Extremely Re-Rich Molybdenite from Porphyry Cu-Mo-Au Prospects in Northeastern Greece: Mode of Occurrence, Causes of Enrichment, and Implications for Gold Exploration

Extremely Re-rich molybdenite occurs with pyrite in sodic–calcic, sodic–sericitic and sericitic-altered porphyritic stocks of granodioritic–tonalitic and granitic composition in the Sapes–Kirki–Esymi, Melitena and Maronia areas, northeastern Greece. Molybdenite in the Pagoni Rachi and Sapes deposits is spatially associated with rheniite, as well as with intermediate (Mo,Re)S2 and (Re,Mo)S2 phases, with up to 46 wt % Re. Nanodomains and/or microinclusions of rheniite may produce the observed Re enrichment in the intermediate molybdenite–rheniite phases. The extreme Re content in molybdenite and the unique presence of rheniite in porphyry-type mineralization, combined with preliminary geochemical data (Cu/Mo ratio, Au grades) may indicate that these deposits have affinities with Cu–Au deposits, and should be considered potential targets for gold mineralization in the porphyry environment. In the post-subduction tectonic regime of northern Greece, the extreme Re and Te enrichments in the magmatic-hydrothermal systems over a large areal extent are attributed to an anomalous source (e.g., chemical inhomogenities in the mantle-wedge triggered magmatism), although local scale processes cannot be underestimated.


Introduction
Rhenium has an average concentration of <1 ppb in the Earth's crust, and there are few places where it is enriched to ore grades.It is among the most expensive metals and is used principally as a high-temperature super-alloy in jet engines.It is mainly obtained as a by-product of molybdenum and copper refinement from Cu-Mo porphyry and sediment-hosted deposits.Re-bearing molybdenite with >1 wt % Re is rare, but has been found in porphyry-style Cu-Mo deposits [1][2][3].Several Re-Mo-Cu sulfides, mostly of scientific interest, are described from magmatic Cu-Ni deposits in mafic-ultramafic complexes in Sweden [4], Canada [5], Montana [6], Karelia [7], and Finland [8].Recently, native rhenium was identified in Precambrian chromite ores hosted in serpentinized dunites-harzburgites of the Ukrainian shield [9].Dzhezkazganite (ReMoCu 2 PbS 6 ) is the main Re-bearing phase in chalcocite ores from the sediment-hosted Dzhezkazgan deposit, Kazakhstan [10].An Os-Cu rheniite [(Re,Cu,Os,Fe)S 2 ] occurs in bornite-rich ore in sandstone at Voronov Bor, Karelia, Russia [11], whereas rheniite (ReS 2 ) was initially discovered in fumarolic sublimates from Kudryavy volcano, Kurile Islands, Russia, which is its type locality [12].
Rheniite was also discovered in the Pagoni Rachi and Konos porphyry prospects, northeastern Greece [13,14], which are the first known occurrences of this mineral in porphyry-style mineralization.The Pagoni Rachi and Konos prospects are part of a cluster of porphyry-type occurrences in northeastern Greece (the others include Maronia, Myli and Melitena), characterized by molybdenite with the highest Re contents yet reported from a porphyry-type deposit (up to 4.7 wt % Re) [2,3,[13][14][15][16][17][18][19].The study of [13] indicated that the Re-rich molybdenite from Pagoni Rachi crystallized as the 2H polytype and not the 3R polytype, as previously hypothesized [1,20], which suggests that Re concentration does not correlate with a specific polytype.In addition, the almost perfect linear correlation between the Mo and Re content of molybdenite from porphyry-type prospects in northeastern Greece supports the previously proposed concept [13,[21][22][23] that Re substitutes for Mo in the structure of molybdenite.
The Re content of molybdenite is controlled by several factors, including the composition of the ore-forming solution and the host rock, the nature and source of the host rock, the total amount of molybdenite in a given deposit, and the physicochemical conditions of ore formation [19,23].Rhenium grades for porphyry Cu-Au and porphyry Mo deposits are comparable (in the range of 0.01 to 0.6 g/t), suggesting that there is no difference between the Re content of Mo-dominant and Cu-dominant porphyry deposits of comparable tonnage [3].The recently discovered Merlin Mo-Re zone, Australia, an unusual vein-replacement deposit, probably genetically related to the Mt.Dore IOCG deposit, is the world's highest grade molybdenum-rhenium deposit, the current grades averaging 0.6 wt % Mo and 10 g/t Re, which is much higher than the majority of porphyry Mo deposits where molybdenum grades are of the order of 0.1 to 0.2 wt % Mo [24,25].
This study evaluates earlier data on Re-enriched molybdenites from porphyry-type deposits in northeastern Greece, presents new information concerning recent molybdenite discoveries, and highlights the reasons for Re-enrichment.On the deposit scale, associated ore mineralogy and geochemistry is tested as a tool for Re and Au-enrichment and the use of these elements in the exploration of Re-bearing porphyry deposits in Greece.

Methods
Seventy-eight thin and polished thin sections of host rocks and sulfide assemblages were studied with an optical microscope and a JEOL JSM 5600 (Tokyo, Japan) scanning electron microscope equipped with back-scattered imaging capabilities, at the Department of Mineralogy and Petrology, University of Athens, Greece.The composition of molybdenite was determined with a Cameca SX 100 (Gennevilliers, France) wavelength-dispersive electron microprobe at the Department of Mineralogy and Petrology, University of Hamburg, Germany.Operating conditions were: 20 kV and 20 nA, counting time 20 to 60 s, beam diameter <2 μm.The following X-ray lines were used: MoLβ, ReMα, CuKα, FeKα and SKα.Rhenium metal (for Re), pyrite and chalcopyrite (for Fe and S), and molybdenite (for Mo), were used as standards.The chemical composition of 87 molybdenite and rheniite grains was also determined using a JEOL JXA 8200 electron microprobe at Iowa State University, Ames, IA, USA.The concentrations of the major and minor elements were determined at an accelerating voltage of 15 kV and a beam current of 15 nA, with 10 s as the counting time (5 s on each background), beam diameter <2 μm.The standards employed were metallic rhenium and molybdenum for Re and Mo, chalcopyrite for Cu, and synthetic pyrite for S. Representative samples from unaltered rocks hosting porphyry-type mineralization were analyzed for major and trace elements by X-ray fluorescence techniques with a Phillips PW 1220 (Amsterdam, The Netherlands) instrument in the Institute of Mineralogy and Petrology, University of Hamburg.Bulk ore analyses of 44 mineralized samples were performed at ACME Labs, Vancouver, Canada using aqua regia digestion and Ultratrace ICP-MS analysis.

Regional Geological and Metallogenetic Setting
The Hellenides, part of the Alpine-Himalayan orogeny, were formed from the Late Jurassic to the present above the Hellenic subduction by the off-scraping of crustal units from the Pelagonian continent in the north and then the Pindos Ocean; the Apulian platform and eastern Mediterranean Ocean further south were subducted below Eurasia after the closure of the Vardar Ocean in the Late Cretaceous period [26][27][28].Since Early Cretaceous, a single subduction zone without slab detachment was active and consumed both oceanic and continental lithospheric mantles beneath the Aegean Sea [26,27,29,30].
In the Rhodope Massif (Bulgaria and Greece), Late Cretaceous-Tertiary exhumation of deep metamorphic rocks along detachment faults resulted in the formation of metamorphic core complexes, supra-detachment sedimentary basins, and widespread Late Eocene to Early Miocene basic to felsic magmatism [31] (Figure 2).Early cooling/exhumation (syn-orogenic exhumation) in the northern and eastern part of the Rhodope Massif was initiated in the Maastrichtian to Early Paleocene (up to ~40 Ma) [27,32].Jolivet et al. [27] and Wüthrich [32] distinguished two following stages of widespread extension and metamorphic core complex formation in the Rhodope Massif: a period of peak-cooling/exhumation from ~40 Ma to Late Eocene-Early Oligocene  without migration of volcanic centers (suggesting that the subduction zone was stationary), followed by a second period (Oligocene-Miocene) of fast back-arc extension that is considered to be the result of slab retreat (post-orogenic slab-rollback driven extension).The two extensional phases may indicate two successive "push-pull" cycles [33], which were caused by the accretion/subduction of continental and oceanic lithosphere fragments within the Aegean subduction zone [32].The first phase of core complex formation in the Rhodope Massif (peak-cooling/exhumation between 40 and 33 Ma) is related to the first "push-pull" cycle (subduction of the Pindos Ocean following accretion of the Pelagonian continental block to the Moesian platform and previous accreted units, e.g., Drama continent, Vardar Ocean), and to voluminous bimodal magmatism and volcanism in the Rhodope Massif [32,34].The Late Eocene-Oligocene (e.g., 35-25 Ma) magmatism in the Rhodope Massif was caused by convective removal of the lithospheric mantle (lithospheric delamination) and subsequent upwelling of the asthenosphere [31].The upwelling asthenosphere provided the heat for hydrous melting of a heterogeneously enriched subcontinental mantle, resulted in the weakening of the crust, and may have produced the accelerated generation of the metamorphic core complexes in the Rhodope Massif [32,34,35].The plutonic rocks (mostly granite to monzonite), and calc-alkaline to shoshonitic and ultra-K volcanic associations have geochemical and isotope affinities with supra-subduction rocks [31,36,37].The second phase of core complex related extension in the Rhodope Massif was coeval with the accretion and subduction of a continental block composed of several microcontinents (Gavrovo-Tripolitsa, Olympos, Ionian, Apulian), and the subsequent subduction of the Mediterranean Sea [32].Slab rollback (since about 35-30 Ma ago) is consistent with the progressive southward migration of magmatic activity from the Tertiary in the Rhodope Massif to the present in the South Aegean volcanic arc, and for the development of the Aegean subduction zone [26,27,32,36,38].The Palaeogene magmatism of the central and eastern Rhodope Massif culminated at about 30-35 Ma and was accompanied by abundant ore-forming hydrothermal activity [31,40].Several styles of mineralization in northern Greece (base-and precious-metal porphyry, reduced intrusion-related systems, epithermal, Carlin-type) are genetically related to arc-related magmatic rocks, and are, in part, controlled by exhumation structures of the Rhodope metamorphic core complex [31,39,[41][42][43][44][45].Porphyry Cu-Mo-Au systems that contain Re-rich molybdenite in northeastern Greece are genetically related to microgranite (Maronia) and granodiorite porphyries (Konos, Pagoni Rachi, Myli) [39,42,[46][47][48] (Figure 2).

Mineralization and Alteration
The alteration assemblages and vein types recognized on the basis of surface relationships are distinguished according to the terminology of Gustafson and Hunt [55], Dilles and Einaudi [56], Muntean and Einaudi [57] for A, B, C and D veins and summarized by Seedorff et al. [58].According to the above classification schemes, A-type veins formed early and consist mostly of quartz, K-feldspar, and sulfide (± anhydrite) that have well-developed K-feldspar alteration generally within the porphyry stock.B-type veins formed after A-type veins along the margins of the stock and into the country rock.They are composed of quartz and sulfides including molybdenite (± anhydrite) and generally lack an alteration halo.C-type veins contain quartz ± chlorite ± biotite ± epidote ± chalcopyrite ± bornite ± pyrite ± molybdenite, while D-type veins usually form very late in the paragenetic sequence of vein formation and are structurally controlled.The sulfide-to-quartz ratio is generally higher than in A-and B-type veins, with banded anhydrite-sulfide textures being common.Sericite replaced K-feldspar and biotite in alteration haloes surrounding veins.Veins dominated by magnetite, actinolite and plagioclase are spatially associated with sodic or calcic alteration in some tonalitic-granodioritic porphyry Cu-(Au-Mo) deposits, such as Park Premier, Utah and Island Copper, British Columbia and were termed M veins by [59].However, magnetite veins can also be related to potassic alteration and range from magnetite-rich veins with biotite and minor quartz, chalcopyrite and K-feldspar envelopes, to sugary quartz veinlets with magnetite and chalcopyrite and no alteration envelopes.Such veinlets were termed A veinlets by [57,58].
The Myli Cu-Mo prospect (Esymi) is characterized by molybdenite-pyrite-bearing quartz stockworks (surface exposure of about 0.5 km × 1 km) that are similar to those spatially associated with Pagoni Rachi and Konos and to sericite-and sericite-carbonate alteration of the host granodiorite-tonalite (Figure 4e).Molybdenite and pyrite deposition formed close to the arsenopyrite/tennantite boundary (Figure 5g-i).Native gold associated with chalcopyrite is enclosed in pyrite (Figure 5j).This is, to our knowledge, only the second occurrence of native gold in porphyry-style stockworks in northeastern Greece other than that in sodic-potassic-and phyllic-altered rocks at Pagoni Rachi.Sericite, calcite-dolomite, and late kaolinite are gangue minerals within the veins (Figure 5g,h).Grades reach values up to 2100 g/t Cu, up to 210 g/t Mo (average Cu/Mo = ~35), and up to 0.2 g/t Au ( [60], Table 1).

Maronia
The Maronia porphyry Cu-Mo-Au deposit occurs in quartz stockworks (surface exposure of about 0.5 km × 0.5 km) crosscutting sericitic-altered microgranite (Figure 4f,g).Molybdenite is associated with pyrite, chalcopyrite, base metal sulfides, and sulfosalts [39].According to [39], bulk ore analyses from porphyry-style ore have an average grade of 0.1 g/t Au, and up to 2763 g/t Cu and 3909 g/t Mo, with an average Cu/Mo ratio of about 15 (Table 1).

Melitena
At the Melitena deposit, molybdenite forms disseminations and fissure fillings, and is associated with quartz veinlets crosscutting transitional sericitic-advanced argillic alteration zone of the granodiorite porphyry [54] (Figure 4h,i).Pyrite and molybdenite are the main metallic minerals (Figure 5k,l), whereas galena, chalcopyrite and pyrrhotite occur in minor amounts in the veins and as disseminations in the wallrocks.Bulk rock chemical analyses [17] reveal a high base and precious metal content (~400 g/t Cu, ~6000 g/t Mo (average Cu/Mo = 0.2), and up to 0.3 g/t Au (average = 0.16 g/t) (Table 1).

Mineralogy of Rhenium-Bearing Phases
The Re content of molybdenite from prospects and deposits in northeastern Greece are presented in Tables 1 and 2 and plotted in Figure 6.Table 1 also includes data from other Greek porphyry systems (e.g., Skouries, Fakos, Sardes, Stypsi) and from molybdenite in porphyry Cu, Cu-Au, Cu-Mo and Mo deposits elsewhere.The almost perfect linear correlation between the Mo and Re content of molybdenite (Figure 6b) indicates that Re substitutes for Mo in the structure of molybdenite as previously suggested [13,[21][22][23].This is also evident from the homogeneous Re distribution in molybdenite (Figure 7a-c).As shown in Figure 5d,i, the Re content of molybdenite can vary greatly within the same crystal.Table 2. Representative electron microprobe analyses and atomic proportions of molybdenite from Konos (1,2), Myli (3,4), Melitena (5,6), Maronia (7,8) and Pagoni Rachi (9,10); bd: below detection.Experimental evidence presented by Drábek et al. [78], suggests that a maximum of 2.7 wt % Re can be dissolved in molybdenite at temperatures up to 1000 °C (and 2.2 wt % at 400 °C).Drábek et al. [78] cautioned that researchers studying molybdenite-bearing mineral assemblages containing anomalously high Re contents should "watch out for a possible excess Re phase."However, our structural analysis of four molybdenite crystals with the highest Re content ever reported in nature (up to 4.7 wt % Re) showed that Re actually enters the molybdenite structure as isovalent substituent of Mo.This observation is corroborated by the progressive shortening of the Mo-S mean bond distance, as well as by the uniform decrease of the unit-cell values with the increase of Re [13].We conclude that the heterogeneity with up to 4.7 wt % Re in molybdenite is the result of solid solution, thus expanding the results of [78].
Pure rheniite and Mo-rich rheniite with up to 6 wt % Mo occurs at Pagoni Rachi and Konos porphyry systems ( [13,14] and this study).It is believed that the Mo content of Mo-rich rheniite, may be derived from associated molybdenite.This is consistent with the experimental work of [78] that indicates a negligible amount of molybdenum in ReS 2 at 400 °C.Intermediate (Mo,Re)S 2 and (Re,Mo)S 2 phases with 10.5-38.5 wt % Re at the Pagoni Rachi porphyry system [13], as well as intermediate phases between molybdenite and rheniite with 37.2-46.1 wt % Re at the Konos porphyry prospect, may be due to the incorporation of rheniite in molybdenite and vice versa, as suggested by [78].
It was proposed by several workers that Re must have redistributed or decoupled within molybdenite after its formation [23,[79][80][81][82].These processes may explain the strong variation in the rhenium concentration within single grains.Rheniite, which rims molybdenite in the Pagoni Rachi and Konos porphyry deposits, resembles low temperature exsolution product of high-Re molybdenite.However, incorporation of the rheniite grains in quartz (Figure 5b,c) instead suggests that may have formed at the same time as molybdenite.
Among the second group of porphyries with a lower (but still high) content of Re in molybdenite (e.g., Skouries, Myli, Fakos, Sardes, Stypsi), the Myli prospect is hosted in granodiorite-tonalite porphyry, similar to that observed at Pagoni Rachi and Konos.The Re content of Myli molybdenite is lower (average = 2733 g/t).It should be noted here that other porphyry Mo and Mo-W systems in Greece are characterized by Re-free or very low-Re molybdenites (Re-content from 10 to 1310 g/t, average = 206 g/t) (e.g., the intrusion-related systems of Lavrion, Serifos, Axioulolis and Kimmeria) (Table 1).

Causes of Re-Enrichment in Molybdenite from Northern Greece
In a previous discussion on the Re content of molybdenite from Greece, Voudouris et al. [14] proposed that Re-poor molybdenite is spatially associated mostly with porphyry Mo and Mo-W systems (e.g., Kimmeria, Lavrion, Serifos and Axioupolis deposits).This proposal is consistent with the findings of [3,23,83,84] who suggested that the metal source of the deposit and the degree of magma fractionation are probable key factors that contribute to the Re content of molybdenite in granitoid-hosted deposits, and that the Re content of molybdenite decreases gradually from mantle sources, to mixtures between mantle and crust, and is the lowest in crustal rocks.It was suggested previously that primary magmatic compositional controls, such as those that influence ore element ratios including Cu:Mo may also be an important control on the Re content of molybdenite [84].According to Blevin [84], the atomic Re:Mo ratios of molybdenite and the degree of fractionation of the related granites have a strongly negative correlation.Molybdenite with low Re:Mo ratios are associated with Mo-dominant mineralizing systems, which are genetically related to more evolved granites, whereas the highest atomic Re:Mo ratio occurs in molybdenite in Cu-Au deposits.Low ppb Re-bearing molybdenite is quite common in reduced intrusion-related systems like the Mactung (Canada) and Pogo (Alaska) deposits [85,86].
All porphyry Mo and Mo-W systems hosting Re-poor molybdenite in Greece are either of crustal origin, or they originated from the mantle but with a major contribution of these metals from the crust.The molybdenite-bearing Kimmeria granodiorite was considered to have formed by dehydration melting of the lower crust without any contribution from mantle components [87,88], whereas the Lavrion granodiorite is regarded to have a mantle origin with a contribution of a crustal felsic component [89,90].The source materials for the leucocratic, S-type Axioupolis and Serifos granites are considered to also have a crustal anatectic origin [91,92].The low Re-content of molybdenite in the Mo-W-enriched Lavrion, Kimmeria, Axioulopis and Serifos systems is consistent with the reducing nature of mineralizing fluids as proposed by the experimental work of [93], who suggested that reducing fluids containing sulfur have a lower capacity for transporting rhenium.Such fluids are, therefore, not favorable for the formation of rhenium-enriched sulfide deposits.To form rhenium deposits, oxidizing fluids must be operative at some stage(s) of ore formation.
Geochemical and stable isotope data suggest that the Oligocene to Miocene plutonic rocks in northern Greece and associated porphyry mineralization with ultrahigh-and high-Re molybdenite, are derived from sub-continental lithospheric mantle and/or the lower crust, with a minimal contribution from the crust [88,94].More precisely: (a) a LILE-and LREE-enriched sub-continental mantle is the proposed magma source for the Pagoni Rachi and Konos granodiorite-tonalite [95]; (b) a lower crustal origin for the Maronia microgranite [49]; (c) a mantle source for the high-K calc-alkaline granodiorite porphyry at Stypsi [94]; (d) dehydration melting of enriched metabasaltic amphibolite at the base of crust, triggered by rising mantle melts from both the asthenosphere and the enriched subcontinental lithosphere for igneous rocks related to porphyry-style mineralization on Limnos island [96]); and (e) a magma genesis in the upper mantle and subsequent crustal contamination of the parental melts during emplacement of the Skouries deposit [97].
The Re-and/or Te-enriched porphyry Cu-Mo-Au deposits found in northeastern Greece (Pagoni Rachi, Konos, Maronia, Melitena) formed in a post-subduction, post-collisional environment, where partial melting of previously metasomatized mantle lithosphere triggered by local ascent of asthenospheric mantle could be the key to metal enrichment [13,43].In post-subduction tectonic settings, similarly to that in northern Greece, arc-like magmas and related porphyry and epithermal ore deposits formed by remelting of previously subduction-modified upper plate lithosphere, and in particular the lower crustal amphibolitic cumulate roots of former arc magmatic complexes [98,99].In the parental magmas of the Pagoni Rachi and Konos intrusives, Re and other metals (e.g., Mo, Cu, Fe) may be derived from the breakdown of amphibole in amphibolites as suggested by [100] for the recently discovered Longmendian Mo deposit, China.A melting of mafic or ultramafic rocks, as well as involvement of mantle underplating or metasomatism in the source rocks, could also contribute to Re-enrichment in molybdenite as proposed elsewhere [23].
Arc magmatism may also be responsible for the enrichment in Re and possibly of other fluid-mobile elements such as Te, Pb, As, Sb, Cu, Au, and the Pd-group elements in the mantle wedge, probably due to fluids released from the subducted slab and associated sediments [99,[101][102][103].The enrichment of Au and Te occurs in several porphyry-and epithermal-style deposits associated with post-subduction magmatism in the Neo-Tethyan belt, including the Rosia Montana (Romania), Larga (Metaliferi Mountains) and Sacarimb, Golden Quadrilateral (Romania) [104][105][106][107], Skouries (Greece) [108], the Grasberg, Ok Tedi, Porgera, Lihir, and Emperor porphyry Cu-Au and epithermal Au deposits in the southwest Pacific, as well as the Golden Sunlight Mo-Au (Montana), and Mount Milligan Cu-Au porphyry deposits, British Columbia [99,109,110].This gold-enrichment may reflect remelting of small amounts of residual sulfide left in the deep lithosphere by arc magmatism [98,99].According to [107], exceptional endowment in Te-Au epithermal deposits is common to provinces associated with extensional magmatism preceded by much older mantle metasomatism.It is suggested here that a strongly metasomatized mantle wedge which triggered magmatism in northeastern Greece, was responsible for the Re-and Te-enrichment in the Greek magmatic-hydrothermal systems.Similar scenarios, where a melt-metal source in fertile mantle and/or juvenile lower crust was responsible for the relatively high Re concentrations (up to 0.35 wt %) in molybdenite from the Majdanpek (Serbia) and Elatsite (Bulgaria) Cu-Mo porphyry deposits, is supported by [111].Rhenium is directly derived from partial melting of metasomatized mantle whereby Re is released during dehydration of the subducted Garzê-Litang oceanic crust to form molybdenite with up to 2434 ppm Re at the Xuejiping deposit, China [112].A release of Re from the asthenospheric mantle wedge by subduction-related fluids and melts similarly to that associated with the formation of Kudryavy volcano, Kurile islands [103] could also explain the enrichment of Re in magmatic systems in northern Greece.
Local physicochemical conditions (e.g., fO 2 , activity of Cl, pressure, and temperature) affect the transportation, deposition and zoning of metals (including Re) in magmatic-hydrothermal systems [19] and likely control the Re content of molybdenite.However, the Re-enriched mineralization in northeastern Greece expands over an area of about 5400 km 2 and this suggests that large-scale processes (e.g., anomalous concentration of Re in source areas is necessary to account for the Re enrichment) in the studied occurrences.

Implications for Re-Au Exploration
According to [3], molybdenite from porphyry Cu-Au deposits in Canada have higher Re contents (up to 3858 ppm Re for Kemess South and 8170 ppm Re for Mitchell) than molybdenite from Canadian porphyry Cu deposits (up to 1863 ppm Re for Island Copper).In non-Canadian porphyry Cu deposits, the Re content of molybdenite can be more than 4000 ppm at Kounrad, Kazakhstan and Morenci, USA and up to 15,400 ppm at the Santo Thomas II porphyry Cu-Au deposit, Philippines [3,76]).The Re content of molybdenite in Canadian porphyry Cu-Mo deposits is generally lower than that in molybdenite from porphyry Cu and Cu-Au deposits (e.g., up to 643 ppm Re from the Maggie deposit; [3]).In non-Canadian Cu-Mo deposits the Re content of molybdenite may reach values up to ~2000 ppm (e.g., Bingham, USA), whereas the Re content of molybdenite in porphyry Mo deposits is characterized by being much lower (up to 397 ppm, e.g., Endako) [3].The Merlin Mo-Re deposit, Australia, is by far the highest grade Re deposit in the world with indicated mineral resources of 6.7 Mt at 1.4% Mo and 23 g/t Re [24,25].However, despite being the highest grade Re deposit in the world, the Mo/Re ratio is about 725 [25] with an average Re content in molybdenite of about 1000 ppm.In general, the plot of Re content of molybdenite versus Cu/Mo ratio and Au grade in porphyry-type deposits (Figure 8a,b) shows a correlation in which the Re content of the molybdenite increases as the Cu/Mo ratio decreases (e.g., in porphyry Cu-Au and Cu deposits [3]).These authors also demonstrated that platinum-group metals, and gold, are concentrated in porphyry deposits along with Re, and that high contents of Re in molybdenite can be an indicator of high levels of platinum-group metals.
Preliminary data on Au grades and Cu/Mo ratios versus the Re content of molybdenite in the porphyry deposits in northeastern Greece (Figure 8a,b) suggest that molybdenite in these occurrences have characteristically high Re contents.The Pagoni Rachi prospect is highly anomalous in Au and ongoing exploration is needed to verify if the rest of the prospects are also enriched in Au and Re, since high Re content of molybdenite correlates with high Au grades in these deposit.However, since these occurrences are not associated with any measured resources, it is premature to consider them as a separate class or subclass of porphyry deposits.The Re-in-molybdenite values from the Skouries deposit (the only porphyry deposit in Greece with a significant resource) average 900 ppm Re, which is low compared to other porphyry Cu-Au deposits worldwide, but this maybe a function of the limited number of molybdenite analyses (i.e., four) obtained here.Our data suggest enrichment of Re in molybdenite throughout the studied deposits and at various levels of exposure in the system ( [13] and this study): Although these deposits are characterized by a strong telescoping (where sericitic alteration overprints earlier sodic-potassic alteration and A, B-type quartz veinlets are reopened by D-type veinlets), molybdenite remains Re-enriched in all mineralized stages, as well as in both the core and distal parts of the systems.Based on available data, there is no indication that the Re values in molybdenite vary with depth and/or laterally in these occurrences, as appears to be the case at, for example, Bingham Canyon by [75], where highest values are present in the upper part of the deposit (up to 2000 ppm Re) and decreased with depth (to less than 200 ppm).
The lower Re content of molybdenite from the Myli deposit may be attributed to the slightly reduced nature of the ore-bearing fluids (presence of arsenopyrite together with molybdenite) as opposed to highly oxidizing conditions at Pagoni Rachi and Konos systems.The reduced nature of the fluid may be due to the intrusion of granodiorite-tonalite porphyry in reduced sediments at Myli.This is also the case for the two porphyry systems on Limnos island (Fakos and Sardes) [14,61,62].Stein et al. [23] and Gilles and Schilling [75] related the Mo contents to mass balance considerations in which the Re content of molybdenite in Cu-Mo deposits was higher than for molybdenite in porphyry Mo deposits because of the lower amount of molybdenite in the former deposits.However, our study indicates that there is no clear relationship between Re content and molybdenite abundance, since molybdenite at the Maronia and Melitena porphyry Mo-Cu deposits is a very abundant mineral that is extremely enriched in Re.
The diaspore + pyrophyllite + aluminum-phosphate-sulfate mineral-bearing root zones of high-temperature advanced argillic lithocaps in northern Greece may be potential exploration targets for porphyry type Cu-Mo-Re-Au mineralization [54].In the Greek Rhodope region, almost all high-sulfidation epithermal systems and associated advanced argillic alteration zones are telescoped onto porphyry-style Cu-Mo-Re-Au mineralization (Figure 9).Careful mapping of advanced argillic alteration assemblages in the Rhodope Massif has the potential to identify new target areas for porphyry-style mineralization.Ongoing exploration and geological, mineralogical and geochemical studies will determine whether or not northeastern Greece represents a Re-Au-Te metallogenetic province.

Conclusions
Molybdenite in Oligocene porphyry systems in northeastern Greece contain extremely high Re content (several tens of thousands of ppm) which is associated with pure rheniite, Mo-rich rheniite, and intermediate (Mo,Re)S 2 and (Re,Mo)S 2 phases with up to 46.1 wt % Re, the latter probably being the result of micro-inclusions of rheniite in molybdenite and vice versa.The molybdenite-rheniite mineralization is hosted by sodic-potassic, sodic-sericitic-and sericitic-altered granodiorite-tonalite and microgranite porphyries and is spatially associated with native gold, Fe-Cu sulfides and oxides.Although local factors at deposit scale may have contributed to Re content in molybdenite, we favor large scale processes, based on the large areal distribution of Re-enrichment in magmatic-hydrothermal systems in northeastern Greece.Strong chemical inhomogenities in the mantle wedge which triggered post-subduction magmatism in northeastern Greece during the Tertiary, resulted in the Re-(and Te, Au) enrichment in these porphyry systems.The widespread presence of precious metal tellurides in the ores, and first detection of Pd in Pagoni Rachi, further suggests that these systems could be exploration targets for porphyry-style Au-Re-Pd-Te ores.

Figure 7 .
Figure 7. Back-scattered electron images showing the distribution of Mo, S and Re in molybdenite grains from Pagoni Rachi.

Figure 8 .
Figure 8.(a) Re content of molybdenite vs. Cu/Mo ratio of porphyry deposits; (b) Re content of molybdenite vs. gold grades of porphyry deposits (modified after [3]).Re content of molybdenite, Cu/Mo ratio and Au grades of studied prospects are also shown.

Figure 9 .
Figure 9. Schematic diagram integrates magma emplacement, degassing and porphyry-epithermal mineralizing processes in northeastern Greece.Four successive magmatic pulses (granodiorite-tonalite porphyry, monzodiorite-monzogabbro, microdiorite and microgranite porphyry) within metamorphic basement and coeval volcanic rocks resulted in superimposed porphyry-epithermal style of mineralization and upgrading of earlier formed mineralization.Magmatic vapors resulted in the formation of high-sulfidation Cu-Au-Te mineralization within advanced argillic alteration lithocaps and fluid inclusion evidence suggests that rheniite, Re-rich molybdenite, chalcopyrite and native gold at Pagoni Rachi and Konos prospects may have precipitated as temperatures fell below 400 °C during phase separation at relatively oxidized conditions, high chlorine activity, and from relatively acid hydrothermal solutions.

Table 1 .
Rhenium content of molybdenite from porphyry-type deposits in Greece and elsewhere.Cu/Mo ratios and Au contents are also included.