Mineralogical Constraints on the Potassic and Sodic-Calcic Hydrothermal Alteration and Vein-Type Mineralization of the Maronia Porphyry Cu-Mo ± Re ± Au Deposit in NE Greece

: The Maronia Cu-Mo ± Re ± Au deposit is spatially related to a microgranite porphyry that intruded an Oligocene monzonite along the Mesozoic Circum-Rhodope belt in Thrace, NE Greece. The magmatic rocks and associated metallic mineralization show plastic and cataclastic features at the south-eastern margin of the deposit that implies emplacement at the ductile-brittle transition, adjacent to a shear zone at the footwall of the Maronia detachment fault. The conversion from ductile to brittle deformation caused a rapid upward magmatic ﬂuid ﬂow and increased the volume of water that interacted with the host rocks through high permeable zones, which produced extensive zones of potassic and sodic-calcic alteration. Potassic alteration is characterized by secondary biotite + K-feldspar (orthoclase) + magnetite + rutile + quartz ± apatite and commonly contains sulﬁdes (pyrite, chalcopyrite, pyrrhotite). Sodic-calcic alteration consists of actinolite + sodic-calcic plagioclase (albite / oligoclase / andesine) + titanite + magnetite + chlorite + quartz ± calcite ± epidote-allanite. The high-oxidation state of the magmas and the hydrothermal ﬂuid circulation were responsible for the metal and sulfur enrichments of the aqueous ﬂuid phase, an increase in O 2 gas content, the breakdown of the magmatic silicates and the production of the extensive potassic and sodic-calcic alterations. Brittle deformation also promoted the rapid upward ﬂuid ﬂow and caused interactions with the surrounding host rocks along the high temperature M-, EB-, A- and B-type veins. of the sodic-calcic alteration in the monzonite; ( b , c ) Magmatic biotite (bt-m) altered to hydrothermal actinolite (act), titanite (tit), chlorite (chl) and magnetite (mt) of the sodic-calcic alteration in the sheared monzonite; ( d ) Euhedral magmatic amphibole selectively altered to hydrothermal biotite (bt) and magnetite (mt). Primary K-feldspar (Kfs) is also present. Potassic alteration in the monzonite; ( e ) Magmatic biotite (bt-m) altered to hydrothermal biotite (bt), K-feldspar (Kfs), magnetite (mt) and pyrite (py) of the potassic alteration in the monzonite; ( f ) Hydrothermal titanite (tit), biotite (bt) and magnetite (mt) of the potassic alteration in the sheared monzonite; ( g , h ) EB-type veins (EB) and hydrothermal biotite (bt) of the potassic alteration in the microgranite porphyry; ( i ) Hydrothermal biotite (bt) and magnetite (mt) in EB-vein enveloped with a quartz-orthoclase (qz + Kfs) halo of the potassic alteration in the microgranite porphyry; ( j ) Hydrothermal orthoclase (Kfs) around granular quartz (qz) from the potassic alteration in microgranite porphyry; ( k ) Needle-like sagenitic rutile (rt-s) in chlorite (chl), which has totally replaced precursor magmatic biotite in potassic altered microgranite; ( l ) Rutile grain in potassic altered microgranite, showing sector zoning due to heterogeneous chemical composition. Brighter areas represent sectors with higher SrO contents (1.81–3.63 wt. %); ( m , n ) Chalcopyrite (cp) and magnetite (mt) associated with secondary biotite (bt) which is replaced by chlorite (chl), in potassic altered microgranite;

. Simplified map of Northern Greece showing the main geotectonic zones, the major tectonic structures within the Hellenides, and the distribution of porphyry-type deposits (after Melfos and Voudouris [8]; Voudouris et al. [9]). The Maronia area is outlined and the geotectonic zones of the Hellenic orogen are shown in the inset.
In this region, numerous porphyry Cu-Mo ± Au ± Re deposits formed from calc-alkaline to shoshonitic magmas derived from partial melting of previously subduction-modified sub-lithospheric mantle and/or the lower crust [8,9]. The deposits are distributed around metamorphic core complexes bound by low-angle detachment faults. Such faults promoted the ascent of magmas and the circulation of hydrothermal fluids, which ultimately resulted in the formation of the porphyry-style mineralization.
One of these porphyry-type deposits occurs in the Rhodope massif, 5 km SE of Maronia village, along the Aegean coast (Figures 1 and 2). The geological setting of the area is characterized by the main geotectonic zones, the major tectonic structures within the Hellenides, and the distribution of porphyry-type deposits (after Melfos and Voudouris [8]; Voudouris et al. [9]). The Maronia area is outlined and the geotectonic zones of the Hellenic orogen are shown in the inset.
In this region, numerous porphyry Cu-Mo ± Au ± Re deposits formed from calc-alkaline to shoshonitic magmas derived from partial melting of previously subduction-modified sub-lithospheric mantle and/or the lower crust [8,9]. The deposits are distributed around metamorphic core complexes bound by low-angle detachment faults. Such faults promoted the ascent of magmas and the circulation of hydrothermal fluids, which ultimately resulted in the formation of the porphyry-style mineralization.
One of these porphyry-type deposits occurs in the Rhodope massif, 5 km SE of Maronia village, along the Aegean coast (Figures 1 and 2). The geological setting of the area is characterized by the low metamorphic grade Mesozoic Circum Rhodope metamorphic belt, which overlies the upper unit of the Paleozoic-Mesozoic high-grade metamorphic basement of the Rhodope massif [28,29]. The Circum Rhodope belt comprises metasedimentary (metaconglomerates, metagreywackes, metaquartzites, marbles, calc-schists) and metavolcanic rocks (greenschists, chlorite schists, talc schists, mica schists, quartzites) of Triassic to Jurassic protolith age.
The Maronia pluton was intruded by the Ktismata microgranite, which is spatially and genetically related to the porphyry style Cu-Mo ± Re ± Au mineralization ( Figure 2). The emplacement of the microgranite was structurally controlled by three sets of faults trending NE-SW, NW-SE, and WNW-ESE. Dips are mostly steep to vertical with a minor strike-slip component, and as a consequence the contacts between the two igneous intrusions are sharp and well-defined. According to Melfos et al. [11], the microgranite porphyry comprises phenocrysts of quartz, K-feldspar and plagioclase in a fine-grained matrix of quartz, plagioclase and K-feldspar. Argillic, sericitic, and propylitic alteration styles were initially recognized by Melfos et al. [11]. Magnetite, pyrite, chalcopyrite, and molybdenite are the main ore minerals. Molybdenite is enriched in Re, with concentrations reaching up to 2.88 wt. % Re [11,12]. Pyrrhotite and bismuthinite are minor minerals. Sphalerite, galena, cubanite, boulangerite, bournonite, tetrahedrite/tennantite, chalcostibite, zinkenite, and famatinite occur in the late high-sulfidation epithermal veins that crosscut the porphyry system [11].
Exploration drilling on the Maronia prospect conducted by the Hellenic Chemical Products and Fertilizers Co Ltd. in 1973 involved four drill holes comprising a total length of 550 m. Despite the encouraging Cu and Au grades, the project was terminated without any resource estimate or feasibility study, possibly because the Maronia prospect is located in an extensive archaeological site or because its immediate vicinity with the Aegean Sea. The cores do not exist anymore and only some reports describe the mineralogical and geochemical results of the drilling [11].
Surface samples contain as much as 5460 ppm Cu, 7600 ppm Mo, >1 ppm Re, 49 ppm Sb, 1.7 ppm Hg, 0.18 ppm Te, 0.9 ppm Ga, and 1 ppm Au [9,11]. Geochemical data from one drill core showed a 10 m intercept grading 2 wt. % Cu, 12 ppm Au, 17 ppm Ag, which was probably associated with the high-sulfidation epithermal overprint. This epithermal event was also characterized by the presence of famatinite, tennantite, and pyrophyllite [8,11]. Boiling was considered as the main mechanism of the ore mineral precipitation occurring at temperatures between 280 and 460 • C, and pressures from 150 to 510 bar [11].

Sampling and Analytical Methods
Surface sampling, with particular focus on spatial relationships among the various types of hydrothermal alteration in different rock types, was carried out in the Maronia porphyry system. A set of 20 samples was collected, representing mainly the intensively altered rocks. Older surface samples from a previous study [11] were also considered for the mineralogical composition of the alterations.

Structural Control on the Mineralization
There is a close relationship of the Maronia detachment fault with the porphyry style mineralization. The approximately E-W striking fault that shallowly dips to the south (Figure 3a) marks the southernmost boundary of the extensive detachment fault system. The fault active from 42 to 12 Ma exhumed the Biala Reka-Kechros high-grade metamorphic core complex in the north [32,36] ( Figure 1). However, only upper crustal low-grade metamorphic rocks, which were described as part of the Circum Rhodope belt, were exhumed in the south [28]. Bonev et al. [29] suggested that the uppermost low-grade Mesozoic Circum Rhodope belt constitutes a part of the hanging wall of the Tertiary extensional detachment system, whereas the lower high-grade metamorphic basement forms the footwall.
Minerals 2020, 10, x FOR PEER REVIEW 6 of 26 the Tertiary extensional detachment system, whereas the lower high-grade metamorphic basement forms the footwall. The detachment fault is characterized by brecciation of the marble in the hanging wall and mylonitization of the calc-schist in the footwall (photographed towards SW); (b) High-angle supra-detachment normal fault in marble in the hanging wall of the detachment along the Aegean coast (photograph towards W); (c) High-angle supra-detachment normal fault along which the microgranite porphyry intruded (photographed towards E); (d) Sodic-calcic alteration consisting of hydrothermal actinolite (act) and biotite (bt) in highly strained and mylonitized monzonite, and an aplitic dike (apl) along the foliation; (e) Weak potassic alteration consisting of fine-grained hydrothermal biotite (bt) in highly strained microgranite porphyry, overprinted by strong sericitic alteration crosscut by a pyrite vein (py); (f) Mineralized aplitic dike (apl) that intruded mylonitized monzonite along the foliation; (g) Potassic alteration of the microgranite porphyry with secondary biotite (bt) at the south-eastern margin of the porphyry system; (h) M-type vein with hydrothermal magnetite (mt) and actinolite (act) associated with the sodic-calcic alteration in the sheared monzonite (mnz) and the aplitic dike (apl). Pyrite (py) (D-type vein) formed due to the re-opening of the vein during later sericitic alteration; (i) Barren A-type uninterrupted sinuous quartz veins associated with potassic alteration crosscut by a B-vein with pyrite-molybdenite mineralization in the microgranite porphyry.
Features that characterize the Maronia detachment include brecciation of the marble in the upper plate of the hanging wall, mylonitization of the calc-schist of the lower plate in the footwall (Figure 3a), and the listric and planar high-angle supra-detachment normal faults that affected metamorphic and magmatic rocks along the Aegean coast (Figures 2 and 3b,c). These supra-detachment faults extend mainly SE-NW. High-angle supra-detachment normal fault in marble in the hanging wall of the detachment along the Aegean coast (photograph towards W); (c) High-angle supra-detachment normal fault along which the microgranite porphyry intruded (photographed towards E); (d) Sodic-calcic alteration consisting of hydrothermal actinolite (act) and biotite (bt) in highly strained and mylonitized monzonite, and an aplitic dike (apl) along the foliation; (e) Weak potassic alteration consisting of fine-grained hydrothermal biotite (bt) in highly strained microgranite porphyry, overprinted by strong sericitic alteration crosscut by a pyrite vein (py); (f) Mineralized aplitic dike (apl) that intruded mylonitized monzonite along the foliation; (g) Potassic alteration of the microgranite porphyry with secondary biotite (bt) at the south-eastern margin of the porphyry system; (h) M-type vein with hydrothermal magnetite (mt) and actinolite (act) associated with the sodic-calcic alteration in the sheared monzonite (mnz) and the aplitic dike (apl). Pyrite (py) (D-type vein) formed due to the re-opening of the vein during later sericitic alteration; (i) Barren A-type uninterrupted sinuous quartz veins associated with potassic alteration crosscut by a B-vein with pyrite-molybdenite mineralization in the microgranite porphyry.
Features that characterize the Maronia detachment include brecciation of the marble in the upper plate of the hanging wall, mylonitization of the calc-schist of the lower plate in the footwall (Figure 3a), and the listric and planar high-angle supra-detachment normal faults that affected metamorphic and Minerals 2020, 10, 182 7 of 26 magmatic rocks along the Aegean coast (Figures 2 and 3b,c). These supra-detachment faults extend mainly SE-NW.
The Maronia microgranite porphyry was emplaced along the footwall of the detachment fault producing horizontal displacement along ductile to brittle shear zones. The most extended exposure of these shear zones occurs at the south-eastern margin of the porphyry system along the coast ( Figure 2). In this site, monzonite and microgranite porphyry are highly strained and mylonitized with well-banded and linear textures (Figure 3d-f). Several aplitic dikes, from a few centimeters to more than 1 m thick, intrude the pluton and extend into the shear zone along ductile ( Figure 3f) and brittle structures. These dikes are attributed to the intrusion of the microgranite porphyry along the foliation in the sheared monzonite. All these events demonstrate a syntectonic emplacement of the microgranite in the Maronia pluton. The aplitic-granitic dikes host miarolitic cavities with sulfides, suggesting volatile-rich residual melts exsolved from the granitic magma, which are probably related to the ore formation.
The following field and petrographic study are mainly based on new results but also considers previous research [11].

Potassic and Sodic-Calcic Hydrothermal Alteration
The hydrothermal alteration of the Maronia deposit is pervasive and consists of five main types in the microgranite and along the contacts with the monzonite. The alteration types include potassic, sodic-calcic, propylitic, sericitic, and argillic alteration, with local silicification (Figure 4). In this study, we focus on the potassic and sodic-calcic alteration. The Maronia microgranite porphyry was emplaced along the footwall of the detachment fault producing horizontal displacement along ductile to brittle shear zones. The most extended exposure of these shear zones occurs at the south-eastern margin of the porphyry system along the coast ( Figure 2). In this site, monzonite and microgranite porphyry are highly strained and mylonitized with well-banded and linear textures (Figure 3d-f). Several aplitic dikes, from a few centimeters to more than 1 m thick, intrude the pluton and extend into the shear zone along ductile ( Figure 3f) and brittle structures. These dikes are attributed to the intrusion of the microgranite porphyry along the foliation in the sheared monzonite. All these events demonstrate a syntectonic emplacement of the microgranite in the Maronia pluton. The aplitic-granitic dikes host miarolitic cavities with sulfides, suggesting volatile-rich residual melts exsolved from the granitic magma, which are probably related to the ore formation.
The following field and petrographic study are mainly based on new results but also considers previous research [11].
At Maronia, mineral assemblages of these two alteration types do not demonstrate any clear temporal relationship, suggesting that they formed simultaneously. Weak potassic alteration is concentrated mainly in the northern, south-western, and south-eastern margins of the microgranite porphyry (Figures 3g and 4). However, a strong sodic-calcic alteration affected the monzonite and aplitic dikes along the south-eastern margin of the porphyry system (Figure 3f). At the south-eastern margin of the porphyry system, an apparent tectonic control on the potassic alteration occurs in the mylonitized microgranite porphyry and the sodic-calcic alteration in the mylonitized monzonite (Figure 3d-f). The potassic and sodic-calcic alteration assemblages in Maronia are superimposed by subsequent sericitic alteration (Figure 3e,g-i and Figure 4). The highest copper and molybdenum grades (up to 5460 ppm Cu and 7600 ppm Mo) are mainly associated with the zones of very dense quartz stockwork (e.g., silicification) in the potassic alteration.

Vein Types
New field observations revealed different types of veins associated with the alteration events in the Maronia deposit ( Table 1). The mylonitized monzonite is crosscut by mineralized aplitic-granitic and microgranitic veins and dikes, varying in thickness from a few centimeters to 1 m, which host pyrite, chalcopyrite and pyrrhotite ( Figure 3f). Additionally, a total of six early to late stage hydrothermal vein types were identified: M-, EB-, A-, B-, D-and epithermal-type veins, using the vein terminology proposed by [1]. Apart from the EB-veins, all the other veins do not exhibit alteration halos. Up to 5 cm wide M-type veins crosscut the monzonite and the aplitic dikes along the shear zone. These veins contain various proportions of secondary actinolite, sodic plagioclase, magnetite, epidote-allanite, calcite and titanite (Figure 3h), and are spatially coincident with the sodic-calcic Minerals 2020, 10, 182 9 of 26 alteration, with which they are likely genetically related. Pyrite rarely occurs in M-type veins, but is present in the subsequent sericitic alteration overprint.
Early biotite (EB) veinlets (1-5 mm wide) consist of fine-grained biotite and magnetite (Figure 5a-c) and are spatially related to potassically altered microgranite porphyry (Figure 5b). A-type uninterrupted sinuous veins and B-type veins (1-3 cm wide) consist of granular quartz and occur in both the deformed and the undeformed parts of the potassically altered microgranite porphyry (Figure 3i). These veins are mostly barren but in places they contain sulfide mineralization consisting of pyrite + magnetite + chalcopyrite + pyrrhotite in the A-type veins, and pyrite + chalcopyrite + molybdenite in the B-type veins (Figure 3i).

Mineral Composition and Chemistry of Potassic and Sodic-Calcic Alteration
Potassic alteration in the microgranite and the monzonite is characterized by secondary biotite + K-feldspar (orthoclase) + magnetite + rutile + quartz ± apatite. This alteration assemblage is the same in the two magmatic rocks, and only variations in the mineral proportions were observed, which are related to the alteration intensity. Sodic-calcic alteration in the monzonite is characterized by the assemblage of actinolite + sodic-calcic plagioclase (albite/oligoclase/andesine) + titanite + magnetite + chlorite + quartz ± calcite ± epidote-allanite.
Two types of biotite were analyzed: (1) primary magmatic biotite from the monzonite and (2) secondary biotite from the monzonite and the microgranite porphyry. Primary magmatic biotite forms platy crystals up to 3 mm in size in the monzonite matrix and is usually well preserved. However, in some places with sodic-calcic alteration, it was dissolved and newly alteration minerals in various amounts precipitated, mainly secondary chlorite, actinolite, and titanite (Figure 5a [42]. For example, the TiO 2 content of biotite ranges from 4.37 to 6.49 wt. % and the Mg/(Mg + Fe) ratios from 0.55 to 0.76.
The most intense development of secondary biotite occurs at the eastern margin of the porphyry system along the sheared monzonite and microgranite porphyry. Macroscopically, the conversion of the greenish gray granular texture of the monzonite to a black biotite-rich mylonite (Figure 3d,f) is the most dramatic visual presentation of the hydrothermal biotite. It is distinguished from magmatic biotite by shreddy texture and fine-grained (typically <100 µm) anhedral morphology in the matrix where it occurs as disseminations, patches, small aggregate clusters or veinlets (EB-veins), up to 1 mm wide (Figure 5d-i). Hydrothermal biotite replaces primary magmatic pyroxene, biotite and hornblende in the monzonite (Figure 5d-f). It also forms veins and veinlets and fills interstices between phenocrysts in the microgranite (Figure 5g-i). Microscopically, hydrothermal biotite commonly co-exists with secondary magnetite, K-feldspar, rutile and quartz. Quartz and orthoclase rarely form alteration halos along the veinlets (Figure 5i).
The chemical composition of hydrothermal biotite is shown in Table 2 and Figures 6 and 7. In a plot of Fe/(Fe + Mg) versus Al IV , the compositions of the biotite in the microgranite porphyry plot along the boundary between biotite and phlogopite fields (with Al IV 1.54-2.10 apfu), whereas biotite in the monzonite partly overlaps the field of the magmatic biotite ( Figure 6a). However, hydrothermal biotite can be discriminated from the magmatic biotite by its lower Al IV content (Figure 6a). The TiO 2 content is lower than that in the magmatic biotite, and similar to the biotite from the microgranite porphyry and monzonite. In both rock types, the hydrothermal biotite contains between 2.64 and 4.66 wt. % TiO 2 , corresponding to 0.30-0.52 apfu Ti (Figure 6b,c).  which has totally replaced precursor magmatic biotite in potassic altered microgranite; (l) Rutile grain in potassic altered microgranite, showing sector zoning due to heterogeneous chemical composition. Brighter areas represent sectors with higher SrO contents (1.81-3.63 wt. %); (m,n) Chalcopyrite (cp) and magnetite (mt) associated with secondary biotite (bt) which is replaced by chlorite (chl), in potassic altered microgranite; (o) Chalcopyrite (cp) and pyrrhotite (po) inclusions in pyrite (py) as part of the potassic alteration mineral assemblage, in microgranite.  (Figure 6b).
A negative correlation between Ti and Si occurs in hydrothermal biotite, but it is not as significant as in the magmatic biotite (Figure 6c). Similar negative correlations of magmatic and hydrothermal biotite were observed in porphyry deposits elsewhere (e.g., [44,45]). The higher SiO2 content in the secondary biotite from the microgranite can be explained by the removal of Si from the quartz-rich host granitic rock. Up to 0.38 wt. % Cl contents in hydrothermal biotite overlap those of A negative correlation between Ti and Si occurs in hydrothermal biotite, but it is not as significant as in the magmatic biotite (Figure 6c). Similar negative correlations of magmatic and hydrothermal biotite were observed in porphyry deposits elsewhere (e.g., [44,45]). The higher SiO 2 content in the secondary biotite from the microgranite can be explained by the removal of Si from the quartz-rich host granitic rock. Up to 0.38 wt. % Cl contents in hydrothermal biotite overlap those of magmatic biotite (up to 0.34 wt. % Cl). The Fe/(Fe + Mg) ratio of the hydrothermal biotite in microgranite porphyry ranges from 0.30 to 0.40 and is relatively higher than that of the magmatic biotite (Figure 6a). However, the ratio in hydrothermal biotite from the monzonite (Fe/(Fe + Mg) = 0.22-0.28) overlaps with that in magmatic biotite.
The ternary diagram 10 × TiO 2 -(FeO + MnO)-MgO proposed by Nachit et al. [46], is used here as a quantitative geochemical tool for discriminating the primary magmatic biotites from those that are newly formed (neoformed) by the action of the hydrothermal fluids. Biotites from Maronia plot in the fields of magmatic and newly formed hydrothermal biotites, demonstrating their different origin ( Figure 7). However, the composition of hydrothermal biotite from the microgranite porphyry is indistinguishable from the hydrothermal biotite in the monzonite. Therefore, all the newly grown biotites in the potassic alteration stage probably formed in equilibrium with the hydrothermal fluids. Similar geochemical behavior has been observed in hydrothermal biotites from other porphyry deposits [47]. microgranite porphyry ranges from 0.30 to 0.40 and is relatively higher than that of the magmatic biotite ( Figure 6a). However, the ratio in hydrothermal biotite from the monzonite (Fe/(Fe + Mg) = 0.22-0.28) overlaps with that in magmatic biotite.
The ternary diagram 10 × TiO2-(FeO + MnO)-MgO proposed by Nachit et al. [46], is used here as a quantitative geochemical tool for discriminating the primary magmatic biotites from those that are newly formed (neoformed) by the action of the hydrothermal fluids. Biotites from Maronia plot in the fields of magmatic and newly formed hydrothermal biotites, demonstrating their different origin ( Figure 7). However, the composition of hydrothermal biotite from the microgranite porphyry is indistinguishable from the hydrothermal biotite in the monzonite. Therefore, all the newly grown biotites in the potassic alteration stage probably formed in equilibrium with the hydrothermal fluids. Similar geochemical behavior has been observed in hydrothermal biotites from other porphyry deposits [47]. Subhedral to anhedral grains of K-feldspar in the potassic alteration of the Maronia porphyry system is mainly orthoclase, which is commonly intergrown with biotite, magnetite, quartz and pyrite. In EB-type veins, orthoclase-quartz assemblages may form alteration halos around biotite (Figure 5i). Orthoclase occurs interstitially around quartz grains in potassically altered microgranite porphyry (Figure 5j). Microanalyses show that the hydrothermal K-feldspar has a composition ranging from Or81 to Or97 (Table 3, Figure 8).
Rutile, which is associated with potassic alteration occurs as disseminated grains up to 0.5 mm in size or more rarely as tiny needle-like sagenite with chlorite in magmatic biotite precursor ( Figure  5k Table 4). Rutile crystals generally exhibit well-developed zoning patterns with sectors of heterogeneous chemical composition related to the enrichment of several minor elements (Figure 5l). These elements were incorporated into the rutile lattice during crystal growth resulting in a sector-zoning pattern, which has previously been described by Rabbia et al. [48] and Rabbia and Hernandez [49] at porphyry-type deposits elsewhere. The light grey zones as revealed by SEM-back scattering images (BSE) are related to high Sr concentrations (SrO 1.81-3.63 wt. %). Subhedral to anhedral grains of K-feldspar in the potassic alteration of the Maronia porphyry system is mainly orthoclase, which is commonly intergrown with biotite, magnetite, quartz and pyrite. In EB-type veins, orthoclase-quartz assemblages may form alteration halos around biotite (Figure 5i). Orthoclase occurs interstitially around quartz grains in potassically altered microgranite porphyry (Figure 5j). Microanalyses show that the hydrothermal K-feldspar has a composition ranging from Or 81 to Or 97 (Table 3, Figure 8).
Rutile, which is associated with potassic alteration occurs as disseminated grains up to 0.5 mm in size or more rarely as tiny needle-like sagenite with chlorite in magmatic biotite precursor (Figure 5k Table 4). Rutile crystals generally exhibit well-developed zoning patterns with sectors of heterogeneous chemical composition related to the enrichment of several minor elements (Figure 5l). These elements were incorporated into the rutile lattice during crystal growth resulting in a sector-zoning pattern, which has previously been described by Rabbia et al. [48] and Rabbia and Hernandez [49] at porphyry-type deposits elsewhere. The light grey zones as revealed by SEM-back scattering images (BSE) are related to high Sr concentrations (SrO 1.81-3.63 wt. %).
Potassic alteration commonly contains magnetite and sulfides (pyrite, chalcopyrite, pyrrhotite) associated with secondary biotite (Figure 5m,n). Magnetite in the potassic alteration zone forms disseminated subhedral to anhedral grains (50-300 µm in size) that are mainly associated with biotite (Figure 5a-f,i,m-o). Magnetite is intergrown with pyrite, chalcopyrite and pyrrhotite (Figure 5m,n). Chalcopyrite is an abundant sulfide in potassically altered rocks and occurs mainly as disseminated anhedral grains (10-500 µm in size) within pyrite grains. It is intergrown with magnetite, pyrite, and pyrrhotite (Figure 5m-o). Disseminated euhedral to subhedral prismatic crystals of hydrothermal actinolite, up to 0.5 mm in length, are associated with the sodic-calcic alteration in the monzonite. Actinolite replaced magmatic biotite and pyroxene, and coexists with hydrothermal plagioclase, titanite, epidote-allanite, magnetite, calcite and chlorite (Figure 5a-c). It is the main mineral phase in actinolite + magnetite ± pyrite ± epidote-allanite ± titanite M-type veins (Figure 3h). Microanalyses show a typical actinolite composition (Table 5), with CaO ranging from 12.41 to 13.80 wt. % and MgO from 19.68 to 19.82 wt. %. Ferric iron was estimated using the method of Schumacher [50]. All analyzed amphiboles belong to the calcic group with (Ca + Na) B ≥ 1.00 and (Na + K) A < 0.50. Fe/(Fe + Mg) ratios are low and range from 0.16 to 0.17.     Secondary plagioclase replaced magmatic plagioclase and potassium feldspar and is associated with hydrothermal actinolite, titanite, apatite, chlorite, and epidote-allanite in the sodic-calcic alteration. It forms unzoned euhedral to subhedral crystals up to 0.1 mm in length. Locally, it appears with chessboard-like twinning, which is common in sodic-calcic alteration of the porphyry systems, demonstrating that K-feldspar was replaced by sodic plagioclase [37]. Microanalyses showed an enrichment in Na (7.14 to 10.67 wt. % Na 2 O) with a composition of plagioclase ranging between Ab 62 and Ab 91 (Table 6, Figure 8). This composition is very similar with the composition Ab 72 of oligoclase that was reported by Carten [37] in the sodic-calcic metasomatic zone in the Yerington porphyry Cu deposit in Nevada (USA). Replacement of Ca by Na in plagioclase of the Maronia porphyry deposit is shown by the negative correlation of these two elements, which is similar to that reported by Carten [37] for secondary plagioclase in the Yerington deposit. Table 6. Representative microanalyses (wt. %) and calculated formulae of hydrothermal plagioclase in sodic-calcic alteration at Maronia. FeO*: total iron as FeO; bdl: below detection limit. Titanite is the dominant Ti-bearing phase in the Na-Ca-alteration assemblage at Maronia. It replaced magmatic minerals including magnetite, ilmenite, biotite and actinolite (Figure 5b,c). These minerals have already been reported from the monzonite [28]. Titanite forms distinctive elongated wedge-shaped crystals up to 1 mm in length ( Figure 5d) and is associated with actinolite, magnetite, plagioclase and chlorite, and more rarely with epidote-allanite and calcite in the sodic-calcic alteration. Titanite contains 35.54 to 38.18 wt. % TiO 2 and 26.70 to 28.09 wt. % CaO ( Table 7). The Al/Fe ratio varies from 1.8 to 3.8.

Structural Control of Permeability and Fluid Flow
The Maronia porphyry system shares features in terms of age, geological and structural setting, hydrothermal alteration distribution, and alteration and ore mineralogy, to many Oligocene porphyry type deposits of the Biala Reka-Kechros metamorphic core complex in Greece (e.g., Pagoni Rachi, Konos Hill, Koryfes, Myli, Aisymi, King Arthur, St. Philippos and Papadokoryfi; [9,17]) (Figure 1). These deposits are related to magmas that were produced in a subduction-related magmatic arc. The magmas were transported to shallow depths through large detachment displacements at an extensional geotectonic regime that affected the Rhodope massif from the Eocene to Oligocene [8,9,20,27,32].
Two large magmatic intrusions occur in the Maronia area. Monzonite with a shoshonitic affinity of Oligocene age (28)(29)(30) was intruded by the younger microgranite [31,33]. The Cu-Mo ± Re ± Au mineralization is genetically related to the microgranite porphyry rather than the monzonite. The ascent of the magma to shallow levels was promoted by the low angle detachment fault in the Maronia area, which also acted as a conduit for the magmatic metalliferous fluids and volatiles. The presence of plastic and cataclastic deformation features at the southeast margin of Maronia microgranite porphyry implies emplacement at the ductile-brittle transition, adjacent to the footwall of the Maronia detachment fault. This detachment produced ductile to brittle shear zones during emplacement of the granitic magma. Ductile deformation is expressed by the highly strained and mylonitized zones and the brittle deformation by the intense random veining of the system.
Fluids are transported through major conduits such as fractures in brittle crust [51][52][53], but they can also migrate through other high permeability conduits such as shear zones [54]. Cataclastic deformation increases significantly the permeability of the host rocks and thus promotes the fluid flow and evolution of the fluid-wallrock interaction. In contrast, the rocks have a low permeability in ductile conditions [53]. However according to Reynolds and Lister [55] and Weis [53], in the ductile regime, the temporally overpressured fluids result in an increase in the permeability of the rocks. In the case of Maronia, hot fluids migrated through the major shear zone, in a ductile-brittle transition regime and caused metasomatic reactions with the primary magmatic minerals of the host rocks that lead to potassic and sodic-calcic alteration. Similar conditions favored the formation of porphyry and epithermal deposits in the neighboring Biga peninsula of NW Turkey, which belong to the same geotectonic regime [27].
However, the change from ductile to brittle deformation was associated with a significant drop in fluid pressure, fracturation of the rock and permitted the escape of overpressured hydrothermal fluids and a rapid upward fluid flow. This process increases the volume of water, which interacts with the surrounding host rocks along the veins [55][56][57]. At Maronia, the high temperature M-, EB-, A-and B-type veins and veinlets probably formed from an aqueous fluid phase, which percolated upward along the shear zone, shortly after magma was emplaced. This process has been previously described by Dilles [58] and Richards [56].

Hydrothermal Alteration
Hydrothermal alteration assemblages in the Maronia porphyry type deposit reflect fluid-rock interactions and multiple overlapping pulses of hydrothermal fluid flow through magmatic fluid discharge zones. The alteration zoning patterns that include potassic, sodic-calcic, propylitic, sericitic and argillic alterations correspond to the recently updated models proposed by Seedorff et al. [1] and Sillitoe [2]. According to these models, potassic alteration is the most promising ore indicator especially for Cu and Au. Sericite alteration overprints the previous potassic assemblages incorporating Mo and other metals, in addition to Cu and sometimes Au (e.g., Santa Rita in New Mexico [59]; Far Southeast in the Philippines [60]; El Teniente in Chile [61]). This is the reason why detailed mapping of the alteration zones at porphyry systems are in many cases a significant exploration tool, driven by the fact that large deposits are likely to be located at greater depths and costly deep exploration drilling is required.
Hydrothermal alteration in Maronia is genetically associated with the microgranite intrusion but it also extends into the adjacent monzonite. The deposit exhibits a nearly symmetrical pattern of alteration zoning. This symmetry is well expressed in the distribution of sericitic and argillic alterations ( Figure 4) and is probably related to the relatively uniform texture of the microgranite and the lack of any significant tilting postdating the formation of the deposit. However, the distribution of the early potassic and sodic-calcic alterations is less symmetrical due to the apparent tectonic control (Figure 4). In addition, the intimate association of the early potassic with the sodic-calcic alteration, and the late sericitic alteration, indicates that the various alteration styles at Maronia developed as part of an evolving magmatic-hydrothermal system.
The sodic-calcic alteration typically postdates potassic alteration [2,19], but in Maronia, these two types of alterations are possibly contemporaneous, since there is not any clear evidence for the relative age from their distinctive mineral assemblages. Both alteration assemblages overprint each other, and secondary biotite, K-feldspar, magnetite and rutile of the potassic alteration coexist with hydrothermal actinolite, sodic-calcic plagioclase (albite/oligoclase/andesine) and titanite of the sodic-calcic alteration. This implies alkali exchange, with increase of Na, Ca and K and variable changes of Mg and Si as previously described by Seedorff et al. [1].
Potassic alteration formed in monzonite and microgranite, by addition of K, Fe and Mg, loss of Na and Ca and variable changes of Ti. These chemical modifications are reflected by the formation of hydrothermal biotite, orthoclase, magnetite, rutile, quartz, and apatite. In the monzonite, this association principally resulted from the reaction of magmatic hornblende, biotite and pyroxene to secondary biotite and of plagioclase to orthoclase. In microgranite porphyry, the newly formed secondary biotite developed due to the circulation of the high-temperature hydrothermal fluids along the grain boundaries. The main sulfide assemblage at the potassic alteration in Maronia includes pyrite, chalcopyrite and pyrrhotite, which are associated with hydrothermal biotite and magnetite.
The most significant chemical factor that distinguishes the two different types of biotite in Maronia is the decrease of Ti and the increase of Si from magmatic to hydrothermal biotite, especially in the microgranite porphyry (Figures 6 and 7). This is consistent with the results from porphyry Cu deposits elsewhere (e.g., in Santa Rita [44]; in Christmas and Elly [62]; in Bingham [63]; in Yerington [58]; in Koryfes, Pagoni Rachi, and Papadokoryfi [64][65][66]).
Conversely, hydrothermal biotite in the potassic alteration of monzonite at Maronia has a similar MgO content than magmatic biotite, but is slightly lower in hydrothermal biotite from the microgranite (Figure 6b). This is in agreement with Hendry et al. [62] who stated that in some porphyry systems (e.g., Christmas and Bingham) primary and secondary biotites have similar concentrations of magnesium, although most studies conclude that secondary biotite contains more MgO compared to magmatic biotite [45,[67][68][69]. The differences in composition between biotite from the microgranite and the monzonite in Maronia are probably attributed to changes in fluid composition with the mineralogical and geochemical variations of the two host magmatic rocks.
In the potassic alteration, the Fe/(Fe + Mg) ratio of hydrothermal biotite is low, reflecting high f O 2 , above the nickel-nickel oxide (NNO) buffer of the early high-temperature aqueous hydrothermal fluids in Maronia, similar to other porphyry Cu deposits [56,58,64,67,70]. The high-oxidizing conditions of the early aqueous fluids are also supported by the presence of hydrothermal magnetite, which is ubiquitous in the potassic and the sodic-calcic alteration zones of the Maronia system, similarly to other porphyry systems in northeastern Greece (e.g., Pagoni Rachi [17]). At those highly oxidized conditions in the potassic alteration, coarse-grained rutile formed due to the breakdown of pre-existing Ti bearing minerals in the monzonite, implying restricted Ti mobility under sub-solidus conditions [48,71]. Titanium is often regarded as an immobile element, and in porphyry systems it can be incorporated in hydrothermal minerals at high temperatures. This indicates that its solubility increases in aqueous fluids in the form of chlorides, fluorides, hydroxides or other complexes, and thus increasing its mobility [48,72]. This increase of Ti solubility and its transportation by the hydrothermal fluids may explain the presence of rutile in the microgranite. Titanium in the microgranite possibly originated from the dissolution of preexisting magmatic biotite and magnetite. The diversity and rarity of trace elements in rutile is due to the broad range of the composition of the fluids and the precursor minerals involved in the formation of rutile [49,71,73]. Potassic alteration in Maronia is characterized by rutile with elevated concentrations of Sr, W and Nb, and irregular amounts of Pb, V and Y. In addition, titanite (with 35.54-38.18 wt. % Ti) from the sodic-calcic alteration of tine monzonite could also contribute Ti for the formation of rutile during potassic alteration. Primary biotite broke down under the high-temperature potassic alteration conditions, and resulted in the exsolution as needle-like sagenitic rutile in chlorite, a process which has been previously described by Czamanske et al. [71] and Rabbia and Hernández [49] in porphyry type deposits.
The major reaction during this early alteration stage involved the conversion of magmatic biotite or pyroxene to actinolite implying an input of Ca in the hydrothermal fluids. Regardless of whether it was derived from biotite or pyroxene, the composition of secondary actinolite does not vary considerably. Additionally, the high albite component of hydrothermal plagioclase can be attributed to the introduction of Na into the system and the development of albite/oligoclase/andesine at the expense of magmatic calcic plagioclase (anorthite) during sodic metasomatism.
Titanite is a common phase in the sodic-calcic alteration associated with the Maronia deposit, implying that Ti and possibly part, or the entire, of Ca were released from primary magmatic minerals. Calcium can also be added externally, by basinal brines or carbonate rocks, which are penetrated by the ascending fluids. The main precursor to titanite could be primary biotite (with 2.87-6.80 wt. % Ti), magnetite (with 0.20-3.10 wt. % Ti), ilmenite (with 46.67-51.42 wt. % Ti), orthopyroxene (with 0.08-0.43 wt. % Ti), clinopyroxene (with up to 0.88 wt. % Ti), and/or amphibole (with 0.09-1.30 wt. % Ti) from the Maronia monzonite. The low Fe/Al ratio (Fe/Al = 0.3-0.6) in the studied titanite is compatible with the Fe/Al ratio of titanite from sodic-calcic alteration in porphyry Cu-Au deposits in British Columbia and indicated high f O 2 conditions of the fluids [76].

Formation Conditions of Potassic and Potassic/Sodic-Calcic Alteration
Oxidized felsic magmas, comparable with the granitic magma of Maronia, can produce fluids rich in K-Na capable to cause potassic and sodic alteration [56,77]. Although in some studies [38,78] basinal brines or meteoric fluids may have been involved in the genesis of sodic-calcic alteration, a magmatic origin of the fluids for this alteration style is not excluded (e.g., Lightning Creek, Northwest Queensland, Australia [79]). However, sodic-calcic alteration of K-rich magmatic rocks is possible when CO 2 is enriched in the magmatic fluids [79]. Previous fluid inclusion study indicates that the volatile fluids were vapor dominated and coexisted with brines with salinities of 28 to 55 wt. % NaCl equiv. and did not contain any appreciable CO 2 [11]. Therefore, in the Maronia deposit, the fluids possibly were enriched in Ca from an external source, either a basinal brine from outside of the magmatic fluid discharge zone, or a carbonate unit.
The high-oxidation state of the hydrothermal fluids resulted in the enrichment of metals in the aqueous phase, a process which has been previously described by Richards [23] and Sillitoe [3]. Chelle-Michou et al. [80] showed that the high-oxidation state of porphyry systems is associated with the deep crustal magmatic processes that control the behavior of S and the chalcophile metals. This high-oxidation state together with a pressure decrease probably favors the escape of H 2 gas and the subsequent dissociation of H 2 O along with an increase of f O 2 [58,69,81]. Another likely cause for the high-oxidation state is the separation of sulfur from the magma under subsolidus magmatic-hydrothermal conditions and its partitioning into the aqueous fluid phase, a process that results in sulfur reacting with metals in the fluid to form sulfides (pyrite, chalcopyrite, pyrrhotite) and to release O 2 gas. The O 2 released from this reaction causes oxidation and breakdown of the magmatic ferromagnesian silicates as described for the Yerington porphyry Cu deposit by Dilles [58] and for the Bajo de la Alumbrera porphyry Cu-Au deposit [81]. The absence of anhydrite in the potassic alteration zone at Maronia is possibly attributed to low contents of sulfate sulfur species in the fluids. According to [1], sulfates, such as anhydrite, precipitate from oxidized sulfur, rather than reduced sulfur, which forms sulfide minerals.
Copper in the potassic alteration partitions into brine rather than the vapor phase at early stages of hydrothermal systems [81][82][83]. The presence of brine in Maronia is apparent from the fluid inclusion assemblage [11], and played a significant role in the precipitation of the A-type pyrite-magnetite-chalcopyrite-pyrrhotite quartz veins at temperatures between 320 and 450 • C (with a distinct peak at 400 • C) under boiling conditions. According to [81,[84][85][86] Cu sulfides precipitate in potassic alteration zones under oxidizing conditions from brines, together with Fe minerals (e.g., magnetite). Copper and iron strongly tend to complex with Cl and partition into brine. The common chalcopyrite daughter crystals in the high salinity aqueous fluid inclusions of the veins at the potassic alteration in Maronia support this assumption.

Conclusions
The Maronia porphyry Cu-Mo ± Re ± Au mineralization is hosted in microgranite porphyry, which intruded the Oligocene monzonite, at the footwall of a detachment fault. This fault formed as a result of the extensional geotectonic regime in the Rhodope massif, which produced magmas in a subduction-related magmatic arc. The detachment and associated supra-detachment faults also acted as conduits of hydrothermal fluids and volatiles along ductile to brittle zones, which is expressed by intense shear and veining of the host rocks. Hot fluids migrated through the ductile to brittle shear zone at the south-eastern margin of the system and caused metasomatic reactions with the primary minerals of the microgranite porphyry and the adjacent monzonite. These reactions lead to potassic and sodic-calcic alteration. Brittle deformation also promoted the rapid upward flow of fluid, which interacted with the surrounding host rocks, along high permeability conduits, to form high temperature M-, EB-, A-and B-type veins. Magma that formed the Maronia microgranite had oxidized felsic affinity and produced fluids enriched in Na, Ca and K that are necessary to form the potassic and sodic-calcic alteration. The high-oxidation state of the hydrothermal fluids and the decrease in pressure resulted in the enrichment of metals and sulfur in the aqueous phase. This process caused precipitation of sulfides (pyrite, chalcopyrite, pyrrhotite), O 2 -gas release, and breakdown of the magmatic silicates to form the potassic and sodic-calcic alteration assemblages.