks, minimal quarry infrastructure, and absence of artefacts aligns with such a scenario and may explain why this material did not enter widespread historical use.

The absence of known artifacts linked to the Ras serpentinite does not necessarily indicate a lack of use. It may rather reflect the current state of archaeological exploration and reporting. This study presents, for the first time, a thorough scientific characterization of this specific serpentinite source, which is essential for any future provenance studies that may aim to match artifacts to specific quarries. The current work serves to raise scholarly awareness of this source and lay the groundwork for future provenance analysis.

The ophicalcite (‘Tinos Green’ or ‘verde di Tinos’) is commonly used as façades, tiles, decorative objects, and architectural constructions (columns, vases, etc.). Contemporary applications of serpentinite are mostly as aggregate material in concrete for construction and conservation/restoration of architectural heritage [15,63,64], but the low strength (<71 MPa indirect UCS) of the Ras serpentinite sets a limit in its potential applications as an architectural element and in the production of sizable extracted blocks (only reaching 45 × 45 × 50 cm). However, as seen in similar lithologies in Spain and Italy [65,66], the ornamental value (color and texture) seems to bypass both the rock’s deterioration trend and general mechanical limitations. The local geological features and discontinuities (foliation, schistosity, faults, mylonitic zones, and shear zones) have greatly affected the latter. Moreover, the physicomechanical properties and behavior of serpentinites are negatively influenced by the degree of their serpentinization and by their mineralogical and textural characteristics [67]. If Ras serpentinite was used as some sort of façade, then the maximum face to be covered is expected to be <2025 cm² with each rock usage. However, due to the low strength of Ras serpentinite, it is thought to have been used as smaller fragments. The absence of accompanying archaeological finds limits further functional hypotheses, though the observed features provide a comprehensive view of ancient quarrying methodologies and their adaptation to material constraints [38].

Lichenometry provides the only viable chronological indicator in the absence of stratified archaeological material and supports the historical contextualization of quarrying activity. Likewise, the mechanical properties offer insight into the workability and potential use of the material—key information in both archaeometric and geoheritage applications. The Ras serpentinite ‘provenance fingerprint’ is summarized in Table S5.

6. Conclusions

This study presents the first complete characterization of serpentinite from the Ras ancient quarry (Tinos, Greece), integrating mineralogical, geochemical, physical, and archaeometric data to define a robust provenance fingerprint. The Ras serpentinite is classified as a highly depleted subduction-related lithology, derived from a harzburgitic protolith subjected to high degrees of partial melting and influenced by oxidizing, slab-derived fluids—conditions reflected by its low REE content, positive fluid-mobile element anomalies, and unique mineral assemblage. Petrographic and spectroscopic analyses reveal a lizardite matrix, crosscut by a late-stage chrysotile network, with relict antigorite and magnetite, and the first documented occurrence of minnesotaite in Greece. These features, along with the absence of carbonate enrichment, distinguish Ras serpentinite from nearby ophicalcite sources such as ‘Tinos Green’, establishing a robust framework for archaeometric fingerprinting. The quarry extraction techniques and localized geological constraints suggest the selective use of serpentinite, likely in small façades or fragments due to its moderate strength. Lichenometric data provide a dating of the last quarrying activities around the 13th century A.D. These findings establish Ras serpentinite as a reference source for future provenance studies and underscore the value of the proposed multi-method approach in archaeometric investigations of ancient stone materials.