Introduction
The terms ‘opal’ and ‘chalcedony’ correspond to amorphous and nano-crystalline types of silica (SiO2·nH2O), respectively (Flörke et al., Reference Flörke, Graetsch, Martin, Röller and Wirth1991; Čurlík and Forgáč, Reference Čurlík and Forgáč1996; Gliozzo, Reference Gliozzo2019). When viewed optically using cross-polarized light, the former appears isotropic (with common reniform and globular fabrics), whereas the latter is anisotropic, with first-order interference colours typical of quartz, but also marked by peculiar fibrous/bundle-like to spherical arrangements of the crystals. In addition, both types of silica are hydrated and/or hydroxylated, as their structures tend to retain variable amounts of water species, i.e. molecular water (H2O) and/or silanol-group water (Si–OH), that can occupy pores, interstitial spaces and/or crystal boundaries (Graetsch et al., Reference Graetsch, Flörke and Miehe1987; Boboň et al., Reference Boboň, Christy, Kluvanec and Illášová2011). The total water content (H2O+Si–OH) is quite variable and ranges from ∼1–2 wt.% in chalcedony and from 2–10 wt.% (up to 20 wt.%) in opaline silica (Graetsch et al., Reference Graetsch, Flörke and Miehe1985; Day and Jones, Reference Day and Jones2008; Li et al., Reference Li, Guo, Wang and Liao2022; Powolny et al., Reference Powolny, Dumańska-Słowik, Szczerbowska-Boruchowska and Woszczyna2024).
On the basis of X-ray diffraction (XRD) and spectroscopic (Raman and FTIR) characteristics, opaline silica can be subdivided into three main types depending on the degree of its disorder. These include: (1) opal-A (highly disordered/amorphous); (2) opal-C (containing well-crystallized cristobalite); and (3) opal-CT (with stacked tridymite-cristobalite domains) (Jones and Segnit, Reference Jones and Segnit1971; Elzea et al., Reference Elzea, Odom and Miles1994; Ghisoli et al., Reference Ghisoli, Caucia and Marinoni2010; Curtis et al., Reference Curtis, Gascooke, Johnston and Pring2019). The nature of opal-CT is the subject of controversy as two different structural models have been proposed, i.e. crystalline and paracrystalline. According to the widely accepted view, opal-CT is made up of disordered interlayering of tridymite and cristobalite units (Flörke, Reference Flörke1955). In turn, the paracrystalline model interprets opal-CT as a disordered material similar to amorphous opal-A, containing only a small volume of crystalline stacking units, mainly of tridymitic nature (Smith, Reference Smith1998; Wilson, Reference Wilson2014; Curtis et al., Reference Curtis, Gascooke, Johnston and Pring2019). A hypothetical opal-T or -TC is also envisaged (Ghisoli et al., Reference Ghisoli, Caucia and Marinoni2010), whilst a low birefringent fibrous (chalcedony-like) variety of opal-CT with pseudo-orthorhombic to orthorhombic symmetry is referred to as lussatite (Mallard, Reference Mallard1890; Jones and Segnit, Reference Jones and Segnit1971; Flörke et al., Reference Flörke, Graetsch and Jones1990; Schindler et al., Reference Schindler, Fayek, Courchesne, Kyser and Hawthorne2017; Fröhlich, Reference Fröhlich2020).
In general, the formation of opaline silica occurs in a wide variety of low-temperature geological settings and can involve both biogenic (diatom- and radiolarian-forming) and non-biogenic activity. The latter can be derived from the circulation of hydrothermal fluids in magmatic to sedimentary rocks, continental weathering, followed by the influx of diluted meteoric waters and precipitation of silica as sinters (geyserites) in the vicinity of hot springs (Chauviré et al., Reference Chauviré, Rondeau and Mangold2017; Pineau et al., Reference Pineau, Chauviré and Rondeau2023). The replacement-related nature of opal has also been asserted for silicified serpentinites and organic materials (e.g. wood) (Lacinska and Styles, Reference Lacinska and Styles2013; Behera and Nayak, Reference Behera and Nayak2023; Mustoe, Reference Mustoe2023).
Chalcedonic silica is composed of fibre-aligned α-quartz crystallites (∼50–100 nm in size), accompanied by a variable admixture of moganite (Heaney and Post, Reference Heaney and Post1992; Heaney et al., Reference Heaney, Veblen and Post1994; Schmidt and Fröhlich, Reference Schmidt and Fröhlich2011). The latter represents a metastable but crystalline polymorph of SiO2 found in relatively young siliceous deposits, in common association with silicified evaporate sequences (Heaney, Reference Heaney1995). As the XRD and spectroscopic characteristics of chalcedony are nearly identical to those of quartz, its further subdivision into so-called length-fast (LF) and length-slow (LS; also known as quartzine) varieties relies on microscopic observations (Cady et al., 1998; Powolny et al., Reference Powolny, Dumańska-Słowik, Szczerbowska-Boruchowska and Woszczyna2024). These subtypes are marked by the crystallographic c axis situated perpendicular and parallel to the fibre elongation, respectively. The LF chalcedony is also characterized by normal (non-twisted) and twisted (‘zebraic’) characters depending on whether the fibres are twisted or not around the c [110] axis. Overall, chalcedony can form not only via the transformation and textural maturation of pristine opaline silica (Kastner et al., Reference Kastner, Keene and Gieskes1977), with or without moganite as a transitional phase (El-Hawat et al., Reference El-Hawat, El-Ghali, McLaren and Kemp2021), but can also crystallize directly from hydrothermal fluid and/or gel-like medium (Heaney, Reference Heaney1993; French et al., Reference French, Worden and Lee2013; Howard and Rabinovitch, Reference Howard and Rabinovitch2018). Chalcedony represents quite a common hydrothermal- and cementation-related phase in volcanic and sedimentary rocks, where it commonly appears as the multi-coloured banded variety termed agate (Götze et al., Reference Götze, Möckel, Kempe, Kapitonov and Vennemann2009; Dumańska-Słowik et al., Reference Dumańska-Słowik, Natkaniec-Nowak, Wesełucha-Birczyńska, Gaweł, Lankosz and Wróbel2013, Reference Dumańska-Słowik, Powolny, Sikorska-Jaworowska, Gaweł, Kogut and Poloński2018; Powolny et al., Reference Powolny, Dumańska-Słowik, Sikorska-Jaworowska and Wójcik-Bania2019a; Natkaniec-Nowak et al., Reference Natkaniec-Nowak, Dumańska-Słowik, Gaweł, Łatkiewicz, Kowalczyk-Szpyt, Wolska, Milovská, Luptáková and Ładoń2020; Başibüyük et al., Reference Başibüyük, Gürbüz and Kaydu Akbudak2023). The coexistence of opal and chalcedony within single agate geodes, though not ubiquitous, has been reported in Moroccan agate deposits (Dumańska-Słowik et al., Reference Dumańska-Słowik, Natkaniec-Nowak, Wesełucha-Birczyńska, Gaweł, Lankosz and Wróbel2013). Both can act as precursors for the formation of recrystallization-related quartz microtextures, such as feathery and mosaic, respectively (Powolny et al., Reference Powolny, Dumańska-Słowik, Szczerbowska-Boruchowska and Woszczyna2024).
Chalcedony and opal can be colourless, but commonly show red, orange, blue, purple and/or black colours. The colouration mechanisms are linked to the presence of specific solid inclusions (carbonaceous material, chrysocolla, Mn-Fe-oxides, sepiolite, etc.), as well as physical effects such as Mie and Rayleigh scattering (Dumańska-Słowik et al., Reference Dumańska-Słowik, Natkaniec-Nowak, Kotarba, Sikorska, Rzymetka, Loboda and Gaweł2008; Başibüyük, Reference Başibüyük2018; Götze et al., Reference Götze, Möckel and Pan2020; Ye and Shen, Reference Ye and Shen2020; Lorenzi et al., Reference Lorenzi, Zullino, Gagliardi, Prosperi, Paleari and Adamo2022; Powolny et al., Reference Powolny, Dumańska-Słowik, Szczerbowska-Boruchowska and Woszczyna2024; Welman-Purchase et al., Reference Welman-Purchase, Wicht, Miller and Roelofse2024). These effects can eventually be enhanced/modified by crystallite size and water content. One of the most valuable chalcedonic and opal species involves chrysoprase, prase opal and chrome chalcedony (also known as ‘mtorolite’ and ‘aquaprase’), all are green or green-blue coloured with variable tone and saturation (Eggleton et al., Reference Eggleton, Fitz Gerald and Foster2011; Başibüyük et al., Reference Başibüyük, Akbudak and Gürbüz2020; Feral, Reference Feral2022; Monico et al., Reference Monico, Cantaluppi, Diella, Gatta, Adamo, Fumagalli and Marinoni2024a, Reference Monico, Adamo, Diella, Melas, Prosperi and Marinoni2024b). Both chrysoprase and prase opal are rich in a Ni-bearing clayey pigment (‘kerolite–pimelite’, willemseite) and/or Ni grafted on the surface silica minerals (Sojka et al., Reference Sojka, Witkowski, Zabiński, Dyrek and Bidzińska2004; Eggleton et al., Reference Eggleton, Fitz Gerald and Foster2011; Čermáková et al., Reference Čermáková, Hradil, Bezdička and Hradilová2017). They are commonly reported in association with altered ultramafic rocks such as weathered (silicified) serpentinites, saprolites and/or laterites (Shigley et al., Reference Shigley, Laurs and Renfro2009; Caucia et al., Reference Caucia, Marinoni, Ghisoli and Leone2016). In contrast, Cr-rich chalcedony/opal is less common in nature, and its actual colouration mechanism (inclusions vs. extra-framework atom) remains poorly resolved.
In the vicinity of Tokat, a town in northern Turkey, silicification has led to the formation of widespread agate (including the rare surprise-egg type), silicified wood and jasper (Kaydu Akbudak et al., Reference Kaydu Akbudak, Başıbüyük and Gürbüz2021; Başibüyük et al., Reference Başibüyük, Gürbüz and Kaydu Akbudak2023; Yüzbaşıoğlu and Kaydu Akbudak, Reference Yüzbaşıoğlu and Kaydu Akbudak2024). Here we report for the first time the abundance of unusual listvenite rocks hosting the coexisting green-coloured (Cr-bearing) opaline silica, saddle dolomite and sulfide mineralization. Using a combined mineralogical and spectroscopic study, our objective was to resolve the structural state of opal and the nature of the agents responsible for the intricate green colouration. This study also highlights the nature of two-stage hydrothermal silicification–carbonatization in ophiolite-related ultramafic rocks, i.e. listvenitization followed by crystallization of chalcedonic bodies (vein agate), as well as their relevance to geodynamic evolution in the study area.
Geological background
The study area, the Tokat Massif, is part of the South-Central Pontides in northern Turkey and preserves the record of the closure of both Paleo- and Neo-Tethyan ocean basins (Şengör and Yilmaz, Reference Şengör and Yilmaz1981; Yilmaz et al., Reference Yilmaz, Serdar, Genc, Yigitbas, Gürer, Elmas, Ylldirim, Bozcu and Gürpinar1997). Here, metamorphic, sedimentary and magmatic rocks belong to three substantial tectonic units, including the Akdağmadeni Group, the Yeşilırmak Group and the Tokat Group (Yilmaz and Yilmaz, Reference Yilmaz and Yilmaz2004; Catlos et al., Reference Catlos, Huber and Shin2013) (Fig. 1a). The former consists mainly of crustal-derived metamorphic rocks (gneiss, amphibolite, schist, quartzite and marble), which were assembled during the Alpine Orogeny and preserve both regional metamorphism up to amphibolite facies and subsequent dynamic metamorphism (Yilmaz and Yilmaz, Reference Yilmaz and Yilmaz2004). The second unit involves the Cretaceous ophiolitic melange of the North Anatolian Ophiolite Belt (Tekeli, Reference Tekeli1981). The Tokat group can be subdivided into the Turhal Metamorphics and Devecidağ Melange. The Turhal Metamorphics consist of a Permian–Triassic metavolcanic-sedimentary succession that has been metamorphosed under greenschist-facies conditions and consists of phyllite, schist, metatuff and metabasite (Yilmaz and Yilmaz, Reference Yilmaz and Yilmaz2004). The Devecidağ Melange includes limestone and ophiolite blocks (Silurian to Triassic) hosted by metaclastic and metavolcanic matrices. The ophiolite blocks, of serpentinized peridotite, are believed to represent the Pre-Liassic component of the ancient oceanic crust (Tekeli, Reference Tekeli1981). The serpentinite formation was later overprinted by abundant listvenitization along the Izmir–Ankara–Erzincan Suture Zone (IAESZ) due to the activity of hydrothermal fluids related to the granite intrusions in the vicinity of the study area (Sarifakioglu, Reference Sarifakioglu2023). The youngest units in the study area are represented by Eocene–Pliocene sedimentary rocks and Quaternary alluviums (Sümengen, Reference Sümengen2013). The opal-bearing listvenite bodies and associated vein agate studied occur within Permian–Triassic phyllites of the Tokat group of the eastern extension of Karakaya complex (Okay and Göncüoǧlu, Reference Okay and Göncüoǧlu2004; Tetiker et al., Reference Tetiker, Yalçın, Bozkaya and Göncüoğlu2015) (Fig. 1b).
Figure 1. (a) Geological map of northern Turkey in the vicinity of Tokat (modified after Yilmaz and Yilmaz, Reference Yilmaz and Yilmaz2004); (b) main lithologies exposed in the study area together with the locations of sampling sites (modified after Sümengen, Reference Sümengen2013).
Materials and methods
Sample description and field observations
Listvenite-rich bodies are observed as veins reaching 10 cm in width and lenses, up to 1 min length, in the alteration zone of serpentinite located at the tectonic contact with Permian–Triassic phyllites (Fig. 2a–c). Their exposures cover an area of ∼0.6 km2. The rocks analysed feature serpentinite-like (mesh/network-like) fabrics resulting from the presence of numerous white to brownish veins comprising carbonate minerals (dolomite), followed by bluish to bluish-white banded chalcedony (agate) (Fig. 2d). The latter commonly exhibits a botryoidal fabric and can also occur as individual bodies, i.e. not associated with dolomite. Green opal-rich areas appear to be massive, homogeneous and nearly opaque with waxy to oily lustre. Pyrite is visible in hand specimen and randomly disseminated throughout the opal-rich matrix of listvenite (Fig. 2d).
Figure 2. (a) Field exposure and (b–c) field views of green-coloured opal-rich dolomite-bearing listvenite containing agate veins, found in the alteration zone close to the tectonic contact with Permian-Triassic schists (phyllites); (d) a closer view of the listvenite distinguished by serpentinite-like fabrics and the presence of bluish agate veins.
Analytical techniques
Polarizing microscopy and cathodoluminescence (OM-CL)
Polished thin sections were examined using an Olympus BX-51 polarizing microscope equipped with 4× to 50× objective lenses (Faculty of Geology, Geophysics and Environmental Protection of AGH University of Krakow, Poland). Photomicrographs were taken using an Olympus DP-12 camera combined with Stream Essential software. Cathodoluminescence observations were done at the Institute of Earth Sciences (University of Silesia in Katowice, Sosnowiec, Poland), using a CITL Mk5 cold-cathode CL device that operated under 10–15 kV voltage and up to 300 µA current.
Powder X-ray diffraction (PXRD)
Standard PXRD analyses of the powdered opal-rich fragments of listvenite (green colouration) were carried out using a Rigaku SmartLab 9kW X-ray diffractometer (Faculty of Geology, Geophysics and Environmental Protection of AGH University of Krakow, Poland) equipped with curved crystal graphite-monochromatised CuKα radiation (λ = 1.54 Å). The PXRD patterns for air-dried, glycolated and heated (550°C) samples were recorded in the range 2–75°2θ. This was required because of the presence of clay-group species within the opaline silica. The step size, counting time, voltage and current were 0.05°, 1 s/step, 45 kV and 200 mA, respectively. The side-loading method was used during the measurements. Phase identification was provided by xRayan (Marciniak and Diduszko, Reference Marciniak and Diduszko1994).
Scanning electron microscopy (SEM) supported by electron microprobe analysis (EMPA)
The analyses were undertaken on carbon-coated thin sections, using a JEOL Super Probe JXA-8230 instrument (AGH University of Krakow-KGHM Polska Miedź S.A.) that operated under the following analytical conditions: high-vacuum mode; acceleration voltage of 15 kV; and a beam current of 20 nA. The instrument is equipped with both energy-dispersive (EDS) and wavelength-dispersive spectrometers (WDS). The former was used for preliminary observations and phase identification, whereas the latter provided the composition of dolomite and element-distribution maps.
For dolomite analysis the elements, standards, analytical crystals and detection limits were as follows: Ca – diopside (CaMgSi2O6), PETJ, 270 ppm; Mg – diopside (CaMgSi2O6), TAPH, 380 ppm, Fe – hematite (Fe2O3); LIFH, 360 ppm; Mn – rhodonite (CaMn3MnSi5O15), LIFH, 350 ppm; Sr – celestine (SrSO4), PETL, 1900 ppm; and Na – albite (NaAlSi3O8), TAP, 300 ppm. The peak count time of 10 s, background time of 5 s and beam size of 5 µm were used for WDS measurements.
Element-distribution WDS maps including Ca, Mg, Fe, Al and Cr were obtained from a rectangular area of 280 µm × 345 µm within an opal-rich area, using a 15 kV acceleration voltage and with a dwell time of 100 ms per pixel (i.e. 1 µm × 1 µm). Characteristic Kα lines of particular elements, measured at PETJ (Ca), TAP (Mg), LIFH (Fe), TAPH (Al) and LIFL (Cr) analytical crystals, were used to determine the semi-quantitative concentration of particular elements.
Raman spectroscopy
The analyses were conducted using a ThermoScientific DXR Raman microscope (Faculty of Geology, Geophysics and Environmental Protection of AGH University of Krakow, Poland) that operated in confocal mode and back-scatter geometry. A laser excitation wavelength of 532 nm was used. The analytical conditions were as follows: 25 µm pinhole, estimated spectral resolution of 3 cm–1 and 1 min exposure time. A deconvolution procedure proposed by Götze et al. (Reference Götze, Nasdala, Kleeberg and Wenzel1998) was applied to establish the concentration of moganite (wt.%), i.e. moganite- and quartz-related bands (501 and 464 cm–1, respectively) were decomposed into two Lorentzian-shaped bands using OMNIC Software, followed by a constant background in the range 600–300 cm–1. Subsequently, band integrals I(501) and I(464), referred to as full-width at half-maximum (FWHM), multiplied by the band height, were obtained. The FWHM was also corrected for the specific apparatus prior to the calculations (Irmer, Reference Irmer1985). Finally, the I(501)/I(464) ratios were plotted on the calibration curve provided by Schmidt et al. (Reference Schmidt, Bellot-Gurlet, Leá and Sciau2013) (Supplementary Fig. S1).
Fourier-transform infrared spectroscopy (FTIR)
The crushed and powdered sample of opaline silica (listvenite-hosted) was analysed in transmission mode using a NICOLET 6700 ThermoScientific instrument (Faculty of Geology, Geophysics and Environmental Protection of AGH University of Krakow, Poland). The ThermoScientific OMNIC Software was also used for spectral analysis (e.g. band positions). The analyses were done in ambient conditions, using the potassium bromide (KBr) pellet method, where 200 mg of KBr was mixed with ∼3 mg of the sample and then pressed into a transparent pellet. The scanning range, spectral resolution and number of scans were 4000–400 cm–1, 4 cm–1 and 64, respectively.
Ultraviolet-Visible-Near Infrared spectroscopy (UV-VIS-NIR)
The analyses were performed using a GemmoSphere UV-VIS-NIR spectrometer located at the Gemological Laboratory of the Faculty of Geology, Geophysics and Environmental Protection of AGH University of Krakow. The acquisition spectra were measured in the range 360–950 nm and used a Xe-boosted light source. The spectral resolution was ∼1.3 nm.
X-ray fluorescence
The XRF analyses were performed to determine the major composition of powdered blue chalcedony and opal-dolomite matrix of listvenite samples using a Thermo Brand ARL 2315 Model XRF device at the Geochemistry Laboratory of the General Directorate of Mineral Research and Exploration in Ankara, Turkey.
Inductively Coupled Plasma Mass Spectrometry (ICP-MS)
The ICP-MS analyses were performed to determine the trace element composition of the green opal (listvenite-hosted) and blue chalcedony. The samples were dissolved using perchloric acid (HClO4), hydrofluoric acid (HF), hydrochloric acid (HCl) and nitric acid (HNO3) and then analysed with the Analytik Jena Brand Plasmaquant MS apparatus located at the General Directorate of Mineral Research and Exploration in Ankara, Turkey.
Results
Petrography and cathodoluminescence supported by scanning electron microscopy
Opal-rich listvenite exhibits unusual fibrous/filamentous microtextures. In this study, the silica matrix exhibits first-order grey interference colours and positive (+) signs of elongation, coupled with dull blue cathodoluminescence (CL) emission (Fig. 3a,b). The chalcedony-like (i.e. fibrous) appearance of opaline silica indicates the presence of lussatite. The lussatite is further colourless to brownish in plane-polarized light and also tends to mantle larger dolomite crystals (Fig. 3c). Another generation of opaline silica (non-fibrous) occurs as secondary veinlets (up to ∼0.1 mm in thickness) with brownish colours and isotropic character, followed by relatively bright-blue CL colours (Fig. 3a). Dolomite is represented by the so-called saddle variety, that is with peculiarly curved crystal faces and sweeping extinction, that reaches ∼1 mm in size. The crystals are characterized by roughly rhombohedral habit (Fig. 3c,d), with cloudy non-luminescent cores and limpid, non-luminescent to weak red-luminescent rims. The saddle dolomite commonly has a tabular/elongated morphology where associated with chalcedonic vein-filling silica (agate).
Figure 3. (a) Photomicrograph and the corresponding CL image (inset, upper left) of lussatite-rich (Op-Lus) matrix (green in the macroscale) intersected by saddle dolomite (sDol), non-twisted length-fast (LF) chalcedony (nCha(-)) and another generation of opal (Op) filling the vein in the central part of the image; (b) example of an enlarged view of fibrous opaline silica (lussatite) accompanied by later opal (isotropic) and quartz-rich (Qz) vein. The inset figure, taken with a gypsum-plate inserted from the lower left, accounts for a positive elongation of the fibres. (c,d) Euhedral to subhedral saddle dolomite with cloudy core and limpid rim surrounded by a lussatite-rich matrix. Note the curved (gneiss-like) alignment of opaline silica around larger saddle dolomite (blue arrows). PX – crossed polars; PPL – plane-polarized light; CL – cathodoluminescence; GPL – gypsum plate inserted.
The interior of the dolomite was rich in Fe and relatively abundant in micropores (Fig. 4a), whereas limpid areas revealed oscillatory zonation owing to the variations in Fe and Mn contents. The boundary between dolomite (the external part of the veins cross-cutting the opaline matrix) and agate bodies (innermost regions) displays truncated growth lines and embayed dissolution contacts with silica (Fig. 4a). The opal-rich matrix hosts spinel-group minerals that probably belong to the chromite–magnetite and chromite–hercynite series (Fig. 4b). Pyrite appears to be homogeneous in back-scattered electron (BSE) SEM images but contains minute cobaltite and larger more common enargite inclusions (Fig. 4c,d). The latter were confirmed by Raman spectroscopy as their composition is similar to that of tennantite-Cu.
Figure 4. SEM-BSE images of the listvenite samples. (a) The boundary between saddle dolomite and chalcedony (agate) vein. Note the Fe-induced growth-related zonation manifested by dark and light areas, as well as dissolution-erosion-related features that indicate silicification of dolomite (blue arrows). (b) Relics of spinel-group minerals such as chromite-magnetite (Chr-Mag) and chromite-hercynite (Chr-Hc) surrounded by opaline silica and saddle dolomite. (c) Pyrite (Py) hosting an enargite (Eng) inclusion. (d) Pyrite impregnated with cobaltite (Cbt).
Chalcedony (agate) veins in opal-bearing listvenite exhibit complex microtextural characteristics, as they contain wall-lining to spherulitic length-fast (LF) chalcedony represented by twisted (‘zebraic’) and non-twisted (‘normal’) types, accompanied by only minor amounts of quartzine (length slow) intergrowths (Fig. 5a,b). Prismatic drusy megaquartz (>20 µm; up to ∼1 mm in length), as well as very thin (<0.1 mm) opaline bands, are sandwiched between chalcedony-rich regions (Fig. 5c,d). The megaquartz exhibits a local flamboyant appearance, i.e. a crystalline core rimmed by chalcedony. It exhibits both reddish and bluish CL and/or patchy zoning represented by irregular reddish and blue areas within single crystals. Chalcedonic spherules display polygonal fabrics in some places, and also exhibit reddish to pinkish CL emissions. This style of emission is common in nano-crystalline silica and can be assigned to the presence of NBOHC (non-bridging oxygen hole centres) associated with silanol-group water (Götze et al., Reference Götze, Plötze, Fuchs and Habermann1999). As shown in Fig. 5e,f, a single (oval-shaped) chalcedonic (partially recrystallized) spherule illustrates the presence of an irregular rectangular-like zonation pattern resembling growth lines of dolomite. As presented in Fig. 5g,h, chalcedonic silica tends to lose its microscopically visible fibrosity and transforms into more or less granular quartz, with feathery and/or mosaic (interpenetrating) fabrics.
Figure 5. Photomicrographs paired with CL images of chalcedony (agate) and saddle dolomite. (a,b) Megaquartz (Qz) followed by twisted (partially recrystallized) chalcedony (tCha(-)), non-twisted chalcedony (nCha(-)) and quartzine (Qzn(+)). Note the growth-related pattern of chalcedony and patchy zoning of megaquartz visible in CL. The inset photomicrograph (upper left) was taken with a gypsum plate inserted. (c, d) The inner part of chalcedony (agate) vein occluded by megaquartz with flamboyant outline (yellow arrows), opaline silica (blue CL emission) and twisted LF chalcedony with polygonal fabrics (visible under the polarizing microscope). Note that botryoidal fabrics of LF chalcedony revealed by cathodoluminescence (blue arrows) do not conform to polygonal fabrics of chalcedony visible in polarized light. (e,f) Recrystallization of chalcedony (or opal?) into mosaic (mQz) and feathery quartz (fQz) that resulted in the loss of its optically visible fibrosity. Note the presence of a zigzag (rhombic-like) pattern (ghost microtexture of dolomite?) that does not follow the spherulitic arrangement of chalcedony visible under polarizing microscopy (crossed polars) and might have resulted from the replacement of dolomite by silica (yellow arrows). (g,h) The boundary between vein-filling dolomite, followed by a chalcedony (agate) body comprising twisted chalcedony (red CL) and blue luminescent megaquartz. Note the presence of red luminescent growth lines in limpid dolomite (yellow arrows) and quartz penetrating through larger dolomite crystals (upper left).
Powder X-ray diffraction (PXRD)
The PXRD analysis of the green-coloured areas of the listvenite reveals the presence of marker reflections for opal-CT (i.e. 4.33 and 4.11Å), followed by a relatively weak peak at 2.51 Å (Fig. 6). According to the literature (e.g. Elzea et al., Reference Elzea, Odom and Miles1994; Ilieva et al., Reference Ilieva, Mihailova, Tsintsov and Petrov2007; Ghisoli et al., Reference Ghisoli, Caucia and Marinoni2010; Fröhlich, Reference Fröhlich2020), the main XRD peak for opal-CT falls between 4.12 and 4.06 Å and shifts towards lower values, ∼4.0 Å, when opal-C becomes a dominant component. The position of the main peak at 4.11 Å, thus marks the presence of highly disordered opal CT with tridymite-dominant characteristics (‘Opal-T’) (Ghisoli et al., Reference Ghisoli, Caucia and Marinoni2010). The full-width at half-maximum (FWHM) of this peak (Fig. 6) attains a relatively low value of 0.2 Å, typical of lussatite, according to Schindler et al. (Reference Schindler, Fayek, Courchesne, Kyser and Hawthorne2017). The reflection at 3.35 Å can be assigned to α-quartz. The presence of dolomite and pyrite are proven by the signals at 2.91 and 2.71 Å, respectively. Low-angle regions of the X-ray pattern (below ∼10°2θ) reveal a relatively weak but broad peak centred at ∼15.65 Å, which shifts towards ∼17 Å after saturation with ethylene glycol and collapses due to heating up to 560°C. This observation suggests the presence of swelling clays (i.e. smectite group). The occurrence of minor weak diffraction at 1.50 Å is consistent with its dioctahedral character typical of the montmorillonite–beidellite series (Emmerich, Reference Emmerich2013). However, the trioctahedral clay-group species (e.g. saponite) cannot be excluded entirely as its eventual diagnostic peak at 1.54 Å coincides with the quartz-related (211) reflection.
Figure 6. PXRD patterns for the opaline (lussatite-rich) matrix of listvenite in the range of 2–40°2θ, 58–64°2θ (inset image, left) and 15–25°2θ (inset image, right). The full width at half maximum (FWHM) of the main opal-related peak was indicated. The AD, GL and HT symbols refer to the curves obtained for air-dried, glycol-solvated and heated (560°C) powders, respectively. Note the presence of marker peaks for α-quartz (Qz), dolomite (Dol) and pyrite (Py), as well as (001) and (060) reflections of smectite-group clays (Sm), found at 15.65(AD)/17.30(GL) Å and 1.50 Å, respectively.
Element-distribution mapping
The distribution of Al, Mg, Fe, Cr and Ca within selected fragments of green-coloured opaline silica (lussatite) elucidated the abundance and character of discrete clay-group species disseminated within it. Even though the given region is still fairly homogeneous under SEM-BSE (with recognizable fibrous/felted microtextures), there is a robust positive correlation among five elements, i.e. Cr, Fe, Mg, Ca and Al (Fig. 7). This type of relationship has confirmed the pigmentation of opaline silica with smectite that has been previously identified via PXRD measurements. As the Cr content appears to be relatively large, this phase can eventually be identified as either volkonskoite (ideally Ca0.3[CrMgFe]2[Si,Al]4O10[OH]2·H2O) or a Cr-bearing smectite of the montmorillonite–beidellite series, although direct spot analyses are precluded owing to the small crystal size.
Figure 7. SEM-BSE image and WDS element-distribution maps showing the distribution of Al, Mg, Fe, Cr and Ca in the rectangular area (280 µm × 345 µm) of green-coloured opal (lussatite)-rich matrix of listvenite.
Dolomite composition
The composition of the saddle dolomite is presented in Table 1 and further depicted in an Fe–Mg–Ca ternary diagram (Fig. S2). The inner (cloudy) areas (see Fig. 3c,d), recognized as ferroan dolomite, are enriched in FeO (7.99–9.04 wt.%) with a stoichiometric Ca content (∼1.0 atoms per formula unit), coupled with traces of Mn (up to 0.23 wt.%). The empirical formula can be roughly approximated as Ca1.01(Mg0.75Fe0.23MnO0.01)(CO3)2. Limpid outer zones (ferroan to nearly pure dolomite), show variable compositions in terms of both Fe and Mn contents that range between 2.32–8.96 and <0.20 wt.%, respectively. The formula of limpid dolomiteis of a stoichiometric character on the basis of its Ca content (∼1.0 apfu), i.e. Ca0.94–1.04(Mg0.79–0.96Fe0.06-0.24MnO0–0.01)(CO3)2. There is also a well-developed positive correlation (R 2 = 0.79) between Fe and Mn contents in the saddle dolomite analysed (Fig. S2).
Table 1. Representative compositions of saddle dolomite (from EMPA) including inner (cloudy) and outer (limpid) areas of the crystals
* calculated from stoichiometry; apfu – atoms per formula unit
Raman spectroscopy
Examples of Raman spectra obtained for green-coloured opaline silica (lussatite) are presented in Fig. 8 and also mark the presence of opal-CT (Smallwood et al., Reference Smallwood, Thomas and Ray1997; Ilieva et al., Reference Ilieva, Mihailova, Tsintsov and Petrov2007; Ivanov et al., Reference Ivanov, Reyes, Fritsch and Faulques2011; Gouzy et al., Reference Gouzy, Rondeau, Gaudin, Louarn, La, Lebeau, Vinogradoff, Clodoré and Chamard-Bois2023). There is a strong and broad band centred at ∼360 cm–1 (with two more or less visible satellite bands at 418 and 265 cm–1) related to O–Si–O bending vibrations, followed by minor bands at 785 and 1063 cm–1 that can be both assigned to the symmetric Si–O–Si stretching vibrations (Ivanov et al., Reference Ivanov, Reyes, Fritsch and Faulques2011). The spectrum in the water-related range (Fig. 8, inset) comprises a broad and asymmetric band between 3800 and 3000 cm–1. This signal originates from water molecules (H2O) bound by both opaline silica and clay-group species (smectite). The presence of silanol-group water (Si–OH) is supported by the weak band at 965 cm–1 (Sodo et al., Reference Sodo, Casanova Municchia, Barucca, Bellatreccia, Della Ventura, Butini and Ricci2016), although another silanol-related marker signal at 3750 cm–1 is absent owing to the possible hydrogen bonding induced by large amounts of molecular water (Gouzy et al., Reference Gouzy, Rondeau, Gaudin, Louarn, La, Lebeau, Vinogradoff, Clodoré and Chamard-Bois2023). Finally, the spectrum (Fig. 8) contains a weak band at ∼705 cm–1, which can be attributed to the smectite-group species (e.g. Cr-montmorillonite, volkonskoite). This signal probably arises from Si–Ob–Si vibration modes (b refers to bridging oxygen) of phyllosilicates and chain silicates (Wang et al., Reference Wang, Freeman and Jolliff2015).
Figure 8. Examples of Raman spectra for green opaline silica (lussatite) of listvenite in the range 1400–150 cm–1. Note that the band at 705 cm–1 (*) appears to be due to the smectite impurity. The spectrum in the water-related region (4000–3000 cm–1) is included as an inset (upper left).
The most prominent Raman band of lussatite, found at ∼360 cm–1, can be further decomposed into at least five sub-bands according to the Voigt function (Fig. 9), with the maxima centred at ∼260, ∼300, ∼360, ∼428 and ∼485 cm–1. Using the intensity and full-width at half maximum (FWHM) of the decomposed representative spectra (Fig. 9), the so-called ξ and η parameters can be obtained. According to Ilieva et al. (Reference Ilieva, Mihailova, Tsintsov and Petrov2007) and Zhao and Bai (Reference Zhao and Bai2020), the ξ and η parameters reflect the participation of the tridymite component and the defect concentration of theoretical tridymitic and cristobalitic nanoregions, respectively. As the values of those parameters correspondingly range between 0.97–1.08 and 0.60–0.67, the tridymite component seems to prevail over the cristobalite one, although they probably share a similar degree of defectiveness.
Figure 9. Deconvolution of Raman spectra of opaline silica (lussatite) in the range 600–150 cm–1, followed by the values of so-called ξ and η parameters (see text for further details). The reference spectra of tridymite (T) and cristobalite (C), attached in the lower part of the image, were taken from the RRUFF database (Lafuente et al., Reference Lafuente, Downs, Yang and Stone2015), i.e. R090042 (tridymite) and R061107 (cristobalite).
Raman spectra obtained from chalcedonic bands with agate banding reveal the presence of coexisting α-quartz and moganite, the diagnostic bands of which (symmetric stretching-bending modes) appear at 464 cm–1 and 501 cm–1, respectively (Kingma and Hemley, Reference Kingma and Hemley1994). The latter can, however, partially overlap with the band originating from the silanol-group water (Schmidt et al., Reference Schmidt, Bellot-Gurlet, Slodczyk and Fröhlich2012). According to Fig. 10a–d, at least three zones (1–3) of different hues are tentatively distinguished within single agate bodies, where moganite addition had been estimated in the overall range of 24–46 wt.% (see Materials and Methods for the principles of the deconvolution-based procedure of quartz and moganite-related bands). In particular, the thin bluish zone 1 at the boundary with encasing saddle dolomite shows elevated moganite content (39–45 wt.%), but these values decrease to 24–33 wt.% once the colour shifts from bluish towards bluish-white in the zone 2. The innermost (bluish) regions of agate bodies, referred to as zone 3, show an increase in moganite admixture that covers the range 37–46 wt.%.
Figure 10. (a) The distribution of moganite-rich (bluish, Zones 1 and 3) and moganite-poor (bluish-white, Zone 2) areas with a single chalcedonic (agate) body surrounded by a dolomite and opal-rich matrix; (b, c, d) average (based on five individual analyses collected from each of the three zones) and deconvoluted spectra and the calculated mean moganite concentration (wt.%).
Raman spectroscopy also confirmed the presence of enargite hosted by the opaline matrix of listvenite (Fig. S3). Here, the strong band at 339 cm–1, accompanied by quite weak signals centred at ∼384 and 278 cm–1, was recognized. According to Berkh et al. (Reference Berkh, Majzlan, Meima, Plášil and Rammlmair2023), these bands correspond to As–S vibrations, i.e. antisymmetric stretching, symmetric stretching and bending modes, respectively.
Fourier-transform infrared spectroscopy (FTIR)
The FTIR spectra of green-coloured opaline silica (lussatite) show three prominent main bands centred at 1100, 789 and 474 cm–1 (Fig. 11a). These bands can be attributed to fundamental vibrations of the SiO4 tetrahedral network, i.e. Si–O–Si antisymmetric stretching, Si–O–Si symmetric stretching and O–Si–O bending modes, respectively (Etchepare et al., Reference Etchepare, Merian and Smetankine1974; Fröhlich, Reference Fröhlich2020). Weak signals at 1880 and 1631 cm–1 are due to overtones of Si–O vibrations and bending of molecular water, respectively. All of these features are consistent with opal-CT characteristics reported elsewhere in the literature (Rice et al., Reference Rice, Freund, Huang, Clouse and Isaacs1995; Adamo et al., Reference Adamo, Ghisoli and Caucia2010; Fröhlich, Reference Fröhlich2020; Ejigu et al., Reference Ejigu, Ketemu, Endalew and Assen2022). Additional bands at 1427, 881 and 728 cm–1 originate from stretching and bending modes of the CO32– group fixed with dolomite (Hsiao et al., Reference Hsiao, Wang, La Plante, Pignatelli, Krishnan, Le Pape, Neithalath, Bauchy and Sant2019). A weak signal at 671 cm–1 can be assigned to the presence of Si–O deformation bands in smectite-group species (Bukka et al., Reference Bukka, Miller and Shabtai1992), whereas the band at 694 cm–1 marks the presence of trace amounts of α-quartz (impurity or intergrowths) (Fig. 11b). In addition, the water-related region of the spectrum (Fig. 11c) shows a broad band with a maximum at ∼3450 cm–1 and quite a sharp peak at 3650 cm–1. These signals are linked to the presence of molecular water (H2O) and silanol-group water (Si–OH), respectively (Graetsch et al., Reference Graetsch, Flörke and Miehe1985; Flörke et al., Reference Flörke, Graetsch, Martin, Röller and Wirth1991), though the spectrum has probably been modified by the vibrations of water molecules hosted by smectite-group species. Two additional but minor bands at 2926 cm–1 and 2856 cm–1 visible in Fig. 11c reflect vibrations of the methylene group C–H (Ejigu et al., Reference Ejigu, Ketemu, Endalew and Assen2022).
Figure 11. (a) FTIR spectrum collected from green-coloured opaline silica (lussatite) in the range 2000–400 cm–1. The second derivative curve was obtained to unravel eventual discrete quartz-related bands. Note that the shape of this curve is quite similar to that obtained from lussatite termed ‘biot opal-CT’ in the work of Fröhlich (Reference Fröhlich2020). (b) The same spectrum in the range 760–640 cm–1; (c) water-related region (4000–2800 cm–1) of the FTIR cm–1 spectrum where Si–OH and H2O-related signals at 3650 and 3450 cm–1 are presented.
Ultraviolet-Visible-Near Infrared (UV-VIS-NIR) spectroscopy
The UV-VIS-NIR spectrum for the analysed green opaline silica (lussatite) has two weak absorption bands at ∼450 nm and ∼490 nm, followed by a broad and asymmetric signal centred at ∼600 nm (Fig. 12). Additionally, there is a small sharp peak centred at 681 nm. These features are consistent with the presence of chromium (Cr3+) in octahedral coordination and were reported for chrome-bearing chalcedony deposits referred to as aquaprase/mtrolite (Willing and Stocklmayer, Reference Willing and Stocklmayer2003; Feral, Reference Feral2022; Lv and Guo, Reference Lv and Guo2023). Such a spectral characteristic is remarkably different compared to chrysoprase (or prase opal), where Ni-bearing hydrous silicates (kerolite–pimelite and/or willemseite) act as colouring agents (Fig. 12). In particular, the spectrum of chrysoprase (added to Fig. 12) exhibits a broad and prominent band at ∼650 nm, followed by a shoulder band at ∼740 nm. According to Lv and Guo (Reference Lv and Guo2023) and Jiang and Guo (Reference Jiang and Guo2021), this type of spectral characteristic can be assigned to octahedrally-coordinated Ni2+ and/or Fe2+-Fe3+ charge transfer. Figure 12 also illustrates the UV-VIS-NIR spectrum obtained from bluish chalcedony (agate). The spectrum is relatively flat with absorption at 450 and 482 nm, followed by a broad and diffuse band at ∼600 nm, that was reported from moganite-rich bluish chalcedony according to Kiefert et al. (Reference Kiefert, Krzemnicki, Fiedler, Sintayehu and Furuya2023).
Figure 12. UV-VIS-NIR spectra for green-coloured opaline silica (lussatite) and bluish chalcedony (agate) from the Tokat area. The reference spectrum for chrysoprase from the Szklary area (Lower Silesia, Poland) is shown to emphasise the difference between Cr- and Ni-related colour enhancements.
Whole-rock composition
The juxtaposition of major-element (XRF-based) and trace-element (ICP-MS-based) compositions obtained from the opal-rich listvenite (green) and vein agate (bluish to white) is shown in Table 2. The former contains SiO2 (53.1 wt.%), with CaO (12.3 wt.%) and MgO (6.7 wt.%) due to the presence of dolomite impurities. Fe2O3 and Al2O3 reach 5.7 and 1.1 wt.%, respectively, reflecting the presence of dolomite, pyrite and smectite impurities (see the PXRD patterns), whereas the high LOI value (19.90 wt.%) arises from the presence of water-hosted components (opal, smectite) and [CO3]2– groups fixed with dolomite. The opal-rich listvenite (green) is enriched in transition metals with chromophore affinities such as: Cr (1615 ppm); Ni (307 ppm); Cu (155 ppm); Mn (1550 ppm); Co (108 ppm); and V (24 ppm), whereas high-field strength elements (U, Th and Hf) are at low and/or below detection limits. The composition of vein agate (blue to white) is dominated by SiO2 (97.1 wt.%) with the minor FeO (0.7 wt.%) and CaO (0.3 wt.%). Here, the amounts of trace elements are quite low compared with the opal-rich green-coloured matrix, except for significant amounts of As (180 ppm) and Mo (7.3 ppm) (Table 2).
Table 2. Main (XRF-based) and trace (ICP-MS-based) whole-rock element composition of green-coloured opal-rich matrix and vein agate
* LOI – loss on ignition
Discussion
Colouring agents and the structural state of opaline and chalcedonic silica
The combined microscopic, PXRD and spectroscopic data argue for the presence of opal-CT (lussatite variety) as the dominant component of the green-coloured opaline matrix of listvenite. The term lussatite was originally coined by Mallard (Reference Mallard1890) and refers to the length-slow fibrous opal-CT with hydrothermal origin (Graetsch et al., Reference Graetsch, Gies and Topalović1994), but can also serve as a discrete mixture of chalcedony and opal-rich domains (Rogers, Reference Rogers1938). The reported opal-CT (lussatite) shows a more ‘tridymitic’ nature as evidenced by the position of the FTIR-related band at 789 cm–1 close to the tridymite reference at 791 cm–1, as well as the position of the XRD peak at 4.11 Å and the shape of the Raman spectra in the low-wavenumber range.
The green colouration of opaline silica from the Tokat region has been enhanced by the presence of chromium, as evidenced by the shape of the UV-VIS-NIR diagnostic absorption band at 681 nm (Feral, Reference Feral2022). The presence of Cr at a concentration level of 1615 ppm, was further confirmed using the ICP-MS technique. Note that, Ni and Fe at 307 and 4430 ppm, respectively, could also act as chromophores and/or modify the saturation/tone of the observed green hue. The powder XRD data combined with elemental WDS mapping (Figs 6 and 7, respectively) suggest strongly that Cr and Fe are sequestered in phyllosilicates such as volkonskoite and/or Cr-rich montmorillonite (both green in the macroscale) which act as colouring agents for opaline silica. The reported colouration mechanism is analogous to that reported from chrysoprase, where a green hue was considered to be due to the presence of nickel ions hosted by clay minerals (willemseite and/or kerolite/pimelite) (Sojka et al., Reference Sojka, Witkowski, Zabiński, Dyrek and Bidzińska2004; Eggleton et al., Reference Eggleton, Fitz Gerald and Foster2011; Čermáková et al., Reference Čermáková, Hradil, Bezdička and Hradilová2017). In addition, Cr was probably derived from the breakdown of pristine spinel-group minerals (e.g. chromite), the relics of which were observed in the opaline matrix of listvenite (Fig. 4b).
In contrast, the bluish-white colouration of chalcedonic (agate) veins was facilitated by physical effects such as Rayleigh and/or Mie scattering (Götze et al., Reference Götze, Möckel and Pan2020; Welman-Purchase et al., Reference Welman-Purchase, Wicht, Miller and Roelofse2024), rather than by mineral or elemental impurities. These effects reflect the scattering of light by tiny chalcedony (quartz+moganite) particles of a specific size, but can also be linked to the presence of water species. There is also a positive correlation between the admixture of moganite and the intensity of a blue tint, as presented in Fig. 12 (cf. Kiefert et al., Reference Kiefert, Krzemnicki, Fiedler, Sintayehu and Furuya2023). The shift from bluish towards bluish-white colouration could have been caused by the presence of recrystallization-related microtextures followed by possible loss of water (i.e. feathery and mosaic quartz) (Powolny et al., Reference Powolny, Dumańska-Słowik, Szczerbowska-Boruchowska and Woszczyna2024).
Origin and crystallization sequence
The observed mineral assemblage of listvenite (opal-CTLUS + saddle dolomite + Fe-As-Co-sulfides + chalcedony) reflects the pervasive influx of medium- to low-temperature hydrothermal fluids with fluctuating/evolving compositions and/or pH, as well as a possible boiling-related environment. The latter is supported by the local bladed morphology of dolomite and colloform banding of chalcedonic layers (Moncada et al., Reference Moncada, Mutchler, Nieto, Reynolds, Rimstidt and Bodnar2012). The detailed and assumed crystallization sequence is presented in Fig. 13, whereas more detailed petrogenetic implications are discussed below.
Figure 13. Assumed crystallization sequence of the listevenite accompanied by possible temperature conditions.
The presence of chromite relics together with secondary Cr-bearing phases (volkonskoite, Cr-montmorillonite), as well as the abundance of transition metals (e.g. Cr, Fe, V, Co, Ni; see Table 2), suggest the presence of an ultramafic precursor (e.g. peridotite, gabbro) that has been later strongly affected by the influx of CO2-bearing fluids, followed by combined in situ two-stage silicification (opalization + formation of chalcedonic bodies) and dolomitization (Foord et al., Reference Foord, Starkey, Taggart and Shawe1987; Boschi et al., Reference Boschi, Dini, Dallai, Ruggieri and Gianelli2009). The reported association of silica+dolomite+pyrite reflects the very final stage of weathering and/or coupled silica-carbonate metasomatism (i.e. listvenitization) of serpentinized ultramafic series (Boskabadi et al., Reference Boskabadi, Pitcairn, Leybourne, Teagle, Cooper, Hadizadeh, Nasiri Bezenjani and Monazzami Bagherzadeh2020; Menzel et al., Reference Menzel, Urai, Ukar, Hirth, Schwedt, Kovács, Kibkalo and Kelemen2022a, Reference Menzel, Urai, Ukar, Decrausaz and Godard2022b, Reference Menzel, Sieber and Godard2024). This process results in the leaching of the main serpentine-hosted elements such as Mg, Si and Al, whilst Cr, Ni and Fe tend to be immobile during ongoing water–rock interactions (Venturelli et al., Reference Venturelli, Contini, Bonazzi and Mangia1997; Boschi et al., Reference Boschi, Dini, Dallai, Ruggieri and Gianelli2009; Lacinska and Styles, Reference Lacinska and Styles2013).
Listvenites form below ∼200°C and may eventually transform into so-called ‘birbirite rocks’, where carbonate minerals might have dissolved due to ongoing hydrothermal activity (Menzel et al., Reference Menzel, Sieber and Godard2024). Thus, the stockwork-like macro-texture of the samples (i.e. dolomitic veins intersecting opaline matrix; Fig. 2d) could have been inherited from both serpentinites and listvenites, where carbonate-rich veins (commonly magnesite) are introduced together with the influx of CO2-rich fluids and the dissolution of serpentinite minerals. As the listvenite analysed does not possess any other unequivocal signs of relic phases (e.g. olivine, serpentinite), and corresponding microtextures (except for chromite remnants), the exact nature of the original rock and silicification mechanisms remains to a certain extent elusive. The proposed petrogenetic hypothesis, however, appears to be valid as ultramafic rocks and their alteration products, belonging to Pre-Liassic and Cretaceous ophiolite complexes (Yilmaz and Yilmaz, Reference Yilmaz and Yilmaz2004; Catlos et al., Reference Catlos, Huber and Shin2013), are present ubiquitously in the study area. Here, the circulation of hydrothermal Si-rich fluids could be further maintained by the presence of a fault system (e.g. strike-slip faults) and erosional unconformities, although their origin (meteoric, deep-seated hydrothermal?) is vague and requires further analysis (e.g. O and C isotope data). The presence of contrasting (high to medium and low temperature – Fig. 13) mineral assemblages suggests a mixing environment, i.e. the migration of high-temperature metal- and/or CO2-rich fluids and cooler Ca-rich meteoric waters containing Al and Si. In particular, the sulfide mineralization observed (cobaltite, pyrite, enargite) reflects an early stage of crystallization (possibly exceeding 200°C), followed by the circulation of S-, Fe-, Cu- and As-bearing fluids in the high-sulfidation environment (Dekov et al., Reference Dekov, Rouxel, Kouzmanov, Bindi, Asael, Fouquet, Etoubleau, Burgaud and Wälle2016; Aluç et al., Reference Aluç, Kuşcu, Peytcheva, Cihan and von Quadt2020). Subsequently dolomite veins were introduced, whilst silicification (opalization) and co-crystallization of smectite-group species (Cr-montmorillonite, volkonskoite) postdated or occurred contemporaneously with fracture-filling dolomitization (Menzel et al., Reference Menzel, Urai, Ukar, Hirth, Schwedt, Kovács, Kibkalo and Kelemen2022a). Syn-kinematic carbonation (i.e. prior to the final silicification/listvenitization) is demonstrated by the deformation of lussatite-rich areas close to adjacent dolomite crystals (Fig. 3c). Dolomite veins could also be partially introduced during late-stage brecciation of already opalized serpentinite. Finally, the two-stage fault-controlled listvenitization, related to either post-granite hydrothermal fluids or meteoric waters, has been discussed widely by Sarifakioglu (Reference Sarifakioglu2023) in the study of altered ophiolitic series disseminated in the area of Turkey. The coexistence of sulfides (pyrite, cobaltite and enargite) and post-listvenite agate mineralization may thus reflect the presence of both processes during the formation of the rocks analysed.
Two questions remain regarding the proposed origin (i.e. silicification and carbonation of serpentinite) and mineral association (saddle dolomite + opal).
(1) Why was the dolomite formed rather than magnesite, as the latter is far more common in listvenitic rocks (Amarbayar et al., Reference Amarbayar, Dandar, Wang, Okamoto, Uno, Batsaikhan, Takayanagi, Iryu and Tsuchiya2023)? This stems from the fact that the magnesite (MgCO3) can readily consume excess Mg released via the dissolution of Mg-bearing silicates of the serpentinite subgroup (chrysotile, antigorite, lizardite), according to the reaction: Mg3Si2O5(OH)4 + 3CO2 = 3MgCO3 + 2SiO2 + 2H2O (Amarbayar et al., Reference Amarbayar, Dandar, Wang, Okamoto, Uno, Batsaikhan, Takayanagi, Iryu and Tsuchiya2023; Menzel et al., Reference Menzel, Sieber and Godard2024).
(2) Why did the opal (lussatite) crystallize instead of quartz as the latter is more common in the listvenite rocks?
First, the formation of Ca-bearing carbonates (dolomite or calcite) in listvenite-type rocks can be linked to the external influx of CaCl2- and CO2-rich waters (Menzel et al., Reference Menzel, Sieber and Godard2024) that might have been mixed with serpentinite-related Mg-bearing (residual) fluids, followed by the subsequent invasion of CO2 (see previous paragraph). Such mixing would reduce Mg/Ca ratios and promote the formation of dolomite, although the in situ dissolution of eventual Ca-rich phases (e.g. pyroxene) is still possible to some extent. In addition, the presence of Jurassic–Cretaceous carbonate rocks in the vicinity of the study area (Catlos et al., Reference Catlos, Huber and Shin2013) could also act as a reservoir for Ca-rich fluids. The alternative scenario involves the replacement of possible pristine magnesite by dolomite (Menzel et al., Reference Menzel, Urai, Ukar, Decrausaz and Godard2022b), although this hypothesis is not supported by petrographic observations such as pseudomorphic-related textures and remnants of magnesite. Moreover, as the dolomite from the listvenite analysed is chiefly represented by the so-called saddle variety, the influx of hydrothermal fluids between ∼60 and 180 °C (Spötl and Pitman, Reference Spötl and Pitman1998; Davies and Smith, Reference Davies and Smith2006; Zhang et al., Reference Zhang, Hu, Qian, Wang, Cao, Zhu, Li and Xie2009) was possible during the listvenitization processes. The distortion of its crystal lattice and peculiar microscopic features (i.e. saddle-shaped crystals, sweeping light extinction) originate from the edge-nucleated crystal growth and significant growth rates, according to Searl (Reference Searl1989). The Na and Sr contents, which are below the detection limit and not listed in Table 1, together with the absence of fluid inclusions suggest that dolomite-forming fluids were of low-salinity and non-evaporitic origin, with the Mg derived from the host serpentinite.
The stoichiometric Ca (∼1.0 apfu; Table 1) and high Mn contents of the dolomite analysed, coupled with the large crystal size and common euhedral habit, could point further to a prolonged crystallization from dilute meteoric (or mixed) waters under reducing crystallization conditions. The ferroan nature of dolomite could be enhanced by sulfide (e.g. pyrite) oxidation and subsequent Fe release, with or without the contribution of microbial activity and/or /oxidation of organic materials.
Secondly, the formation of opaline silica (lussatite) rather than quartz suggests the presence of low-temperature, highly silica-supersaturated and polymerized fluids (Fournier, Reference Fournier, Berger and Bethke1985; Boschi et al., Reference Boschi, Dini, Dallai, Ruggieri and Gianelli2009). Opalization could also be facilitated by the presence of positively charged hydroxyl complexes (Williams and Crerar, Reference Williams and Crerar1985). The origin of lussatite from the Tokat region could be related to either direct precipitation from hydrothermal fluids (Nagase and Akizuki, Reference Nagase and Akizuki1997; Schindler et al., Reference Schindler, Fayek, Courchesne, Kyser and Hawthorne2017; Mustoe, Reference Mustoe2023), or recrystallization from the pristine amorphous opal-A (Ulrich et al., Reference Ulrich, Muñoz, Guillot, Cathelineau, Picard, Quesnel, Boulvais and Couteau2014; Liesegang and Tomaschek, Reference Liesegang and Tomaschek2020). The former scenario is more plausible as there is no petrographic evidence for opal-A to -CT recrystallization, whereas tridymite stacking seems to prevail over cristobalite (note that the latter should be introduced during the textural maturation of opaline silica; Rice et al., Reference Rice, Freund, Huang, Clouse and Isaacs1995). We speculate that the presence of clay-group species has further inhibited the opal-CT→ chalcedony (quartz) transition.
In turn, the formation of chalcedonic (agate) bodies represents the final and low-temperature (post-listvenite) stage of hydrothermal alteration, accompanied by the ongoing reduction in pH and CO2 activity maintained by the former pervasive crystallization of dolomite. On the basis of microscopic and CL data, there was an inward (syntaxial) growth of chalcedonic bands towards the centre of the vug. Here, chalcedony can be regarded as the product of direct precipitation from fluid, though the observed colloform banding and feathery quartz (see Fig. 5d) might also indicate the presence of an amorphous gel-like precursor (Yilmaz et al., Reference Yilmaz, Duschl and Di Genova2016). Moreover, the observed presence of opaline silica between chalcedonic bands (visible in Fig. 5c,d) is not only uncommon in agate deposits worldwide, but can be quite supportive of the rapid changes in silica supersaturation in an agate-forming medium.
The formation of chalcedony (agate) and another generation of opal (i.e. veins crosscutting lussatite; Fig. 3a,b) is probably related to the brecciation of already silicified (opalized) and/or carbonatized serpentinite (listvenite). However, a contemporaneous dissolution of the earlier dolomitic veins by silica should be endorsed based by such petrographic observations as corrosion-related features (Fig. 4a), ghost cathodoluminescence zonation (Fig. 5e,f) and impregnation of dolomite with silica (Fig. 5g–h). Prismatic quartz is commonly observed not only in the centre of chalcedony (agate) bodies, but also in the external regions of the veins (i.e. sandwiched between chalcedonic layers; Fig. 5c,d). This type of sequence is not typical of agate deposits worldwide, where prismatic euhedral quartz tends to occupy the centre of the amygdales/veins due to a progressive decrease in the concentration of dissolved silica (French et al., Reference French, Worden and Lee2013; Dumańska-Słowik et al., Reference Dumańska-Słowik, Powolny, Sikorska-Jaworowska, Gaweł, Kogut and Poloński2018; Powolny et al., Reference Powolny, Dumańska-Słowik, Sikorska-Jaworowska and Wójcik-Bania2019b, Reference Powolny, Dumańska-Słowik, Szczerbowska-Boruchowska and Woszczyna2024). Thus, the formation of chalcedonic (agate) veins from the Tokat area originated from remarkable cyclic fluctuations (degree of polymerization and supersaturation, temperature, etc.) of the Si-rich medium. This conclusion is also supported by the local abundance of thin blue-luminescent opal-rich bands between chalcedony-quartz layers, as well as alternating (moganite-rich and moganite-poor) domains (Fig. 10). Furthermore, the presence of moganite (up to 46 wt.%) can be ascribed to the circulation of alkaline fluids (Heaney and Post, Reference Heaney and Post1992; Kayama et al., Reference Kayama, Tomioka, Ohtani, Seto, Nagaoka, Götze, Miyake, Ozawa, Sekine, Miyahara, Tomeoka, Matsumoto, Shoda, Hirao and Kobayashi2018), but can also be connected with the young age of silica mineralization and lack of pronounced age-related textural maturity (Moxon and Ríos, Reference Moxon and Ríos2004; Moxon, Reference Moxon2017). Otherwise, the scarcity of LS chalcedony (quartzine) and the preservation of LF chalcedony might suggest an acidic (Mg- and sulfate-poor) crystallization environment (Folk and Pittman, Reference Folk and Pittman1971), but the exact controls on the formation of these chalcedony types remains ambiguous (Keene, Reference Keene1983; Powolny et al., Reference Powolny, Dumańska-Słowik, Szczerbowska-Boruchowska and Woszczyna2024). Finally, the blue CL of prismatic megaquartz is typical of a low-temperature hydrothermal environment during the formation of vein agate (Götze et al., Reference Götze, Möckel, Kempe, Kapitonov and Vennemann2009).
Conclusions
(1) There were at least two stages in the formation of opal-bearing listvenite and vein agate from the Tokat area, both related to the circulation of medium- to low-temperature hydrothermal and/or meteoric waters within parental rocks which were possibly serpentinized ultramafics. First, the early post-serpentinite stage (i.e. listvenitization) involved the formation of sulfide mineralization (pyrite, enargite and cobaltite), as well as the crystallization of fracture-filling saddle dolomite and concomitant or later opal–Cr-smectite association. Secondly, the already carbonatized and/or opalized rock (listvenite) was affected by cyclic influxes of Si-rich fluids that precipitated chalcedony (agate) in newly formed vugs, although replacement of the former saddle dolomite by silica should be envisaged. The observed associations suggest the circulation of external CO2-rich and Ca-rich fluids within Mg-rich parental rock (serpentinite), followed by the later pervasive influx of Si-rich agate-forming solutions.
(2) The green-coloured opaline matrix of listvenite has been recognised as the opal-CT (lussatite subvariety, i.e. with chalcedony-like microfabrics, with a positive sign of elongation) that bears a rather ‘tridymitic’ than ‘cristobalitic’ character based on the observed spectroscopic and X-ray diffraction characteristics. The formation of the listvenite was probably related to direct precipitation from supersaturated fluids in a Mg-Ca-Al-rich environment. The cross-cutting chalcedonic (agate) veins consist of α-quartz with a variable moganite admixture (up to 46 wt.%). The latter is responsible for the intensity of a bluish hue within chalcedonic bands but could also maintain recrystallization of silica into mosaic and/or feathery quartz.
(3) The green colour of listvenite-hosted opal-CT (lussatite) is triggered by dispersed Cr-bearing smectite (volkonskoite or Cr-montmorillonite) whilst the chromophore elements (mainly Cr, but also Ni, Fe, Co) were derived from a spinel-bearing ultramafic precursor (serpentinized and now listvenitized). Additionally, the reported colouration mechanism is akin to that reported from chrysoprase, where Ni-bearing clays act as colouring agents. Moreover, Cr-related green colouration produces an exclusive diagnostic UV-VIS-NIR absorption band at ∼691 nm, the presence of which is absent from chrysoprase.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1180/mgm.2025.10174.
Acknowledgements
This work was supported financially by AGH Statutory Grant No.16.16.140.315. The authors appreciate the assistance of Adam Włodek during the electron microprobe measurements. Peter Heaney and an anonymous reviewer are acknowledged for their comments on the initial version of the text. The authors also thank Principal Editor, Roger Mitchell, for his assistance in handling the manuscript.
Conflict of interest
The Authors declare no conflicts of interest regarding the present work.
Open access
