Introduction

Kaolin group minerals (kaolinite, dickite, nacrite) having 1:1 phyllosilicate structure with a composition of Al2Si2O5(OH)4 are common hydrothermal alteration minerals under acidic conditions, that formed by alteration of primary igneous minerals as well as direct precipitation (neoformation) from the hydrothermal fluids (e.g., Murray 1988; Murray and Keller 1993; Inoue 1995). Hydrothermal alteration of volcanic and volcanoclastic rocks with acidic and intermediate composition commonly leads to the formation of industrial clays (e.g., Murray 1988). For instance, there are several industrial kaolin deposits in volcanic-hosted rocks related to acidic conditions in several regions of Turkey (Karakaya et al. 2001; Sayın 2007; Karakaya 2009; Ece et al. 2013; Kadir and Erkoyun 2013; Kadir et al. 2014; Ünal-Ercan et al. 2016; Başıbüyük and Yalçın 2019). The study area is located in the Cappadocia Volcanic Province (CAVP) that contains Miocene-aged volcanogenic units (Innocenti et al. 1975; Temel 1992) (Fig. 1).

Fig. 1
figure 1

Geological map and stratigraphic section of the Avanos-Nevşehir area (modified from Aydar 2012) and location of the studied samples

Full size image

Previous studies on CAVP units generally focused on the tectonic, petrological, geochemical, and volcanological aspects (e.g., Innocenti et al. 1975; Schumacher et al. 1990; Le Pennec et al. 1994; Mues-Schumacher and Schumacher 1996; Schumacher and Mues-Schumacher 1996; Temel et al. 1998; Le Pennec et al. 2005; Viereck-Götte et al. 2010; Schmitt et al. 2011; Aydar et al. 2012) while only a few studies were made on clay occurrences in CAVP (Ertek and Öner 2008; Gurel and Kadir 2008; Kadir et al. 2013). Ertek and Öner (2008) mentioned the possible hydrothermal alteration effect on the ignimbrites and stated an intense alteration caused the kaolin clay occurrences and their industrial uses effect, but did not provide any detail for hydrothermal alteration. The origin and conditions of volcanic-hosted clay-bearing alteration can provide important data and/or tools for determining potential industrial clay occurrences for widely distributed volcaniclastic layers in CAVP. This study is focused on the hydrothermal alteration of the lowermost parts of ignimbrite beds. The Kavak ignimbrite (Pasquaré 1968) represents the earliest volcanogenic stages (9.12 Ma, Aydar et al. 2012), and its ignimbrite layers have a potential alteration state through the later volcanic activities that extended up to 2.5 Ma (Aydar et al. 2012). In this context, this study aimed to investigate detailed mineralogical and geochemical characteristics of altered ignimbrites and to understand the origin and occurrence of clay-rich alteration zones through hydrothermal alteration.

Geological setting and lithology

The study area in the central part of the CAVP (Fig. 1) has been investigated throughout the last few decades in terms of its volcanology context (Aydar et al. 2012 and references therein). The CAVP as a plateau with an altitude of 1400–1500 m is dated as Miocene to Upper Holocene in age. The CAVP has been reported to have a few Quaternary stratovolcanoes (Erciyes Mountain and Hasan Mountain) and numerous monogenetic outlets (cinder cones, maars, domes) (Aydar and Gourgaud 1998).

Miocene volcanic complexes and lava flows are also associated with ignimbrites (e.g., Aydar and Gourgaud 1998; Aydar et al. 2012). This unit is underlain by sedimentary, magmatic, and metamorphic basement rocks (Fig. 1). The region has a topographical slope in the north direction, bounded by the Taurus Mountains in the south, and two distinct Quaternary stratovolcanoes; Hasan Mountain and Erciyes Mountain, respectively in the west and east. Miocene-Pliocene ignimbrites cover a total area of 20,000 km2 (Le Pennec et al. 1994; Aydar et al. 2012).

Large pyroclastic discharge deposits were formed by covering the surface with high-intensity pyroclastic currents caused by the explosion of magmas ranging from volcanic rocks with composition of dacites to rhyolites (Smith 1960; Wilson et al. 1980; Aydar et al. 2012). The pyroclastic layers are distributed as very common and distinct deposits and these discharge deposits may have individually traveled long distances (e.g., Wilson et al. 1995; Aydar et al. 2012). The nomenclature of ignimbrite was taken from Le Pennec et al. (1994), in which 10 ignimbrite members of Ürgüp Formation were identified from old to young: Kavak, Zelve, Sarımadentepe, Sofular, Cemilköy, Tahar, Gördeles, Kızılkaya, Valibabatepe, and Kumtepe. They are separated by abundant rhyolitic-dacitic ignimbrites, lacustrine deposits, earth, or lava flows (Pasquaré 1968). This study covers the Upper Miocene Kavak ignimbrite member, which is dated as the oldest ignimbrite unit in the stratigraphic sequence (Aydar et al. 2012). However, the Kavak ignimbrite was previously defined as the upper and lower Göreme ignimbrite (Schumacher et al. 1990; Ercan et al. 1994), but the Kavak ignimbrite term is preferred in recent studies (e.g., Aydar et al. 2012 and references therein). Considering the stratigraphic position of the Kavak ignimbrite and K/Ar dating data, Upper Miocene age is preferred. In the K/Ar analysis of biotite minerals of the Kavak ignimbrite, the following ages were obtained by previous researchers: 8.6 ± 1.7 Ma (Innocenti et al. 1975); 11.2 ± 2.5 Ma (Temel 1992), and 8.96 ± 0.2 Ma (Schumacher et al. 1990). The mean age was determined by these researchers and in recent studies an age of 9.12 ± 0.09 Ma was reported (Aydar et al. 2012).

Field observations show that ignimbrites have lapilli structures and parts resembling karstic gaps are observed in some parts of the white-gray altered unit (Fig. 2). The upper parts of the Kavak ignimbrite unit exhibits fine-grained stratified facies at lower parts and massive ignimbrite facies at uppermost parts (e.g., Guzmán et al. 2020) (Fig. 2a–c). The erosional differences related to silica-rich (slightly-eroded) in massive ignimbrite facies and clay-rich (highly-eroded) levels of fine-grained stratified parts show a similar appearance to fairy chimneys (Fig. 2a, b). Highly altered clay-rich (argillized) ignimbrite layers mostly exhibit clean white colors (Fig. 2b, c). Iron oxy-hydroxide (hematite, goethite/limonite) zones within the altered ignimbrite layers display orange-yellow colors due to the effect of hydrothermal fluids circulation especially throughout the fault planes and fractures (Fig. 2d, e).

Fig. 2
figure 2

Field observations of the Kavak ignimbrite unit. a Erosional differences for altered ignimbrite layers related to silica-rich (slightly-eroded) and clay-rich (highly-eroded) levels known as fairy chimneys. b, c Highly altered (argillized) ignimbrite layers. d, e Iron oxy-hydroxide (hematite, goethite/limonite) zones along the cracks and/or fault zones within the altered ignimbrite layers

Full size image

Materials and methods

A total of 30 samples were collected from different levels and locations between Nevşehir and Avanos (Fig. 1) and analyzed by several methods: optical microscopy (OM), scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM-EDX), X-ray fluorescence spectroscopy (XRF), and inductively coupled plasma mass spectrometry (ICP-MS) methods. The X-ray diffraction (XRD), XRF, SEM-EDS, and ICP-Ms analyses were conducted at the Central Research Laboratory of Niğde Ömer Halisdemir University, Turkey. XRD analyses of some samples were performed at the Geology Department of the University of Cincinnati, USA. The oxygen and hydrogen isotope analyses were performed at the Stable Isotope Laboratory of Cornell University, USA.

OM studies were conducted by Olympus binocular transmitted light microscopy to determine textural and mineralogical features of altered ignimbrites.

XRD analysis was performed using a Panalytical EMPYREAN Advance Diffractometer for determining the whole rock and clay fraction mineral composition. The analysis were made by using CuKα (λ = 1.541871 Å) radiation. Diffraction patterns were recorded from 5 to 70° 2θ with a scan rate of 3°min−1 for whole-rock, 5 to 30° 2 θ with a scan rate of 2°min−1 for clay fraction. Samples for clay analysis (< 2 μm) were prepared by separation of the clay fraction by sedimentation, followed by centrifugation of the suspension, after overnight dispersion in distilled water. The clay particles were dispersed by ultrasonic vibration for about 15 min. Three oriented clay specimens were prepared from each sample by the following protocol: air dried, ethylene-glycol-solvated at 60 °C for 2 h, and thermally treated at 550 °C for 2 h, respectively. Rietveld optimization for XRD data was performed with PanAlytical software X’Pert Highscore Plus.

SEM–EDX analysis was performed in a Zeiss EVO 40XVP instrument. Representative clay-rich samples were prepared for SEM–EDX analysis by gluing the fresh, broken surface of the sample onto an aluminum sample holder that had been covered with double-sided tape and coated with a thin film (~ 350 Å) of gold, using a Giko ion coater.

pH values of the hydrothermally altered samples were determined from homogenized powder samples within the distilled water (suspension) by using a pH meter (Isolab/pH-Orp). Fourier-transform infrared spectroscopy (FTIR) analyses were achieved from Bruker Vertex 70 Brand instruments.

The major oxide chemical analyses were obtained for twenty-one representative pure or nearly pure kaolin samples by XRF (Philips Panalytical Zetium). Chemical analyses were performed using the rock standards supplied by MBH Reference Materials and Breitlander companies. The accuracy for elements was ± 2%. Loss on ignition (LOI) of each sample was also determined by drying the samples at 105 °C overnight, followed by calculation of their H2O content and other volatiles at 1050 °C. Trace element analyses were acquired by inductively couple plasma-mass spectroscopy (ICP-MS). The detection limits ranged from 0.1 to 5 ppm.

The stable isotope analyses were conducted in a Thermo Delta V isotope ratio mass spectrometer (IRMS) interfaced to a temperature conversion elemental analyzer (TC/EA). The results were given as δ18O and δ2H or δD standardized with Vienna Standard Mean Ocean Water (V-SMOW). Delta values are measured in units of per mil (‰). Precision is ability of the instrument to consistently measure the same value for a known sample throughout the course of the analytic run. The standard deviation for the internal standard Inter Benzoic Acid for δ2H is 4‰, and for δ18O is 0.09‰. Corrections were performed using two-point normalization (regression) based on standards USGS 601 and USGS 602 for δ18O and IAEA CH7 and EMA-P1 for δ2H. Detailed analytical procedures are given in the laboratory website (www.cobsil.com).

Results and discussion

Petrography results

Optical microscopy

The glassy ash tuffs of the Kavak ignimbrite unit has a porphyritic texture containing phenocrysts of plagioclase, quartz, hornblende, biotite, opaque mineral (iron oxide), and alkali feldspar within the devitrified volcanic groundmass associated with pumice grains. According to chemical compositions, ignimbrites have mostly dacite and trachyandesite, and partly rhyolite composition (Temel et al. 1998; Ertek and Öner 2008). Plagioclase phenocrysts having polysynthetic twinning show fractured and/or cracking texture indicating the pyroclastic eruption (Fig. 3a). Pyroclastic eruption also deformed the biotite phenocrysts as shown by the development of fringes along the cleavage planes (Fig. 3b). The quartz minerals exhibit primary magmatic texture which was corroded by the volcanic groundmass (Fig. 3c).

Fig. 3
figure 3

Optical microscope images of ignimbrites. a, b Fractured plagioclase phenocryst within the devitrified (altered) volcanic groundmass and pumice (a: crossed nicols, b: open nicol). c Deformed biotite phenocryst with fringe texture along the cleavage planes within the siliceous volcanic groundmass (Crossed nicols). d Corroded quartz phenocrysts within the siliceous volcanic groundmass (crossed nicols)

Full size image

The optical microscopic observations of intensely altered ignimbritic rocks show silicification and argillization (Fig. 4). Primary magmatic quartz (Qm) commonly corroded the glassy volcanic groundmass which was extremely devitrified to silica and clays (Fig. 4a–d). The pumice grains having porous texture were also altered to silica and clay minerals (Fig. 4a, b). Biotite phenocrysts were extremely altered to opaque (iron oxide) minerals (opacitized) and some grains were completely replaced by opaque minerals and exhibit a pseudomorphic texture (Fig. 4c, d). Glass shards are mainly observed as completely silicified grains (Qzh), whereas clay occurrences are completely related to the volcanic glassy matrix (Fig. 4c–f). The petrographic observations indicate that the common alteration products are cryptocrystalline silica (quartz) and clay (kaolinite/dickite) minerals. Hydrothermal alteration related silica minerals were precipitated as both pore filling (euhedral spherical relicts) and replacement of glass shards (sharp edged grain), where clays were developed due to transformation from the glassy volcanic matrix.

Fig. 4
figure 4

Optical microscope images of altered ignimbrites. a, b Quartz phenocrysts (primary magmatic quartz-Qm) corroded by silicified (hydrothermal quartz-Qh) and argillized (kaolinite/dickite-Kln/Dkt) volcanic groundmass (glassy ash), and microporous pumice fragments (a: Crossed nicols, b: Open nicol). c, d Corroded quartz and opacitized biotite phenocrsyts associated with the argillized volcanic groundmass (c: Crossed nicols, d: Open nicol). e, f Silicified glass shards and pore-filled fine-grained hydrothermal quartz and argillized volcanic groundmass (e: crossed nicols, f: open nicol)

Full size image

Scanning electron microscopy

The scanning electron microscopy (SEM) investigation of argillized samples shows vermicular (vermiform) kaolinite/dickite booklets (Fig. 5a). The stacking thickness of booklets reached up to 30 μm. Kaolinite/dickite minerals exhibit irregular stacking of pseudo hexagonal kaolinite platelets, in which kaolinite and dickite minerals are distinguished by uniform and prolonged shape of plates (Fig. 5b). Fine-grained euhedral alunite crystals (approximately 1 μm in size) are also associated with the kaolinite/dickite minerals (Fig. 5c). The thicknesses of the individual kaolinite/dickite plates smaller than 1 μm (~ 200 nm), and horizontal and/or lateral diameters of pseudo-hexagonal plates are around 2 μm (Fig. 5d). The vermicular packing of kaolinite/dickite plates exhibits sequential (or serrated) stacking.

Fig. 5.
figure 5

SEM photomicrographs of kaolinite/dickites. a Vermicular (vermiform) euhedral kaolinite/dickite booklets. b Tightly irregularly stacked kaolinite/dickite booklets, pseudohexagonal plates of kaolinites and dickites exhibit uniform and prolonged shapes, respectively. c Euhedral alunite crystals and prolonged dickites. d Tightly packed kaolinite/dickite booklets

Full size image

Mineralogy results

X-ray diffraction

The altered ignimbrite samples are divided into three groups, as slightly, moderately, and highly altered, with respect to their alteration grades on the basis of their mesoscopical (plasticity related to argillization), petrographical (primary magmatic vs. alteration related components), mineralogical (clay content), and chemical (loss on ignition values related to clay content) properties. X-ray diffraction (XRD) analyses of the samples having different alteration grades collected from several profiles and/or layers are given in Table 1.

Table 1 Mineralogical composition of bulk and clay-size fraction (< 2 μm) of the samples
Full size table

The samples contain mainly quartz and clay, and minor amounts of alunite and jarosite. The clay fraction is composed of kaolinite and dickite and rarely illite and mixed-layered illite/smectite (I-S) (Fig. 6). The absence of feldspar minerals seems to be related to their being completely transformed to clays during hydrothermal alteration. Subordinate amounts of illite (or mica) and I-S could be associated with the magmatic mica relicts (see Fig. 4c, d) and their transformation product from mica though I-S to finally smectite.

Fig. 6
figure 6

Whole rock and clay fraction of altered ignimbrite sample

Full size image

The kaolinite and dickite minerals are characterized by two reflections at 7.15 and 3.57 Å in the range of 2θ = 2–30° (Fig. 6). The kaolinite and dickite peaks are not affected by ethylene glycol treatments but collapsed upon heating to 550 °C (e.g., Moore and Reynolds 1989). The highly altered ignimbrites have somewhat higher clay (kaolinite and dickite), alunite, and jarosite contents than moderately and slightly altered ones (Table 1, Fig. 7). The diagnostic peaks for both kaolinite and dickite minerals were used for distinguishing peaks for both minerals from unoriented powder patterns as previously preferred by Bailey (1980, 1988) (Fig. 7). The representative diagnostic peaks of kaolinite and dickite minerals are clearly distinguished from schematic X-ray powder diffraction diagrams (Brown and Brindley 1980; p. 330; Fig. 5.2) in the 2θ ranges of 19–27° and 34–42° (Fig. 7).

Fig. 7
figure 7

Representative XRD patterns of whole rock and clay fraction of altered ignimbrite samples. Diagnostic peaks used for distinguishing kaolinite and dickite minerals are taken from Bailey (1980, 1988)

Full size image

Fourier-transform infrared spectrometer

The Fourier-transform infrared spectroscopy (FTIR) properties of altered ignimbrite samples indicate sharp stretching OH regression bands typical for both kaolinite and dickite at 3689, − 3670, − 3662, and 3620 cm−1 (Fig. 8). The 3620 cm−1 band belongs to the internal stresses of the -OH groups between the layers of kaolinite structure. The 1005–1030 cm−1 bands observed in the fingerprint region belong to the Si-O stresses at the center of the kaolinite/dickite crystals. The band of the hydroxyl group on the inner surface of the kaolinite/dickite is 911 cm−1. The weak bands at 745 and 700 cm−1 in the spectra are probably related to the perpendicular Si-O vibrations of the surface hydroxyl layer. The bands around 3000 cm−1 are thought to indicate the presence of alunite (Fig. 8).

Fig. 8
figure 8

FTIR spectra of altered ignimbrite samples

Full size image

Geochemistry results

Major element geochemistry

The major element compositions of altered ignimbrite samples are given in Table 2 and Fig. 9. According to loss on ignition values, three alteration groups are divided as slightly altered (5.00–7.76 wt.%), moderately altered (8.00–8.36 wt.%), and highly altered (10.90–13.08 wt.%) ignimbrites. In order to understand the chemical differences related to alteration, major element concentrations of altered ignimbrites are compared with the non-altered ignimbrite composition (Temel et al. 1998). The major oxide compositions of the altered ignimbrite samples exhibit a similar trend, but some differences with respect to the non-altered sample (Fig. 10). The mean values of TiO2, Al2O3, and P2O5 increase, whereas SiO2, MnO, MgO, CaO, Na2O, and K2O decrease in the altered samples (Table 2, Fig. 9). The enrichment in Al2O3 is resulting from hydrothermal alteration, i.e., kaolinization. From slightly altered to highly altered ignimbrites and finally to theoretical pure kaolinite, SiO2 slightly decreases but Al2O3 and LOI distinctly increase together with the increasing degree of alteration (Fig. 10). The unclear decreasing of the SiO2 contents may be caused by the cryptocrystalline quartz occurrences together with clays. Kaolinite/dickite-rich clay fractions in highly altered ignimbrite samples have excess Si and Fe because of the presence of cryptocrystalline quartz and goethite in the clay fraction, respectively; therefore, their structural formulas did not calculate because of the charge imbalance.

Table 2 Major (wt %) and trace (ppm) element composition of the altered ignimbrites
Full size table
Fig. 9
figure 9

Major oxide (wt. %) concentrations of kaolinites/dickites from ignimbrite samples having different stage of alteration. Fresh (non-altered) ignimbrite composition is taken from Temel et al. (1998)

Full size image
Fig. 10
figure 10

Al2O3, SiO2 and LOI distributions of kaolinites/dickites with regard to different alteration stages of ignimbrites. Fresh (non-altered) ignimbrite and glass composition are taken from Temel et al. (1998)

Full size image

Trace-element and REE compositions

The trace element concentrations of altered ignimbrites are given in Table 2. Some trace element compositions are evaluated on the chondrite-normalized multiple trace element and REE spider diagram (Fig. 11a, b). All elements, except for Rb, Sr, and P, show obvious enrichments. Rb, Ba, and Sr show a broad range, i.e., decreased 0.3 times and enriched up to c. 9400 times, with respect to chondrite (Fig. 11a). However, the contents of trace elements of altered ignimbrites exhibit similar trends with host-rock composition; there are some differences for different grades of alterations with respect to the non-altered host-rock. Chondrite-normalized (Sun and Mc Donough 1989) Rb, K, and Yb average contents are noticeably lower than ignimbrite host-rock (Fig. 11b), indicating element mobility during acidic hydrothermal alteration. Contrary to this, La, Ce, Nd, Sm, and Ti are enriched in altered ignimbrites, which seem to be related to increasing the hydrothermal clay content. In addition to this, some elements (Ba, Sr, Eu, Ti, Yb) show gradual increase together with the alteration degrees. In other words, the concentrations of Ba, Sr, and Eu increase, but Ti and Yb decrease together with the increasing alteration degree (Fig. 11b).

Fig. 11
figure 11

Chondrite-normalized multiple trace element diagram of altered ignimbrite samples. a All samples. b Average values for different alteration degrees. Fresh (non-altered) ignimbrite composition is taken from Temel et al. (1998)

Full size image

The chondrite-normalized REE values of altered ignimbrites have a similar trend with the non-altered ignimbrite sample but exhibit a wide range spider diagrams associated with the different degrees of alteration (Fig. 12a). All of the altered ignimbrites as well as the host-rock exhibit clear enrichment of REE with respect to the chondrite. The REE spider diagram of altered ignimbrites shows a distinct decreasing trend from La to Gd (light rare earth elements, LREE) and a rather stable distribution from Dy to Lu (heavy rare earth elements, HREE) without any negative Eu anomaly (Fig. 12b). The altered ignimbrites have enrichment for LREE but depletion for HREE, when compared to the non-altered ignimbrite host-rock. The highly altered samples have lower REE values than those of slightly and moderately altered ignimbrites, except for Eu, Gd, and Er.

Fig. 12
figure 12

Chondrite-normalized REE compositions of kaolinites/dickites from ignimbrite samples having different stage of alteration. Fresh (non-altered) ignimbrite composition is taken from Temel et al. (1998)

Full size image

Oxygen and hydrogen isotope geochemistry

Oxygen and hydrogen isotope compositions of twelve pure or nearly pure clay-sized (< 2 μm) kaolin samples show that the δ18O and δD ‰ (SMOW) values vary from 3.10 to 9.18 and from 79.97 to 94.31, respectively (Table 3). The calculated δ18O and δD ‰ (SMOW) for mineral-forming water composition (assuming a temperature of 200 °C) from the equations of Sheppard and Gilg (1996) and Gilg and Sheppard (1996) are also given in Table 3. The isotope data of dickite/kaolinite samples and calculated waters are presented in a δ18O vs δD ‰ binary diagram (Fig. 13a), together with the meteoric water line (Craig 1961), supergene-hypogene line (Sheppard et al. 1969), kaolinite line (Savin and Epstein 1970), and magmatic water field (Taylor 1974). Isotopic values of kaolinite/dickite fall completely in the hypogene area indicating hydrothermal conditions rather than weathering processes. The relatively higher dickite-bearing samples (i.e., dickite/(kaolinite+dickite, Dck*> 40%, Table 1) show relatively higher δ18O and lower δD ‰ values than kaolinite-rich (Dck*< 40%) samples. The calculated water compositions on the basis of the assumed 200 °C temperature conditions locate the area between magmatic water field and meteoric water line, representing the mixture of magmatic and meteoric waters (Fig. 13a). The water composition of relatively higher dickite-bearing (Dck*> 40%) samples fall close to magmatic water, whereas kaolinite-rich (Dck*< 40%) samples fall relatively far from magmatic water composition. According to temperature-related kaolinite-water isotopic fractionation, possible local meteoric water composition is determined as δ18O = − 10‰ and δD = − 60‰ (SMOW). If clay-forming water was assumed to be completely magmatic water, the minimum temperature conditions of hydrothermal fluids should be around 250 °C (Fig. 13b). If clay-forming water originated from meteoric waters, the maximum temperature should be around 100 °C. The presence of dickite indicates intermediate phase conditions between kaolinite and pyrophyllite which is relatively higher temperatures than those of kaolinite. Since the formation temperature of dickite should be higher than 100 °C, the clay-forming water could not have originated from entirely meteoric waters; hence, the fluids should be a mixing of meteoric and magmatic waters.

Table 3 δ18O and δD values of kaolinite/dickite and calculated water compositions.
Full size table

pH conditions

The measured pH values of the studied samples (Table 3) change between 4.45 and 6.60 and generally show moderately low values around pH 5. The differences in pH values for samples seem to be related to different amounts of kaolin group minerals, i.e., associated minerals such as quartz and illite.

Discussion

Hydrothermal clay occurrences

The hydrothermally altered tuff lithologies of the Kavak ignimbrite unit have mainly preserved primary porphyry texture, that indicate the in-situ alteration in a temperature dominant (steam-heated) environment. The Kavak ignimbrite unit is the oldest and lowermost level of the Ürgüp formation (Fig. 1), therefore following the deposition of the Kavak ignimbrite unit (~ 9 Ma, Aydar et al. 2012) the volcanic activity continued to the Quaternary (Temel 1992) with duration of ~ 6 Ma, at least. Hydrothermal alteration of quartz and clay (kaolinite and dickite) occurred. Fine-grained (cryptocrystalline) quartz was developed within the pores in the groundmass and replaced by glass shards, as well. The kaolinite/dickite and association with alunite indicate the volcanism-related steam-heated surficial environments (Reyes 1990; Hedenquist et al. 2000; Simón et al. 2005; Ece et al. 2013; Ercan 2016; Mayer et al. 2016). Alunite minerals are formed as alteration products of Al-bearing minerals in relatively oxidizing, acid-sulfate conditions in the advanced argillic alteration zone (Hemely 1959; Knight 1977; Scott 1990). Its occurrences under surficial environments indicates the presence of acid sulfate fluids from H2S gas, which evolved from a boiling hydrothermal system at depth (Reyes 1990). The very fine-grained pseudo cubic alunite crystals (Fig. 5c) seem to be related to acidic steam-heated waters.

According to petrographical observations, volcanic groundmass and volcanic glass shards as well as pumice grains completely devitrified and transformed to silica and kaolin minerals. This process could be explained as volcanic glass + water → hydrous Al-silicate gel → kaolinite/dickite reaction that is similar to those of tonsteins (Bohor and Triplehorn 1993; Spears 2012) as well as volcanic-hosted kaolin occurrences (Murray 1988; Yalçın and Bozkaya 2003):

$$ {\displaystyle \begin{array}{c}2\mathrm{Si}\mathrm{O}2.\mathrm{Al}2\mathrm{O}3+7\mathrm{H}2\mathrm{O}\to \mathrm{Al}2\left(\mathrm{OH}\right)6.2\mathrm{Si}\left(\mathrm{OH}\right)4\to \mathrm{Al}2\mathrm{Si}2\mathrm{O}2\left(\mathrm{OH}\right)2+5\mathrm{H}2\mathrm{O}\kern0.5em \mathrm{Volcanic}\ \mathrm{glass}\\ {}\mathrm{Al}-\mathrm{silicate}\ \mathrm{gel}\kern2.25em \mathrm{Kaolinite}/\mathrm{dickite}\end{array}} $$

The absence of feldspars in altered ignimbrites indicates that the kaolin minerals also developed from feldspars under acidic environments:

$$ 2\left(\mathrm{K},\mathrm{Na},\mathrm{Ca}\right)\mathrm{AlSi}2-3\mathrm{O}8+\mathrm{H}2\mathrm{O}+2\mathrm{H}+\to \mathrm{Al}2\left(\mathrm{OH}\right)6.2\mathrm{Si}\left(\mathrm{OH}\right)4\to 2\mathrm{Si}\mathrm{O}2+2\left(\mathrm{K}+,\mathrm{Na}+,\mathrm{Ca}2+\right) $$

Regarding the kaolinite-to-dickite transformation in sedimentary environments, the temperature at which kaolinite transforms to dickite is 80 to 200 °C (Dunoyer de Segonzac 1970; Ruiz Cruz and Reyes 1998), 120–130 (Ehrenberg et al. 1993), and 140 °C (Keller 1988). However, this transformation in the hydrothermal system is still unknown because mixtures of dickite with kaolinite are not commonly encountered in this environment. Several studies indicate some temperature data for this transformation, i.e., the temperature of hydrothermal alteration was estimated from 150 °C (Marumo 1989) to 270–300 °C (Koh and Chang 1997; Papoulis and Tsolis-Katagas 2008). Palinkaš et al. (2009) determined 210–250 °C and 290–330 °C conditions from fluid inclusion data for kaolinite and dickite, respectively, in a metal sulfur deposit. The kaolinite-to-dickite transformation was controlled mostly by temperatures nearing the onset of hypogene conditions > 100 °C or dickite in kaolin-alunite hydrothermal deposits in Çanakkale, Turkey (Ercan et al. 2016). According to the zonation of hydrothermal kaolin group minerals with increasing temperature geothermal systems (Reyes 1990), temperature distributions were given as kaolinite (< 120 °C), dickite ± kaolinite (120–200 °C), dickite + pyrophyllite (200–250 °C), and pyrophyllite + illite (230–320 °C). Thus, dickite forms in a transitional setting between kaolinite and pyrophyllite temperature ranges. Therefore, in this study, the maximum temperature condition for kaolinite + dickite is assumed as 200 °C for estimating of the origin of hydrothermal fluids (see Fig. 13a).

pH conditions of hydrothermal fluids should be moderately acidic (approximately pH=4) because of the developments of kaolin group minerals and alunite (Reyes 1990). The pH values of the studied samples are around 5 and correspond to acidic hydrothermal alteration conditions. The acidic condition was the principal control for obstructing the smectite clays, as well as depletion of soluble elements (Na, K, and Ca) ions from ignimbrites (Fig. 9) under an acid leaching environment. The mineralogical composition (dominance with silica and kaolin group minerals, and scarce alunite and jarosite) and pH data indicate typical advanced argillic to argillic alteration conditions under the conditions of moderate to high (150–200 °C) temperature and low pH conditions.

SiO2 vs. Al2O3 relationships between slightly altered and highly altered ignimbrites show a clear geochemical trend from ignimbrite to hydrothermal kaolin transformation (Fig. 10). In addition to this, the broad ranges of Ba and Sr contents of altered ignimbrites are other tools for alteration degree determination (Table 2, Fig. 11). The anomalous increase of these elements in the highest grade altered samples (LOI > 12.5% wt.) should be associated with the grade of hydrothermal alteration as previously mentioned by Dill et al. (1997, 2000). Chondrite normalized LREE contents increase, whereas HREE contents decrease in altered ignimbrites with respect to non-altered ignimbrites. The different contents of REE values for different grades of alterations, such as lower REE values for highly altered ignimbrites, are other geochemical indicators for alteration.

The dominance of kaolinite/dickite, but negligible amounts of illite and mixed-layered illite-smectite minerals, should be controlled by several parameters: (1) low pH and alkalinity conditions of hydrothermal solutions, (2) intense acid leaching, (3) high silica saturation, (4) high-grade activities of ions (H+, H4SiO4, Al(OH)4, etc.) and waters, (5) high dissolution rates of volcanic materials, and (5) composition of volcanic material, i.e., low Fe and/or Mg, whereas high Si and Al, and cation ratios such as K/Na = 0, Si/Al = 1 (Murray 1988; Murray and Keller 1993; Yalçın and Bozkaya 2003; Kadir et al. 2013).

Origin of hydrothermal fluids

The major and trace element geochemical data of altered ignimbrites show a similar composition to non-altered host-rocks, except for somewhat increasing the immobile elements and decreasing mobile elements (Figs. 9, 11, and 12). This data indicates the slight leaching by silica-rich acidic (pH 4–6) hydrothermal solutions. The acidic character of hydrothermal fluids should be derived from silica-rich magmatic fluids which circulated through the acidic pyroclastics rocks with dacite and/or rhyolite composition (Temel 1992; Ertek and Öner 2008).

Acid-rich magmatic vapors seem to be responsible for leaching of the host rock and the advanced argillic alteration. The oxygen and hydrogen isotope data demonstrate that magmatic vapor phases were mixed with meteoric waters for assumed temperature conditions as high as 200 °C (Fig. 13a). The steam-related volcanic environments are the main source responsible for magmatic vapors which circulated along the crack zones within the earliest ignimbrite strata during the approximately 6 Ma duration magmatic period of the Ürgüp formation (Temel 1992).

Fig. 13
figure 13

The distributions of the oxygen and hydrogen isotope data of kaolinite/dickite minerals in δ18O ‰ vs. δD ‰ diagram (Supergene-hypogene line: Sheppard et al. 1969; kaolinite line: Savin and Epstein 1970; Magmatic water box: Taylor 1979; Local meteoric water is assumed from kaolinite-water equilibrium according to equations of (Savin and Lee 1988 for oxygen and Gilg and Sheppard 1996 for hydrogen). Dck* = dickite/(dickite + kaolinite)

Full size image

Conclusions

Miocene-Quaternary volcanic complexes associated ignimbrites from the Cappadocia Volcanic Province (CAVP) have been subjected to hydrothermal alteration and produced industrial kaolin group (kaolinite, dickite) clay occurrences. The mineralogical composition of hydrothermally altered samples are represented mainly by quartz and kaolin group clays (kaolinite and dickite), and minor amounts of goethite/limonite, alunite, jarosite, illite, and mixed-layered illite-smectite. Kaolinite, dickite, and quartz, and scarce alunite, jarosite associations suggest that the hydrothermal alteration occurred in acidic (pH < 5) and moderate temperature (approximately 200 °C) conditions, i.e., argillic alteration. The fine-grained kaolinite and dickite and their associations with the fine-grained alunite and jarosite minerals are characteristic for steam-heated related shallow depth environments. Kaolinite and dickite minerals exhibit uniform and prolonged pseudo-hexagonal platelets, respectively, as tightly packed vermicular booklets. Hydrothermal alteration-related silica and kaolin minerals were developed as the mechanism of replacement of volcanic glass/groundmass and neoformation in the pores from the acidic hydrothermal solutions.

The altered samples are classified into three groups based on their loss on ignition values, i.e., slightly, moderately, and highly altered samples. Altered ignimbrites have enrichment for major oxides of TiO2, Al2O3, and P2O5, but depletion for SiO2, MnO, MgO, CaO, Na2O, and K2O with respect to host-rock composition. Most of the trace element concentrations were also enriched in altered samples except for Rb, Sr, and P. The noticeable decrease in some elements, such as Rb, K, and Yb, are related element mobility during acidic hydrothermal alteration. Chondrite-normalized REE values show a similar trend but have some differences for altered and host-rock compositions. The contents of LREE increase, but HREE decrease in altered ignimbrites. The oxygen and hydrogen isotope data of kaolinite/dickite demonstrate that magmatic vapor phases mixed with meteoric waters. The steam-heated igneous environment with a duration of 6 Ma seems to be the reason for hot magmatic vapors circulating along the cracks zones through the acidic volcanogenic rocks, and consequently the acidic (low pH) conditions which caused to kaolin-dominant clay alteration.