Schweizerische Mineralogische und Petrographische Mitteilungen 84 59ndash78 2004

0036-769904008459 copy2004 Schweiz Mineral Petrogr Ges

The Ada Tepe deposit a sediment-hosted detachment fault-controlled low-sulfidation gold deposit in the

Eastern Rhodopes SE Bulgaria

Peter Marchev1 Brad S Singer 2 Danko Jelev 3 Sean Hasson 3 Robert Moritz4 and Nikolay Bonev5

Abstract

The Ada Tepe gold deposit 230 km SE of Sofia formed in the eastern part of the Rhodope Mountains that under-went extension and metamorphic core complex formation followed by normal faulting basin subsidence and silicicto mafic magmatism during the MaastrichtianndashOligocene The region comprises numerous volcanic-hosted epi-thermal and base-metal vein deposits spatially and temporally associated with the Oligocene magmatism Ada Tepeis a typical low-sulfidation epithermal gold deposit unusual in that it is older than adjacent magmatic-related depos-its and is hosted in MaastrichtianndashPaleocene sedimentary rocks above a detachment fault contact with underlyingPaleozoic metamorphic rocksGold mineralization is located in (1) a massive tabular ore body above the detachment fault and (2) open space-filling ores along predominantly eastndashwest oriented listric faults The ores are zones of intensive silicification andbrecciation synchronous with detachment faulting Brittle deformation opened spaces in which bands of opalinesilica and electrum quartz pyrite massive and bladed carbonates were deposited Mineralization is exclusively a Ausystem with AuAg ~3 trace As and no base metals Alteration consists of quartz adularia chlorite sericite calcitepyrite and clay minerals Adularia and abundant bladed carbonates indicate boiling within the entire span of thedeposit whereas bands of opaline silica with dendritic gold suggest that silica and gold were transported as colloidsThe physical setting of formation of the Ada Tepe deposit was very shallow and low temperature The Sr and Pbisotope ratios of carbonates and pyrite reflect hydrothermal fluid signatures derived predominantly from the meta-morphic rocksThe age of mineralization and association with the detachment fault suggest that gold mineralization at Ada Tepe ismore closely linked to the Kessebir metamorphic core complex rather than to local magmatism

Keywords detachment fault sediment-hosted low-sulfidation epithermal Eastern Rhodopes Bulgaria

1 Introduction

The Rhodope metallogenic province (Stoyanov1979) comprises numerous volcanic-hosted epi-thermal (Madjarovo Chala Zvezdel Losen Per-ama Hill Sappes) and base-metal vein (MadanLuki Popsko Thermes) deposits hosted by meta-morphic basement (Marchev et al 2000) Thespatial and temporal association of these depositswith Oligocene magmatic centers is documentedby the radioisotopic dating of Singer and Marchev(2000) Marchev and Singer (2002) Rohrmeier etal (2002) and Ovtcharova et al (2003)

The Ada Tepe deposit is one of a newly discov-ered group of gold deposits in the Eastern Rho-

dopes These share many features with the adu-laria class of epithermal deposits but they areatypical in that they are hosted mostly in sedi-mentary rocks and do not show direct relation-ships to local magmatic activity Three of them in-cluding Ada Tepe Rozino and Stremtsi (Fig 1)are major targets for commercial exploration Ex-ploration carried out in the last 2 years shows thatAda Tepe is the major potential economic depositin the region

The Ada Tepe low-sulfidation epithermal golddeposit is located 4 km SE of Krumovgrad in theEastern Rhodopes 230 km SE of Sofia (Fig 1)Balkan Minerals and Mining AD (BMM ownedby Dundee Precious Acquisitions Inc) operates

1 Geological Institute Bulgarian Academy of Sciences Acad G Bonchev St 1113 Sofia Bulgarialtpmarchevgeologybasbggt

2 Department of Geology and Geophysics University of Wisconsin-Madison 1215 West Dayton St Madison WI53706 USA

3 Balkan Mineral and Mining AD 24 Saedinenie St Krumovgrad Bulgaria4 Section des Sciences de la Terre Universiteacute de Genegraveve rue des Maraicircchers 13 CH 1211 Genegraveve Switzerland5 Department of Geology and Paleontology Faculty of Geology and Geography Sofia University ldquoSt Kliment

Ohridskirdquo 15 Tzar Osvoboditel Bd 1504 Sofia Bulgaria

P Marchev et al60

this project The deposit is unusual in that it is lo-cated along a regional detachment fault More-over it is geochemically distinctive with respect tothe adjacent deposits owing to high AuAg ratiosand low base metal content (Marchev et al 2003)

Here we summarize results from recent stud-ies on the surface geology structure alteration

types and textural and mineralogical characteris-tics of the gold mineralization The paper provideslimited Sr and Pb isotope data for Ada Tepe anddiscusses the age relationships among gold miner-alization local igneous activity and thermalevents in the nearby Kessebir metamorphic dome(Figs 1 2)

Fig 1 Schematic geological map of the Eastern Rhodope showing metamorphic dome structures major volcanicareas and dyke swarms Compiled from Ricou et al (1998) Yanev et al (1998a) and Marchev et al (2004) Insetshows location of Krumovgrad within Bulgaria

61Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

2 Historical background and previous studies

Small adits and large dumps at the eastern part ofthe Ada Tepe hill (Fig 3a) provide tangible evi-dence for the historical mining activity probablyof Thracian and Roman times Regional map-ping of the Krumovgrad area was conducted in1965 (Shabatov et al 1965) and recently be-tween 1995 and 1996 (Sarov et al 1996) by theState exploration Company ldquoGeology and Geo-

physicsrdquo at a 125 000 scale Studies by Goranovand Atanassov (1992) and a PhD thesis byBonev (2002) have clarified the stratigraphicand tectonic relationships of the Krumovgradarea The latter work advocates a detachment-fault model for the formation of the Kessebirmetamorphic dome Limited exploration in thearea of the Ada Tepe deposit was followed by thefirst description of its alteration and mineraliza-tion and its classification as AundashAg-polymetallic

Fig 2 Geological map of Kessebir dome (after Bonev 2002) and location of Ada Tepe deposit and Synap prospect

P Marchev et al62

Fig 3 (a) 3D view of Ada Tepe deposit (b) Cross section of Ada Tepe deposit Major ore bodies (c) Massivetabular ore body (ldquoThe Wallrdquo) above the detachment fault (d) Open space-filling ores along predominantly EndashWoriented listric faults

63Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

occurrence of the adularia-sericite type (Kunovet al 2001)

Since June 2000 the Ada Tepe deposit hasbeen included in the Krumovgrad license areaand is explored by Balkan Mineral and MiningAD Based on the results of regional mapping andsurface sampling BMM conducted two drillingprograms with 145 drill holes using both diamondand reverse circulation drilling techniques for atotal length of 12 440 m A total of 10 218 m ofsurface channel samples were also collected dur-ing 2001ndash2002 Resource calculations by RSGGlobal of Perth Australia show that Ada Tepe isa high-grade AundashAg deposit containing 615 Mt ofore at 46 gt Au for 910 000 ounces of Au using a10 gt cutoff grade

On the basis of initial results from the explora-tion program and the geographical proximity ofthe Ada Tepe deposit to the base metal mineraliza-tion at the Oligocene Zvezdel volcano Crummy(2002) suggested a direct genetic link between thebase and precious metal districts Marchev et al(2003) favoured an Upper Eocene age based onprecise 40Ar39Ar data and a detachment faultingmodel for the origin of the Ada Tepe deposit

3 Geologic setting

31 District geology

Basement metamorphic rocks in the Krumovgradarea build up the elongated Kessebir dome in NndashNE direction (Kozhoukharov et al 1988 Ricou etal 1998 Bonev 2002) (Figs 1 2) Two major tec-tonostratigraphic units (complexes) have beenrecognized on the basis of composition and tec-tonic setting of the metamorphic rocks a Gneiss-migmatite complex and a Variegated Complex(Kozhoukharov et al 1988 Haydoutov et al2001) which correspond to the Continental andMixed units of Ricou et al (1998) respectivelyThe structurally lower Gneiss-migmatite complexcrops out in the core of the Kessebir metamorphicdome It is dominated by igneous protoliths in-cluding metagranites migmatites and migmatizedgneisses overlain by a series of pelitic gneissesand rare amphibolites Eclogites and eclogitic am-phibolites have been described in the Greek partof the Kessebir dome (Mposkos and Krohe2000) RbndashSr ages of 3346 plusmn 34 Ma (Mposkosand Wawrzenitz 1995) from metapegmatites inGreece and RbndashSr (328 plusmn 25 Ma) and UndashPb zir-con ages of 310 plusmn 55 Ma and 319 plusmn 9 Ma (Pey-tcheva et al 1995 1998) from metagranites of theKessebir dome in Bulgaria indicate that theGneiss-migmatite complex is formed of Variscanor older continental basement

The overlaying Variegated complex consists ofa heterogeneous assemblage of pelitic schists para-gneisses amphibolites marbles and ophiolitebodies (Kozhoukharova 1984 Kolcheva and Es-kenazi 1998) Metamorphosed ophiolitic peridot-ites and amphibolitized eclogites are intruded bygabbros gabbronorites plagiogranites and dio-rites UndashPb zircon dating of a gabbro from theneighbouring Biala Reka dome (Fig 1) yields aLate Neoproterozoic age of 572 plusmn 5 Ma for thecore and outer zone recording a Hercynian meta-morphic event at ~300ndash350 Ma (Carrigan et al2003) Volumetrically minor plutonic bodies ofpresumably Upper Cretaceous age intrude theVariegated complex and represent an importantphase of igneous activity in the area (Belmustako-va et al 1995) The rocks are medium to fine-grained light grey two-mica granites of massiveto slightly foliated structure

The rocks of the Variegated complex in theKrumovgrad area are overlain by the Maastrich-tianndashPaleocene syndetachment Shavarovo For-mation (Goranov and Atanasov 1992 see be-low) which is in turn overlain by Upper EocenendashLower Oligocene coal-bearing-sandstone andmarl-limestone formations Lava flows anddomes of the ~35 Ma andesites (Lilov et al1987) of the Iran Tepe volcano are exposednortheast of Krumovgrad (Fig 1) This magmaticactivity was followed by scarce dykes of latitic torhyolitic composition in the northern part of theKessebir dome dated at 3182 plusmn 020 Ma and fi-nally by 26ndash28 Ma old intra-plate basaltic mag-matism in the southern part of the dome(Marchev et al 1997 1998)

32 Deposit geology

The Ada Tepe deposit is hosted by the syndetach-ment MaastrichtianndashPaleocene Shavarovo For-mation overlaying the northeastern closure ofthe Kessebir metamorphic core complex (Fig 2)The contact between poorly consolidated sedi-mentary rocks and underlying metamorphicrocks corresponds to the location of a regionallydeveloped low-angle normal fault named theTockachka detachment fault (TDF) by Bonev(1996) The fault dips 10ndash15deg to the north-north-east It can be traced further southwest for morethan 40 km from Krumovgrad to the Bulgarian-Greek frontier and is well exposed in the areasouthwest of Synap village about 2 km west ofAda Tepe In the latter area the rocks in the hang-ing wall of the TDF consist of metamorphicblocks breccia conglomerates sandstone marlsand argillaceous limestone This unit was first rec-ognized by Atanasov and Goranov (1984) in the

P Marchev et al64

Fig 4 Mineralization of The Wall showing multiple periods of silicification and veining (a) Brecciated early micro-crystalline silica (early stage) with a single gneiss fragment (b) Breccia of the microcrystalline silica (early stage)cemented by calcite (third stage) (c) Subvertical colloform finely banded vein Early stage gray (g) microcrystallinesilica on the margins of the veins followed by the massive microcrystalline (mq) cross-cut by the third stage sugaryquartz and massive and bladed calcite (qc) Late veinlets of dolomite-ankerite (da) cut early vein mineralization(d) Transition of subhorizontal opaline carbonate vein into subvertical step-like vein in the overlying sediments

65Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

Krumovgrad area and described as the ShavarovoFormation of the Krumovgrad Group in a laterpaper (Goranov and Atanasov 1992) The totalthickness of the type section of the ShavarovoFormation (~350 m) was measured along the Kal-djic river near the Shavar village Blocks andclasts of breccia and conglomerate are mixtures ofgneiss amphibolite and marble derived primarilyfrom the Variegated complex

4 Structure

The dominant structure in the Ada Tepe prospectis the TDF It is a mapable tectonic contact bestexpressed along the western slope of Ada Tepehill (Fig 3a) In outcrop and drill cores the fault isrecognized as a contact between Paleozoic meta-morphic rocks and MaastrichtianndashPaleocene sed-imentary rocks of the Shavarovo Formation Im-mediately above the detachment fault there is a05 to 20 m thick zone of massive silicification Indrill core the lower limit of the structure ismarked by a 10ndash20 cm-thick layer of tectonic clayGneiss migmatite marble and the amphibolite ofthe metamorphic basement beneath the detach-ment surface are transformed into a cataclasticrock over a thickness of several meters Thesouthwestern slope of the Ada Tepe hill is cut by aNWndashSE-oriented high-angle fault (Fig 2) whichformed a small half graben fill by older Maas-trichtianndashPaleocene syndetachment sedimentaryrocks overlain in the eastern part by younger Pria-bonianndashOligocene sedimentary rocks

Major structures cutting the basement rocks ofthe Ada Tepe deposit are rare although severalEndashW and NndashS-oriented subvertical faults markedby a zone of silicification occur to the south eastand west of the Ada Tepe deposit The longest NndashS-striking zone has been traced for almost 200 mThe EndashW and NndashS faults coincide with the majordirections of the mineralized veins in the sedi-mentary cover and can be interpreted as feederstructures for the mineralizing fluids

5 Mineralization

51 Ore geometry and ore bodies

The Ada Tepe deposit is 600 m along strike and300ndash350 m wide Gold mineralization occurs as(1) a massive tabular body located immediatelyabove the detachment fault (Figs 3b c) and (2)open space-filling within the breccia-conglomer-ate and sandstone along predominantly east-westoriented subvertical listric faults within the hang-

ing wall (Figs 3b d) In cross section the tabularbody dips (10ndash15deg) to the north-northeast whichis consistent with the low angle TDF The orebody is so strongly silicified company geologistsrefer to it as ldquoThe Wallrdquo Detailed observations onThe Wall indicate a complicated sequence ofevents with multiple periods of silicification andveining Silicification began with the deposition ofmassive white to light grey silica above the de-tachment fault partly or totally replacing the orig-inal rocks Original rock textures were destroyedwith the exception of ghosted gneiss fragments(Fig 4a) The latter are rounded with veins ofquartz or carbonate cutting or surrounding theclasts Early microcrystalline silica was commonlyveined by subvertical banded veins or brecciatedby later fault movements and cemented by calcitesugary quartz (Fig 5b) or by pyrite

In the less silicified parts The Wall consists ofnumerous differently oriented veins ranging inwidth from less than 1 cm up to 30 cm (Fig 4c)Subvertical veins which were formed by deposi-tion of colloform silica layers in open space canbe traced from The Wall into the listric faults inthe overlying sedimentary rocks (Fig 4d) On thesurface they form EndashW-trending ridges separatedby argillic zones (Figs 3a b d)

52 Ore and gangue mineralogy andstages of mineralization

The ore mineralogy of the Ada Tepe deposit is sim-ple consisting mainly of electrum and subordinatepyrite with traces of galena and gold-silver tellu-rides Gangue minerals consist of silica polymorphs(microcrystalline fine-grained sugary quartz andopaline silica) various carbonates (calcite dolo-mite ankerite and siderite) and adularia Telluridesinclude hessite (Ag2Te) and petzite (Ag3AuTe2) Allsamples with Au content higher than 1 gt have al-most constant AuAg ratios ~ 3 reflecting the com-position of electrum (76ndash73 wt Au)

Unlike the massive silicification of The Wallbanded veins in and above it provide an excellentframework for identifying the different stages ofmineralization Mineralization in Ada Tepe canbe subdivided into 4 broad generally overlappingstages of quartz and carbonate veins on the basisof cross-cutting relationships changes in mineral-ogy texture and gold grades

The earliest stage is characterized by deposi-tion of microcrystalline to fine-grained grayish towhite massive or banded quartz with rare pyriteand adularia (Figs 5andashd) In some veins silica depo-sition starts with a thin band of almost isotropicopaline (Fig 4 c) followed by thin band of quartzwith rare bladed calcite Breccia texture is typical

P Marchev et al66

for the massive silicification of The Wall Theproducts of the early stage form microfracturefilling in the host rocks near the margins of the

Fig 5 Microphotographs and SEM images of silica and bonanza ore textures (a) Bands of early stage microcrystal-line (mq) and jigsaw quartz (jq) (b) Electrum band (e) deposited on the interface between early microcrystallinequartz (mq) and third stage sugary quartz and radiating bladed calcite (qc) (c) Band of electrum (e) and isotropicopaline Electrum is located on the bottom of the band (d) Third stage coarse crystalline quartz (ccq) cutting theopaline (o) and microcrystalline quartz (mq) which in turn is cut by calcite (c) (e) Subparallel bands of electrum (e)and microcrystalline adularia and silica separated by a calcite vein (c) The arrow shows the vertical direction (f)SEM image of the outlined area

veins or rock fragments in The Wall causing thesilica-adularia-pyrite replacement in them Thegold grade of this stage is unknown

67Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

The economic gold is related to the secondstage of mineralization It starts at the end of themicrocrystalline quartz and the beginning of thecoarser grained (sugary) quartz of the third stagebelow (Fig 5b) Usually it forms black millimeter-scale bands of opaline or microcrystalline chal-cedony-adularia and electrum (Figs 5cndashf) Locallyopaline silica crosscuts the microcrystallinequartz (Figs 5cndashd) of the early stage and in turn iscut by the coarser grained quartz and carbonatesof the third and forth stages (Figs 5dndashe) Petro-graphic examination of the electrum-rich bandsreveals that electrum commonly occurs towardsthe bottom of the opaline or microcrystallinebands (Figs 5c e) Where opaline bands are notpresent electrum precipitates at the boundary offirst and third stages (Fig 5b) The electrum hasirregular forms as the result of crystallization inthe quartz interstices being relatively well-shapedin the coarser-grained quartz crystals The size ofelectrum grains ranges from less than 1 to 20 micromHowever the single crystals form chains reachingup to 650 microm (Figs 5b 6a) Electrum in gold-richbands etched with HF is characterized by botryoi-dal surfaces that appear to have formed by theaggregation of spherical electrum particles (Figs6andashb) as observed in the Sleeper deposit Nevada(Saunders 1990 1994) Electrum in bonanza veinscan occupy 50 vol (Figs 5f 6c) accompanied byrare As-bearing pyrite Occasionally micron orsubmicron size electrum inclusions form zoneswithin the As-bearing pyrite (Figs 6d) Tiny AundashAg tellurides have been observed at the contactsbetween electrum grains and small galena crystals(Fig 6f)

The third stage starts with deposition of sugaryquartz (up to 15 mm large Figs 5d 7a) followedby bands of quartz and bladed calcite (Figs 5b 7b)and finally by massive calcite with small amountof quartz (Fig 5d) Locally calcite fills quartz ge-odes or forms small veins (Fig 7a) Microscopical-ly the late calcite corrodes and replaces quartzcrystals Small amount of adularia and pyrite arealso typical for this stage

The hydrothermal mineralization ended withthe deposition of pinkish ankerite-dolomite andsiderite which was preceded by precipitation ofmassive pyrite in some parts of the deposit (Figs7cndashd)

The mineralization of Ada Tepe exhibits sever-al important spatial and textural features thatprovide insights to the physical environment ofore deposition In its upper part quartz becomesthe predominant gangue mineral whereas car-bonates are subordinate This is particularly obvi-ous in the last stage where in the surface samplesdolomite and ankerite are absent in the geodes

and bladed minerals consist of microcrystallinequartz only Another spatial variation is the di-minishing of banding textures with the decreasein grade of Au mineralization to the north Animportant feature is the absence of bladed car-bonates in the veinlets beneath The Wall in con-trast to the entire extent of the deposit above thedetachment surface

53 Oxidized zone

The uppermost part of the Ada Tepe deposit hasbeen oxidized by supergene processes with thedepth of oxidation being dependent upon inten-sity of faulting Due to extensive silicification TheWall itself is generally not oxidized however a 2to 5 m thick zone of oxidation invariably occursabove The Wall Narrow zones of oxidation occurdown to this zone utilizing high angle listric faultsThese faults and fractures most likely acted asconduits for the surface waters to reach the im-permeable silicified zone forming the oxidizedzone above The Wall

Most of the pyrite in the oxidized zones hasbeen converted into limonite identified asgoethite Kaolinite is widely distributed in thiszone forming locally small veinlets Electrum inthe oxidized zone however shows no evidence ofAu enrichment

6 Alteration

In the poorly mineralized zones hydrothermal al-teration is weak with most of the detrital minerals(quartz muscovite and feldspars) remaining unal-tered Alteration consists of chlorite calcite kaoli-nite and subordinate albite pyrite and ankerite-dolomite It seems that calcite and ankerite-dolo-mite are late however it is difficult to distinguishreplacement calcite from vein calcite Quartzadularia calcite pyrite and dolomite-ankerite-siderite plusmn sericite and clay minerals in variableproportions occur in The Wall and immediateproximity to the subvertical quartz and carbonateveins in the strongly mineralized zones Theyformed through replacement of the original meta-morphic minerals included in the sandstone andconglomerate and are associated with relict de-trital quartz and muscovite This complex mineralassemblage is the result of the evolution of thehydrothermal activity

Hydrothermal alteration of the basementmetamorphic rocks is developed within a haloseveral meters in extent beneath The Wall and ismore pronounced around the subvertical veinsThe alteration assemblage includes quartz adu-

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

P Marchev et al60

this project The deposit is unusual in that it is lo-cated along a regional detachment fault More-over it is geochemically distinctive with respect tothe adjacent deposits owing to high AuAg ratiosand low base metal content (Marchev et al 2003)

Here we summarize results from recent stud-ies on the surface geology structure alteration

types and textural and mineralogical characteris-tics of the gold mineralization The paper provideslimited Sr and Pb isotope data for Ada Tepe anddiscusses the age relationships among gold miner-alization local igneous activity and thermalevents in the nearby Kessebir metamorphic dome(Figs 1 2)

Fig 1 Schematic geological map of the Eastern Rhodope showing metamorphic dome structures major volcanicareas and dyke swarms Compiled from Ricou et al (1998) Yanev et al (1998a) and Marchev et al (2004) Insetshows location of Krumovgrad within Bulgaria

61Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

2 Historical background and previous studies

Small adits and large dumps at the eastern part ofthe Ada Tepe hill (Fig 3a) provide tangible evi-dence for the historical mining activity probablyof Thracian and Roman times Regional map-ping of the Krumovgrad area was conducted in1965 (Shabatov et al 1965) and recently be-tween 1995 and 1996 (Sarov et al 1996) by theState exploration Company ldquoGeology and Geo-

physicsrdquo at a 125 000 scale Studies by Goranovand Atanassov (1992) and a PhD thesis byBonev (2002) have clarified the stratigraphicand tectonic relationships of the Krumovgradarea The latter work advocates a detachment-fault model for the formation of the Kessebirmetamorphic dome Limited exploration in thearea of the Ada Tepe deposit was followed by thefirst description of its alteration and mineraliza-tion and its classification as AundashAg-polymetallic

Fig 2 Geological map of Kessebir dome (after Bonev 2002) and location of Ada Tepe deposit and Synap prospect

P Marchev et al62

Fig 3 (a) 3D view of Ada Tepe deposit (b) Cross section of Ada Tepe deposit Major ore bodies (c) Massivetabular ore body (ldquoThe Wallrdquo) above the detachment fault (d) Open space-filling ores along predominantly EndashWoriented listric faults

63Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

occurrence of the adularia-sericite type (Kunovet al 2001)

Since June 2000 the Ada Tepe deposit hasbeen included in the Krumovgrad license areaand is explored by Balkan Mineral and MiningAD Based on the results of regional mapping andsurface sampling BMM conducted two drillingprograms with 145 drill holes using both diamondand reverse circulation drilling techniques for atotal length of 12 440 m A total of 10 218 m ofsurface channel samples were also collected dur-ing 2001ndash2002 Resource calculations by RSGGlobal of Perth Australia show that Ada Tepe isa high-grade AundashAg deposit containing 615 Mt ofore at 46 gt Au for 910 000 ounces of Au using a10 gt cutoff grade

On the basis of initial results from the explora-tion program and the geographical proximity ofthe Ada Tepe deposit to the base metal mineraliza-tion at the Oligocene Zvezdel volcano Crummy(2002) suggested a direct genetic link between thebase and precious metal districts Marchev et al(2003) favoured an Upper Eocene age based onprecise 40Ar39Ar data and a detachment faultingmodel for the origin of the Ada Tepe deposit

3 Geologic setting

31 District geology

Basement metamorphic rocks in the Krumovgradarea build up the elongated Kessebir dome in NndashNE direction (Kozhoukharov et al 1988 Ricou etal 1998 Bonev 2002) (Figs 1 2) Two major tec-tonostratigraphic units (complexes) have beenrecognized on the basis of composition and tec-tonic setting of the metamorphic rocks a Gneiss-migmatite complex and a Variegated Complex(Kozhoukharov et al 1988 Haydoutov et al2001) which correspond to the Continental andMixed units of Ricou et al (1998) respectivelyThe structurally lower Gneiss-migmatite complexcrops out in the core of the Kessebir metamorphicdome It is dominated by igneous protoliths in-cluding metagranites migmatites and migmatizedgneisses overlain by a series of pelitic gneissesand rare amphibolites Eclogites and eclogitic am-phibolites have been described in the Greek partof the Kessebir dome (Mposkos and Krohe2000) RbndashSr ages of 3346 plusmn 34 Ma (Mposkosand Wawrzenitz 1995) from metapegmatites inGreece and RbndashSr (328 plusmn 25 Ma) and UndashPb zir-con ages of 310 plusmn 55 Ma and 319 plusmn 9 Ma (Pey-tcheva et al 1995 1998) from metagranites of theKessebir dome in Bulgaria indicate that theGneiss-migmatite complex is formed of Variscanor older continental basement

The overlaying Variegated complex consists ofa heterogeneous assemblage of pelitic schists para-gneisses amphibolites marbles and ophiolitebodies (Kozhoukharova 1984 Kolcheva and Es-kenazi 1998) Metamorphosed ophiolitic peridot-ites and amphibolitized eclogites are intruded bygabbros gabbronorites plagiogranites and dio-rites UndashPb zircon dating of a gabbro from theneighbouring Biala Reka dome (Fig 1) yields aLate Neoproterozoic age of 572 plusmn 5 Ma for thecore and outer zone recording a Hercynian meta-morphic event at ~300ndash350 Ma (Carrigan et al2003) Volumetrically minor plutonic bodies ofpresumably Upper Cretaceous age intrude theVariegated complex and represent an importantphase of igneous activity in the area (Belmustako-va et al 1995) The rocks are medium to fine-grained light grey two-mica granites of massiveto slightly foliated structure

The rocks of the Variegated complex in theKrumovgrad area are overlain by the Maastrich-tianndashPaleocene syndetachment Shavarovo For-mation (Goranov and Atanasov 1992 see be-low) which is in turn overlain by Upper EocenendashLower Oligocene coal-bearing-sandstone andmarl-limestone formations Lava flows anddomes of the ~35 Ma andesites (Lilov et al1987) of the Iran Tepe volcano are exposednortheast of Krumovgrad (Fig 1) This magmaticactivity was followed by scarce dykes of latitic torhyolitic composition in the northern part of theKessebir dome dated at 3182 plusmn 020 Ma and fi-nally by 26ndash28 Ma old intra-plate basaltic mag-matism in the southern part of the dome(Marchev et al 1997 1998)

32 Deposit geology

The Ada Tepe deposit is hosted by the syndetach-ment MaastrichtianndashPaleocene Shavarovo For-mation overlaying the northeastern closure ofthe Kessebir metamorphic core complex (Fig 2)The contact between poorly consolidated sedi-mentary rocks and underlying metamorphicrocks corresponds to the location of a regionallydeveloped low-angle normal fault named theTockachka detachment fault (TDF) by Bonev(1996) The fault dips 10ndash15deg to the north-north-east It can be traced further southwest for morethan 40 km from Krumovgrad to the Bulgarian-Greek frontier and is well exposed in the areasouthwest of Synap village about 2 km west ofAda Tepe In the latter area the rocks in the hang-ing wall of the TDF consist of metamorphicblocks breccia conglomerates sandstone marlsand argillaceous limestone This unit was first rec-ognized by Atanasov and Goranov (1984) in the

P Marchev et al64

Fig 4 Mineralization of The Wall showing multiple periods of silicification and veining (a) Brecciated early micro-crystalline silica (early stage) with a single gneiss fragment (b) Breccia of the microcrystalline silica (early stage)cemented by calcite (third stage) (c) Subvertical colloform finely banded vein Early stage gray (g) microcrystallinesilica on the margins of the veins followed by the massive microcrystalline (mq) cross-cut by the third stage sugaryquartz and massive and bladed calcite (qc) Late veinlets of dolomite-ankerite (da) cut early vein mineralization(d) Transition of subhorizontal opaline carbonate vein into subvertical step-like vein in the overlying sediments

65Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

Krumovgrad area and described as the ShavarovoFormation of the Krumovgrad Group in a laterpaper (Goranov and Atanasov 1992) The totalthickness of the type section of the ShavarovoFormation (~350 m) was measured along the Kal-djic river near the Shavar village Blocks andclasts of breccia and conglomerate are mixtures ofgneiss amphibolite and marble derived primarilyfrom the Variegated complex

4 Structure

The dominant structure in the Ada Tepe prospectis the TDF It is a mapable tectonic contact bestexpressed along the western slope of Ada Tepehill (Fig 3a) In outcrop and drill cores the fault isrecognized as a contact between Paleozoic meta-morphic rocks and MaastrichtianndashPaleocene sed-imentary rocks of the Shavarovo Formation Im-mediately above the detachment fault there is a05 to 20 m thick zone of massive silicification Indrill core the lower limit of the structure ismarked by a 10ndash20 cm-thick layer of tectonic clayGneiss migmatite marble and the amphibolite ofthe metamorphic basement beneath the detach-ment surface are transformed into a cataclasticrock over a thickness of several meters Thesouthwestern slope of the Ada Tepe hill is cut by aNWndashSE-oriented high-angle fault (Fig 2) whichformed a small half graben fill by older Maas-trichtianndashPaleocene syndetachment sedimentaryrocks overlain in the eastern part by younger Pria-bonianndashOligocene sedimentary rocks

Major structures cutting the basement rocks ofthe Ada Tepe deposit are rare although severalEndashW and NndashS-oriented subvertical faults markedby a zone of silicification occur to the south eastand west of the Ada Tepe deposit The longest NndashS-striking zone has been traced for almost 200 mThe EndashW and NndashS faults coincide with the majordirections of the mineralized veins in the sedi-mentary cover and can be interpreted as feederstructures for the mineralizing fluids

5 Mineralization

51 Ore geometry and ore bodies

The Ada Tepe deposit is 600 m along strike and300ndash350 m wide Gold mineralization occurs as(1) a massive tabular body located immediatelyabove the detachment fault (Figs 3b c) and (2)open space-filling within the breccia-conglomer-ate and sandstone along predominantly east-westoriented subvertical listric faults within the hang-

ing wall (Figs 3b d) In cross section the tabularbody dips (10ndash15deg) to the north-northeast whichis consistent with the low angle TDF The orebody is so strongly silicified company geologistsrefer to it as ldquoThe Wallrdquo Detailed observations onThe Wall indicate a complicated sequence ofevents with multiple periods of silicification andveining Silicification began with the deposition ofmassive white to light grey silica above the de-tachment fault partly or totally replacing the orig-inal rocks Original rock textures were destroyedwith the exception of ghosted gneiss fragments(Fig 4a) The latter are rounded with veins ofquartz or carbonate cutting or surrounding theclasts Early microcrystalline silica was commonlyveined by subvertical banded veins or brecciatedby later fault movements and cemented by calcitesugary quartz (Fig 5b) or by pyrite

In the less silicified parts The Wall consists ofnumerous differently oriented veins ranging inwidth from less than 1 cm up to 30 cm (Fig 4c)Subvertical veins which were formed by deposi-tion of colloform silica layers in open space canbe traced from The Wall into the listric faults inthe overlying sedimentary rocks (Fig 4d) On thesurface they form EndashW-trending ridges separatedby argillic zones (Figs 3a b d)

52 Ore and gangue mineralogy andstages of mineralization

The ore mineralogy of the Ada Tepe deposit is sim-ple consisting mainly of electrum and subordinatepyrite with traces of galena and gold-silver tellu-rides Gangue minerals consist of silica polymorphs(microcrystalline fine-grained sugary quartz andopaline silica) various carbonates (calcite dolo-mite ankerite and siderite) and adularia Telluridesinclude hessite (Ag2Te) and petzite (Ag3AuTe2) Allsamples with Au content higher than 1 gt have al-most constant AuAg ratios ~ 3 reflecting the com-position of electrum (76ndash73 wt Au)

Unlike the massive silicification of The Wallbanded veins in and above it provide an excellentframework for identifying the different stages ofmineralization Mineralization in Ada Tepe canbe subdivided into 4 broad generally overlappingstages of quartz and carbonate veins on the basisof cross-cutting relationships changes in mineral-ogy texture and gold grades

The earliest stage is characterized by deposi-tion of microcrystalline to fine-grained grayish towhite massive or banded quartz with rare pyriteand adularia (Figs 5andashd) In some veins silica depo-sition starts with a thin band of almost isotropicopaline (Fig 4 c) followed by thin band of quartzwith rare bladed calcite Breccia texture is typical

P Marchev et al66

for the massive silicification of The Wall Theproducts of the early stage form microfracturefilling in the host rocks near the margins of the

Fig 5 Microphotographs and SEM images of silica and bonanza ore textures (a) Bands of early stage microcrystal-line (mq) and jigsaw quartz (jq) (b) Electrum band (e) deposited on the interface between early microcrystallinequartz (mq) and third stage sugary quartz and radiating bladed calcite (qc) (c) Band of electrum (e) and isotropicopaline Electrum is located on the bottom of the band (d) Third stage coarse crystalline quartz (ccq) cutting theopaline (o) and microcrystalline quartz (mq) which in turn is cut by calcite (c) (e) Subparallel bands of electrum (e)and microcrystalline adularia and silica separated by a calcite vein (c) The arrow shows the vertical direction (f)SEM image of the outlined area

veins or rock fragments in The Wall causing thesilica-adularia-pyrite replacement in them Thegold grade of this stage is unknown

67Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

The economic gold is related to the secondstage of mineralization It starts at the end of themicrocrystalline quartz and the beginning of thecoarser grained (sugary) quartz of the third stagebelow (Fig 5b) Usually it forms black millimeter-scale bands of opaline or microcrystalline chal-cedony-adularia and electrum (Figs 5cndashf) Locallyopaline silica crosscuts the microcrystallinequartz (Figs 5cndashd) of the early stage and in turn iscut by the coarser grained quartz and carbonatesof the third and forth stages (Figs 5dndashe) Petro-graphic examination of the electrum-rich bandsreveals that electrum commonly occurs towardsthe bottom of the opaline or microcrystallinebands (Figs 5c e) Where opaline bands are notpresent electrum precipitates at the boundary offirst and third stages (Fig 5b) The electrum hasirregular forms as the result of crystallization inthe quartz interstices being relatively well-shapedin the coarser-grained quartz crystals The size ofelectrum grains ranges from less than 1 to 20 micromHowever the single crystals form chains reachingup to 650 microm (Figs 5b 6a) Electrum in gold-richbands etched with HF is characterized by botryoi-dal surfaces that appear to have formed by theaggregation of spherical electrum particles (Figs6andashb) as observed in the Sleeper deposit Nevada(Saunders 1990 1994) Electrum in bonanza veinscan occupy 50 vol (Figs 5f 6c) accompanied byrare As-bearing pyrite Occasionally micron orsubmicron size electrum inclusions form zoneswithin the As-bearing pyrite (Figs 6d) Tiny AundashAg tellurides have been observed at the contactsbetween electrum grains and small galena crystals(Fig 6f)

The third stage starts with deposition of sugaryquartz (up to 15 mm large Figs 5d 7a) followedby bands of quartz and bladed calcite (Figs 5b 7b)and finally by massive calcite with small amountof quartz (Fig 5d) Locally calcite fills quartz ge-odes or forms small veins (Fig 7a) Microscopical-ly the late calcite corrodes and replaces quartzcrystals Small amount of adularia and pyrite arealso typical for this stage

The hydrothermal mineralization ended withthe deposition of pinkish ankerite-dolomite andsiderite which was preceded by precipitation ofmassive pyrite in some parts of the deposit (Figs7cndashd)

The mineralization of Ada Tepe exhibits sever-al important spatial and textural features thatprovide insights to the physical environment ofore deposition In its upper part quartz becomesthe predominant gangue mineral whereas car-bonates are subordinate This is particularly obvi-ous in the last stage where in the surface samplesdolomite and ankerite are absent in the geodes

and bladed minerals consist of microcrystallinequartz only Another spatial variation is the di-minishing of banding textures with the decreasein grade of Au mineralization to the north Animportant feature is the absence of bladed car-bonates in the veinlets beneath The Wall in con-trast to the entire extent of the deposit above thedetachment surface

53 Oxidized zone

The uppermost part of the Ada Tepe deposit hasbeen oxidized by supergene processes with thedepth of oxidation being dependent upon inten-sity of faulting Due to extensive silicification TheWall itself is generally not oxidized however a 2to 5 m thick zone of oxidation invariably occursabove The Wall Narrow zones of oxidation occurdown to this zone utilizing high angle listric faultsThese faults and fractures most likely acted asconduits for the surface waters to reach the im-permeable silicified zone forming the oxidizedzone above The Wall

Most of the pyrite in the oxidized zones hasbeen converted into limonite identified asgoethite Kaolinite is widely distributed in thiszone forming locally small veinlets Electrum inthe oxidized zone however shows no evidence ofAu enrichment

6 Alteration

In the poorly mineralized zones hydrothermal al-teration is weak with most of the detrital minerals(quartz muscovite and feldspars) remaining unal-tered Alteration consists of chlorite calcite kaoli-nite and subordinate albite pyrite and ankerite-dolomite It seems that calcite and ankerite-dolo-mite are late however it is difficult to distinguishreplacement calcite from vein calcite Quartzadularia calcite pyrite and dolomite-ankerite-siderite plusmn sericite and clay minerals in variableproportions occur in The Wall and immediateproximity to the subvertical quartz and carbonateveins in the strongly mineralized zones Theyformed through replacement of the original meta-morphic minerals included in the sandstone andconglomerate and are associated with relict de-trital quartz and muscovite This complex mineralassemblage is the result of the evolution of thehydrothermal activity

Hydrothermal alteration of the basementmetamorphic rocks is developed within a haloseveral meters in extent beneath The Wall and ismore pronounced around the subvertical veinsThe alteration assemblage includes quartz adu-

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

61Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

2 Historical background and previous studies

Small adits and large dumps at the eastern part ofthe Ada Tepe hill (Fig 3a) provide tangible evi-dence for the historical mining activity probablyof Thracian and Roman times Regional map-ping of the Krumovgrad area was conducted in1965 (Shabatov et al 1965) and recently be-tween 1995 and 1996 (Sarov et al 1996) by theState exploration Company ldquoGeology and Geo-

physicsrdquo at a 125 000 scale Studies by Goranovand Atanassov (1992) and a PhD thesis byBonev (2002) have clarified the stratigraphicand tectonic relationships of the Krumovgradarea The latter work advocates a detachment-fault model for the formation of the Kessebirmetamorphic dome Limited exploration in thearea of the Ada Tepe deposit was followed by thefirst description of its alteration and mineraliza-tion and its classification as AundashAg-polymetallic

Fig 2 Geological map of Kessebir dome (after Bonev 2002) and location of Ada Tepe deposit and Synap prospect

P Marchev et al62

Fig 3 (a) 3D view of Ada Tepe deposit (b) Cross section of Ada Tepe deposit Major ore bodies (c) Massivetabular ore body (ldquoThe Wallrdquo) above the detachment fault (d) Open space-filling ores along predominantly EndashWoriented listric faults

63Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

occurrence of the adularia-sericite type (Kunovet al 2001)

Since June 2000 the Ada Tepe deposit hasbeen included in the Krumovgrad license areaand is explored by Balkan Mineral and MiningAD Based on the results of regional mapping andsurface sampling BMM conducted two drillingprograms with 145 drill holes using both diamondand reverse circulation drilling techniques for atotal length of 12 440 m A total of 10 218 m ofsurface channel samples were also collected dur-ing 2001ndash2002 Resource calculations by RSGGlobal of Perth Australia show that Ada Tepe isa high-grade AundashAg deposit containing 615 Mt ofore at 46 gt Au for 910 000 ounces of Au using a10 gt cutoff grade

On the basis of initial results from the explora-tion program and the geographical proximity ofthe Ada Tepe deposit to the base metal mineraliza-tion at the Oligocene Zvezdel volcano Crummy(2002) suggested a direct genetic link between thebase and precious metal districts Marchev et al(2003) favoured an Upper Eocene age based onprecise 40Ar39Ar data and a detachment faultingmodel for the origin of the Ada Tepe deposit

3 Geologic setting

31 District geology

Basement metamorphic rocks in the Krumovgradarea build up the elongated Kessebir dome in NndashNE direction (Kozhoukharov et al 1988 Ricou etal 1998 Bonev 2002) (Figs 1 2) Two major tec-tonostratigraphic units (complexes) have beenrecognized on the basis of composition and tec-tonic setting of the metamorphic rocks a Gneiss-migmatite complex and a Variegated Complex(Kozhoukharov et al 1988 Haydoutov et al2001) which correspond to the Continental andMixed units of Ricou et al (1998) respectivelyThe structurally lower Gneiss-migmatite complexcrops out in the core of the Kessebir metamorphicdome It is dominated by igneous protoliths in-cluding metagranites migmatites and migmatizedgneisses overlain by a series of pelitic gneissesand rare amphibolites Eclogites and eclogitic am-phibolites have been described in the Greek partof the Kessebir dome (Mposkos and Krohe2000) RbndashSr ages of 3346 plusmn 34 Ma (Mposkosand Wawrzenitz 1995) from metapegmatites inGreece and RbndashSr (328 plusmn 25 Ma) and UndashPb zir-con ages of 310 plusmn 55 Ma and 319 plusmn 9 Ma (Pey-tcheva et al 1995 1998) from metagranites of theKessebir dome in Bulgaria indicate that theGneiss-migmatite complex is formed of Variscanor older continental basement

The overlaying Variegated complex consists ofa heterogeneous assemblage of pelitic schists para-gneisses amphibolites marbles and ophiolitebodies (Kozhoukharova 1984 Kolcheva and Es-kenazi 1998) Metamorphosed ophiolitic peridot-ites and amphibolitized eclogites are intruded bygabbros gabbronorites plagiogranites and dio-rites UndashPb zircon dating of a gabbro from theneighbouring Biala Reka dome (Fig 1) yields aLate Neoproterozoic age of 572 plusmn 5 Ma for thecore and outer zone recording a Hercynian meta-morphic event at ~300ndash350 Ma (Carrigan et al2003) Volumetrically minor plutonic bodies ofpresumably Upper Cretaceous age intrude theVariegated complex and represent an importantphase of igneous activity in the area (Belmustako-va et al 1995) The rocks are medium to fine-grained light grey two-mica granites of massiveto slightly foliated structure

The rocks of the Variegated complex in theKrumovgrad area are overlain by the Maastrich-tianndashPaleocene syndetachment Shavarovo For-mation (Goranov and Atanasov 1992 see be-low) which is in turn overlain by Upper EocenendashLower Oligocene coal-bearing-sandstone andmarl-limestone formations Lava flows anddomes of the ~35 Ma andesites (Lilov et al1987) of the Iran Tepe volcano are exposednortheast of Krumovgrad (Fig 1) This magmaticactivity was followed by scarce dykes of latitic torhyolitic composition in the northern part of theKessebir dome dated at 3182 plusmn 020 Ma and fi-nally by 26ndash28 Ma old intra-plate basaltic mag-matism in the southern part of the dome(Marchev et al 1997 1998)

32 Deposit geology

The Ada Tepe deposit is hosted by the syndetach-ment MaastrichtianndashPaleocene Shavarovo For-mation overlaying the northeastern closure ofthe Kessebir metamorphic core complex (Fig 2)The contact between poorly consolidated sedi-mentary rocks and underlying metamorphicrocks corresponds to the location of a regionallydeveloped low-angle normal fault named theTockachka detachment fault (TDF) by Bonev(1996) The fault dips 10ndash15deg to the north-north-east It can be traced further southwest for morethan 40 km from Krumovgrad to the Bulgarian-Greek frontier and is well exposed in the areasouthwest of Synap village about 2 km west ofAda Tepe In the latter area the rocks in the hang-ing wall of the TDF consist of metamorphicblocks breccia conglomerates sandstone marlsand argillaceous limestone This unit was first rec-ognized by Atanasov and Goranov (1984) in the

P Marchev et al64

Fig 4 Mineralization of The Wall showing multiple periods of silicification and veining (a) Brecciated early micro-crystalline silica (early stage) with a single gneiss fragment (b) Breccia of the microcrystalline silica (early stage)cemented by calcite (third stage) (c) Subvertical colloform finely banded vein Early stage gray (g) microcrystallinesilica on the margins of the veins followed by the massive microcrystalline (mq) cross-cut by the third stage sugaryquartz and massive and bladed calcite (qc) Late veinlets of dolomite-ankerite (da) cut early vein mineralization(d) Transition of subhorizontal opaline carbonate vein into subvertical step-like vein in the overlying sediments

65Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

Krumovgrad area and described as the ShavarovoFormation of the Krumovgrad Group in a laterpaper (Goranov and Atanasov 1992) The totalthickness of the type section of the ShavarovoFormation (~350 m) was measured along the Kal-djic river near the Shavar village Blocks andclasts of breccia and conglomerate are mixtures ofgneiss amphibolite and marble derived primarilyfrom the Variegated complex

4 Structure

The dominant structure in the Ada Tepe prospectis the TDF It is a mapable tectonic contact bestexpressed along the western slope of Ada Tepehill (Fig 3a) In outcrop and drill cores the fault isrecognized as a contact between Paleozoic meta-morphic rocks and MaastrichtianndashPaleocene sed-imentary rocks of the Shavarovo Formation Im-mediately above the detachment fault there is a05 to 20 m thick zone of massive silicification Indrill core the lower limit of the structure ismarked by a 10ndash20 cm-thick layer of tectonic clayGneiss migmatite marble and the amphibolite ofthe metamorphic basement beneath the detach-ment surface are transformed into a cataclasticrock over a thickness of several meters Thesouthwestern slope of the Ada Tepe hill is cut by aNWndashSE-oriented high-angle fault (Fig 2) whichformed a small half graben fill by older Maas-trichtianndashPaleocene syndetachment sedimentaryrocks overlain in the eastern part by younger Pria-bonianndashOligocene sedimentary rocks

Major structures cutting the basement rocks ofthe Ada Tepe deposit are rare although severalEndashW and NndashS-oriented subvertical faults markedby a zone of silicification occur to the south eastand west of the Ada Tepe deposit The longest NndashS-striking zone has been traced for almost 200 mThe EndashW and NndashS faults coincide with the majordirections of the mineralized veins in the sedi-mentary cover and can be interpreted as feederstructures for the mineralizing fluids

5 Mineralization

51 Ore geometry and ore bodies

The Ada Tepe deposit is 600 m along strike and300ndash350 m wide Gold mineralization occurs as(1) a massive tabular body located immediatelyabove the detachment fault (Figs 3b c) and (2)open space-filling within the breccia-conglomer-ate and sandstone along predominantly east-westoriented subvertical listric faults within the hang-

ing wall (Figs 3b d) In cross section the tabularbody dips (10ndash15deg) to the north-northeast whichis consistent with the low angle TDF The orebody is so strongly silicified company geologistsrefer to it as ldquoThe Wallrdquo Detailed observations onThe Wall indicate a complicated sequence ofevents with multiple periods of silicification andveining Silicification began with the deposition ofmassive white to light grey silica above the de-tachment fault partly or totally replacing the orig-inal rocks Original rock textures were destroyedwith the exception of ghosted gneiss fragments(Fig 4a) The latter are rounded with veins ofquartz or carbonate cutting or surrounding theclasts Early microcrystalline silica was commonlyveined by subvertical banded veins or brecciatedby later fault movements and cemented by calcitesugary quartz (Fig 5b) or by pyrite

In the less silicified parts The Wall consists ofnumerous differently oriented veins ranging inwidth from less than 1 cm up to 30 cm (Fig 4c)Subvertical veins which were formed by deposi-tion of colloform silica layers in open space canbe traced from The Wall into the listric faults inthe overlying sedimentary rocks (Fig 4d) On thesurface they form EndashW-trending ridges separatedby argillic zones (Figs 3a b d)

52 Ore and gangue mineralogy andstages of mineralization

The ore mineralogy of the Ada Tepe deposit is sim-ple consisting mainly of electrum and subordinatepyrite with traces of galena and gold-silver tellu-rides Gangue minerals consist of silica polymorphs(microcrystalline fine-grained sugary quartz andopaline silica) various carbonates (calcite dolo-mite ankerite and siderite) and adularia Telluridesinclude hessite (Ag2Te) and petzite (Ag3AuTe2) Allsamples with Au content higher than 1 gt have al-most constant AuAg ratios ~ 3 reflecting the com-position of electrum (76ndash73 wt Au)

Unlike the massive silicification of The Wallbanded veins in and above it provide an excellentframework for identifying the different stages ofmineralization Mineralization in Ada Tepe canbe subdivided into 4 broad generally overlappingstages of quartz and carbonate veins on the basisof cross-cutting relationships changes in mineral-ogy texture and gold grades

The earliest stage is characterized by deposi-tion of microcrystalline to fine-grained grayish towhite massive or banded quartz with rare pyriteand adularia (Figs 5andashd) In some veins silica depo-sition starts with a thin band of almost isotropicopaline (Fig 4 c) followed by thin band of quartzwith rare bladed calcite Breccia texture is typical

P Marchev et al66

for the massive silicification of The Wall Theproducts of the early stage form microfracturefilling in the host rocks near the margins of the

Fig 5 Microphotographs and SEM images of silica and bonanza ore textures (a) Bands of early stage microcrystal-line (mq) and jigsaw quartz (jq) (b) Electrum band (e) deposited on the interface between early microcrystallinequartz (mq) and third stage sugary quartz and radiating bladed calcite (qc) (c) Band of electrum (e) and isotropicopaline Electrum is located on the bottom of the band (d) Third stage coarse crystalline quartz (ccq) cutting theopaline (o) and microcrystalline quartz (mq) which in turn is cut by calcite (c) (e) Subparallel bands of electrum (e)and microcrystalline adularia and silica separated by a calcite vein (c) The arrow shows the vertical direction (f)SEM image of the outlined area

veins or rock fragments in The Wall causing thesilica-adularia-pyrite replacement in them Thegold grade of this stage is unknown

67Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

The economic gold is related to the secondstage of mineralization It starts at the end of themicrocrystalline quartz and the beginning of thecoarser grained (sugary) quartz of the third stagebelow (Fig 5b) Usually it forms black millimeter-scale bands of opaline or microcrystalline chal-cedony-adularia and electrum (Figs 5cndashf) Locallyopaline silica crosscuts the microcrystallinequartz (Figs 5cndashd) of the early stage and in turn iscut by the coarser grained quartz and carbonatesof the third and forth stages (Figs 5dndashe) Petro-graphic examination of the electrum-rich bandsreveals that electrum commonly occurs towardsthe bottom of the opaline or microcrystallinebands (Figs 5c e) Where opaline bands are notpresent electrum precipitates at the boundary offirst and third stages (Fig 5b) The electrum hasirregular forms as the result of crystallization inthe quartz interstices being relatively well-shapedin the coarser-grained quartz crystals The size ofelectrum grains ranges from less than 1 to 20 micromHowever the single crystals form chains reachingup to 650 microm (Figs 5b 6a) Electrum in gold-richbands etched with HF is characterized by botryoi-dal surfaces that appear to have formed by theaggregation of spherical electrum particles (Figs6andashb) as observed in the Sleeper deposit Nevada(Saunders 1990 1994) Electrum in bonanza veinscan occupy 50 vol (Figs 5f 6c) accompanied byrare As-bearing pyrite Occasionally micron orsubmicron size electrum inclusions form zoneswithin the As-bearing pyrite (Figs 6d) Tiny AundashAg tellurides have been observed at the contactsbetween electrum grains and small galena crystals(Fig 6f)

The third stage starts with deposition of sugaryquartz (up to 15 mm large Figs 5d 7a) followedby bands of quartz and bladed calcite (Figs 5b 7b)and finally by massive calcite with small amountof quartz (Fig 5d) Locally calcite fills quartz ge-odes or forms small veins (Fig 7a) Microscopical-ly the late calcite corrodes and replaces quartzcrystals Small amount of adularia and pyrite arealso typical for this stage

The hydrothermal mineralization ended withthe deposition of pinkish ankerite-dolomite andsiderite which was preceded by precipitation ofmassive pyrite in some parts of the deposit (Figs7cndashd)

The mineralization of Ada Tepe exhibits sever-al important spatial and textural features thatprovide insights to the physical environment ofore deposition In its upper part quartz becomesthe predominant gangue mineral whereas car-bonates are subordinate This is particularly obvi-ous in the last stage where in the surface samplesdolomite and ankerite are absent in the geodes

and bladed minerals consist of microcrystallinequartz only Another spatial variation is the di-minishing of banding textures with the decreasein grade of Au mineralization to the north Animportant feature is the absence of bladed car-bonates in the veinlets beneath The Wall in con-trast to the entire extent of the deposit above thedetachment surface

53 Oxidized zone

The uppermost part of the Ada Tepe deposit hasbeen oxidized by supergene processes with thedepth of oxidation being dependent upon inten-sity of faulting Due to extensive silicification TheWall itself is generally not oxidized however a 2to 5 m thick zone of oxidation invariably occursabove The Wall Narrow zones of oxidation occurdown to this zone utilizing high angle listric faultsThese faults and fractures most likely acted asconduits for the surface waters to reach the im-permeable silicified zone forming the oxidizedzone above The Wall

Most of the pyrite in the oxidized zones hasbeen converted into limonite identified asgoethite Kaolinite is widely distributed in thiszone forming locally small veinlets Electrum inthe oxidized zone however shows no evidence ofAu enrichment

6 Alteration

In the poorly mineralized zones hydrothermal al-teration is weak with most of the detrital minerals(quartz muscovite and feldspars) remaining unal-tered Alteration consists of chlorite calcite kaoli-nite and subordinate albite pyrite and ankerite-dolomite It seems that calcite and ankerite-dolo-mite are late however it is difficult to distinguishreplacement calcite from vein calcite Quartzadularia calcite pyrite and dolomite-ankerite-siderite plusmn sericite and clay minerals in variableproportions occur in The Wall and immediateproximity to the subvertical quartz and carbonateveins in the strongly mineralized zones Theyformed through replacement of the original meta-morphic minerals included in the sandstone andconglomerate and are associated with relict de-trital quartz and muscovite This complex mineralassemblage is the result of the evolution of thehydrothermal activity

Hydrothermal alteration of the basementmetamorphic rocks is developed within a haloseveral meters in extent beneath The Wall and ismore pronounced around the subvertical veinsThe alteration assemblage includes quartz adu-

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

P Marchev et al62

Fig 3 (a) 3D view of Ada Tepe deposit (b) Cross section of Ada Tepe deposit Major ore bodies (c) Massivetabular ore body (ldquoThe Wallrdquo) above the detachment fault (d) Open space-filling ores along predominantly EndashWoriented listric faults

63Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

occurrence of the adularia-sericite type (Kunovet al 2001)

Since June 2000 the Ada Tepe deposit hasbeen included in the Krumovgrad license areaand is explored by Balkan Mineral and MiningAD Based on the results of regional mapping andsurface sampling BMM conducted two drillingprograms with 145 drill holes using both diamondand reverse circulation drilling techniques for atotal length of 12 440 m A total of 10 218 m ofsurface channel samples were also collected dur-ing 2001ndash2002 Resource calculations by RSGGlobal of Perth Australia show that Ada Tepe isa high-grade AundashAg deposit containing 615 Mt ofore at 46 gt Au for 910 000 ounces of Au using a10 gt cutoff grade

On the basis of initial results from the explora-tion program and the geographical proximity ofthe Ada Tepe deposit to the base metal mineraliza-tion at the Oligocene Zvezdel volcano Crummy(2002) suggested a direct genetic link between thebase and precious metal districts Marchev et al(2003) favoured an Upper Eocene age based onprecise 40Ar39Ar data and a detachment faultingmodel for the origin of the Ada Tepe deposit

3 Geologic setting

31 District geology

Basement metamorphic rocks in the Krumovgradarea build up the elongated Kessebir dome in NndashNE direction (Kozhoukharov et al 1988 Ricou etal 1998 Bonev 2002) (Figs 1 2) Two major tec-tonostratigraphic units (complexes) have beenrecognized on the basis of composition and tec-tonic setting of the metamorphic rocks a Gneiss-migmatite complex and a Variegated Complex(Kozhoukharov et al 1988 Haydoutov et al2001) which correspond to the Continental andMixed units of Ricou et al (1998) respectivelyThe structurally lower Gneiss-migmatite complexcrops out in the core of the Kessebir metamorphicdome It is dominated by igneous protoliths in-cluding metagranites migmatites and migmatizedgneisses overlain by a series of pelitic gneissesand rare amphibolites Eclogites and eclogitic am-phibolites have been described in the Greek partof the Kessebir dome (Mposkos and Krohe2000) RbndashSr ages of 3346 plusmn 34 Ma (Mposkosand Wawrzenitz 1995) from metapegmatites inGreece and RbndashSr (328 plusmn 25 Ma) and UndashPb zir-con ages of 310 plusmn 55 Ma and 319 plusmn 9 Ma (Pey-tcheva et al 1995 1998) from metagranites of theKessebir dome in Bulgaria indicate that theGneiss-migmatite complex is formed of Variscanor older continental basement

The overlaying Variegated complex consists ofa heterogeneous assemblage of pelitic schists para-gneisses amphibolites marbles and ophiolitebodies (Kozhoukharova 1984 Kolcheva and Es-kenazi 1998) Metamorphosed ophiolitic peridot-ites and amphibolitized eclogites are intruded bygabbros gabbronorites plagiogranites and dio-rites UndashPb zircon dating of a gabbro from theneighbouring Biala Reka dome (Fig 1) yields aLate Neoproterozoic age of 572 plusmn 5 Ma for thecore and outer zone recording a Hercynian meta-morphic event at ~300ndash350 Ma (Carrigan et al2003) Volumetrically minor plutonic bodies ofpresumably Upper Cretaceous age intrude theVariegated complex and represent an importantphase of igneous activity in the area (Belmustako-va et al 1995) The rocks are medium to fine-grained light grey two-mica granites of massiveto slightly foliated structure

The rocks of the Variegated complex in theKrumovgrad area are overlain by the Maastrich-tianndashPaleocene syndetachment Shavarovo For-mation (Goranov and Atanasov 1992 see be-low) which is in turn overlain by Upper EocenendashLower Oligocene coal-bearing-sandstone andmarl-limestone formations Lava flows anddomes of the ~35 Ma andesites (Lilov et al1987) of the Iran Tepe volcano are exposednortheast of Krumovgrad (Fig 1) This magmaticactivity was followed by scarce dykes of latitic torhyolitic composition in the northern part of theKessebir dome dated at 3182 plusmn 020 Ma and fi-nally by 26ndash28 Ma old intra-plate basaltic mag-matism in the southern part of the dome(Marchev et al 1997 1998)

32 Deposit geology

The Ada Tepe deposit is hosted by the syndetach-ment MaastrichtianndashPaleocene Shavarovo For-mation overlaying the northeastern closure ofthe Kessebir metamorphic core complex (Fig 2)The contact between poorly consolidated sedi-mentary rocks and underlying metamorphicrocks corresponds to the location of a regionallydeveloped low-angle normal fault named theTockachka detachment fault (TDF) by Bonev(1996) The fault dips 10ndash15deg to the north-north-east It can be traced further southwest for morethan 40 km from Krumovgrad to the Bulgarian-Greek frontier and is well exposed in the areasouthwest of Synap village about 2 km west ofAda Tepe In the latter area the rocks in the hang-ing wall of the TDF consist of metamorphicblocks breccia conglomerates sandstone marlsand argillaceous limestone This unit was first rec-ognized by Atanasov and Goranov (1984) in the

P Marchev et al64

Fig 4 Mineralization of The Wall showing multiple periods of silicification and veining (a) Brecciated early micro-crystalline silica (early stage) with a single gneiss fragment (b) Breccia of the microcrystalline silica (early stage)cemented by calcite (third stage) (c) Subvertical colloform finely banded vein Early stage gray (g) microcrystallinesilica on the margins of the veins followed by the massive microcrystalline (mq) cross-cut by the third stage sugaryquartz and massive and bladed calcite (qc) Late veinlets of dolomite-ankerite (da) cut early vein mineralization(d) Transition of subhorizontal opaline carbonate vein into subvertical step-like vein in the overlying sediments

65Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

Krumovgrad area and described as the ShavarovoFormation of the Krumovgrad Group in a laterpaper (Goranov and Atanasov 1992) The totalthickness of the type section of the ShavarovoFormation (~350 m) was measured along the Kal-djic river near the Shavar village Blocks andclasts of breccia and conglomerate are mixtures ofgneiss amphibolite and marble derived primarilyfrom the Variegated complex

4 Structure

The dominant structure in the Ada Tepe prospectis the TDF It is a mapable tectonic contact bestexpressed along the western slope of Ada Tepehill (Fig 3a) In outcrop and drill cores the fault isrecognized as a contact between Paleozoic meta-morphic rocks and MaastrichtianndashPaleocene sed-imentary rocks of the Shavarovo Formation Im-mediately above the detachment fault there is a05 to 20 m thick zone of massive silicification Indrill core the lower limit of the structure ismarked by a 10ndash20 cm-thick layer of tectonic clayGneiss migmatite marble and the amphibolite ofthe metamorphic basement beneath the detach-ment surface are transformed into a cataclasticrock over a thickness of several meters Thesouthwestern slope of the Ada Tepe hill is cut by aNWndashSE-oriented high-angle fault (Fig 2) whichformed a small half graben fill by older Maas-trichtianndashPaleocene syndetachment sedimentaryrocks overlain in the eastern part by younger Pria-bonianndashOligocene sedimentary rocks

Major structures cutting the basement rocks ofthe Ada Tepe deposit are rare although severalEndashW and NndashS-oriented subvertical faults markedby a zone of silicification occur to the south eastand west of the Ada Tepe deposit The longest NndashS-striking zone has been traced for almost 200 mThe EndashW and NndashS faults coincide with the majordirections of the mineralized veins in the sedi-mentary cover and can be interpreted as feederstructures for the mineralizing fluids

5 Mineralization

51 Ore geometry and ore bodies

The Ada Tepe deposit is 600 m along strike and300ndash350 m wide Gold mineralization occurs as(1) a massive tabular body located immediatelyabove the detachment fault (Figs 3b c) and (2)open space-filling within the breccia-conglomer-ate and sandstone along predominantly east-westoriented subvertical listric faults within the hang-

ing wall (Figs 3b d) In cross section the tabularbody dips (10ndash15deg) to the north-northeast whichis consistent with the low angle TDF The orebody is so strongly silicified company geologistsrefer to it as ldquoThe Wallrdquo Detailed observations onThe Wall indicate a complicated sequence ofevents with multiple periods of silicification andveining Silicification began with the deposition ofmassive white to light grey silica above the de-tachment fault partly or totally replacing the orig-inal rocks Original rock textures were destroyedwith the exception of ghosted gneiss fragments(Fig 4a) The latter are rounded with veins ofquartz or carbonate cutting or surrounding theclasts Early microcrystalline silica was commonlyveined by subvertical banded veins or brecciatedby later fault movements and cemented by calcitesugary quartz (Fig 5b) or by pyrite

In the less silicified parts The Wall consists ofnumerous differently oriented veins ranging inwidth from less than 1 cm up to 30 cm (Fig 4c)Subvertical veins which were formed by deposi-tion of colloform silica layers in open space canbe traced from The Wall into the listric faults inthe overlying sedimentary rocks (Fig 4d) On thesurface they form EndashW-trending ridges separatedby argillic zones (Figs 3a b d)

52 Ore and gangue mineralogy andstages of mineralization

The ore mineralogy of the Ada Tepe deposit is sim-ple consisting mainly of electrum and subordinatepyrite with traces of galena and gold-silver tellu-rides Gangue minerals consist of silica polymorphs(microcrystalline fine-grained sugary quartz andopaline silica) various carbonates (calcite dolo-mite ankerite and siderite) and adularia Telluridesinclude hessite (Ag2Te) and petzite (Ag3AuTe2) Allsamples with Au content higher than 1 gt have al-most constant AuAg ratios ~ 3 reflecting the com-position of electrum (76ndash73 wt Au)

Unlike the massive silicification of The Wallbanded veins in and above it provide an excellentframework for identifying the different stages ofmineralization Mineralization in Ada Tepe canbe subdivided into 4 broad generally overlappingstages of quartz and carbonate veins on the basisof cross-cutting relationships changes in mineral-ogy texture and gold grades

The earliest stage is characterized by deposi-tion of microcrystalline to fine-grained grayish towhite massive or banded quartz with rare pyriteand adularia (Figs 5andashd) In some veins silica depo-sition starts with a thin band of almost isotropicopaline (Fig 4 c) followed by thin band of quartzwith rare bladed calcite Breccia texture is typical

P Marchev et al66

for the massive silicification of The Wall Theproducts of the early stage form microfracturefilling in the host rocks near the margins of the

Fig 5 Microphotographs and SEM images of silica and bonanza ore textures (a) Bands of early stage microcrystal-line (mq) and jigsaw quartz (jq) (b) Electrum band (e) deposited on the interface between early microcrystallinequartz (mq) and third stage sugary quartz and radiating bladed calcite (qc) (c) Band of electrum (e) and isotropicopaline Electrum is located on the bottom of the band (d) Third stage coarse crystalline quartz (ccq) cutting theopaline (o) and microcrystalline quartz (mq) which in turn is cut by calcite (c) (e) Subparallel bands of electrum (e)and microcrystalline adularia and silica separated by a calcite vein (c) The arrow shows the vertical direction (f)SEM image of the outlined area

veins or rock fragments in The Wall causing thesilica-adularia-pyrite replacement in them Thegold grade of this stage is unknown

67Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

The economic gold is related to the secondstage of mineralization It starts at the end of themicrocrystalline quartz and the beginning of thecoarser grained (sugary) quartz of the third stagebelow (Fig 5b) Usually it forms black millimeter-scale bands of opaline or microcrystalline chal-cedony-adularia and electrum (Figs 5cndashf) Locallyopaline silica crosscuts the microcrystallinequartz (Figs 5cndashd) of the early stage and in turn iscut by the coarser grained quartz and carbonatesof the third and forth stages (Figs 5dndashe) Petro-graphic examination of the electrum-rich bandsreveals that electrum commonly occurs towardsthe bottom of the opaline or microcrystallinebands (Figs 5c e) Where opaline bands are notpresent electrum precipitates at the boundary offirst and third stages (Fig 5b) The electrum hasirregular forms as the result of crystallization inthe quartz interstices being relatively well-shapedin the coarser-grained quartz crystals The size ofelectrum grains ranges from less than 1 to 20 micromHowever the single crystals form chains reachingup to 650 microm (Figs 5b 6a) Electrum in gold-richbands etched with HF is characterized by botryoi-dal surfaces that appear to have formed by theaggregation of spherical electrum particles (Figs6andashb) as observed in the Sleeper deposit Nevada(Saunders 1990 1994) Electrum in bonanza veinscan occupy 50 vol (Figs 5f 6c) accompanied byrare As-bearing pyrite Occasionally micron orsubmicron size electrum inclusions form zoneswithin the As-bearing pyrite (Figs 6d) Tiny AundashAg tellurides have been observed at the contactsbetween electrum grains and small galena crystals(Fig 6f)

The third stage starts with deposition of sugaryquartz (up to 15 mm large Figs 5d 7a) followedby bands of quartz and bladed calcite (Figs 5b 7b)and finally by massive calcite with small amountof quartz (Fig 5d) Locally calcite fills quartz ge-odes or forms small veins (Fig 7a) Microscopical-ly the late calcite corrodes and replaces quartzcrystals Small amount of adularia and pyrite arealso typical for this stage

The hydrothermal mineralization ended withthe deposition of pinkish ankerite-dolomite andsiderite which was preceded by precipitation ofmassive pyrite in some parts of the deposit (Figs7cndashd)

The mineralization of Ada Tepe exhibits sever-al important spatial and textural features thatprovide insights to the physical environment ofore deposition In its upper part quartz becomesthe predominant gangue mineral whereas car-bonates are subordinate This is particularly obvi-ous in the last stage where in the surface samplesdolomite and ankerite are absent in the geodes

and bladed minerals consist of microcrystallinequartz only Another spatial variation is the di-minishing of banding textures with the decreasein grade of Au mineralization to the north Animportant feature is the absence of bladed car-bonates in the veinlets beneath The Wall in con-trast to the entire extent of the deposit above thedetachment surface

53 Oxidized zone

The uppermost part of the Ada Tepe deposit hasbeen oxidized by supergene processes with thedepth of oxidation being dependent upon inten-sity of faulting Due to extensive silicification TheWall itself is generally not oxidized however a 2to 5 m thick zone of oxidation invariably occursabove The Wall Narrow zones of oxidation occurdown to this zone utilizing high angle listric faultsThese faults and fractures most likely acted asconduits for the surface waters to reach the im-permeable silicified zone forming the oxidizedzone above The Wall

Most of the pyrite in the oxidized zones hasbeen converted into limonite identified asgoethite Kaolinite is widely distributed in thiszone forming locally small veinlets Electrum inthe oxidized zone however shows no evidence ofAu enrichment

6 Alteration

In the poorly mineralized zones hydrothermal al-teration is weak with most of the detrital minerals(quartz muscovite and feldspars) remaining unal-tered Alteration consists of chlorite calcite kaoli-nite and subordinate albite pyrite and ankerite-dolomite It seems that calcite and ankerite-dolo-mite are late however it is difficult to distinguishreplacement calcite from vein calcite Quartzadularia calcite pyrite and dolomite-ankerite-siderite plusmn sericite and clay minerals in variableproportions occur in The Wall and immediateproximity to the subvertical quartz and carbonateveins in the strongly mineralized zones Theyformed through replacement of the original meta-morphic minerals included in the sandstone andconglomerate and are associated with relict de-trital quartz and muscovite This complex mineralassemblage is the result of the evolution of thehydrothermal activity

Hydrothermal alteration of the basementmetamorphic rocks is developed within a haloseveral meters in extent beneath The Wall and ismore pronounced around the subvertical veinsThe alteration assemblage includes quartz adu-

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

63Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

occurrence of the adularia-sericite type (Kunovet al 2001)

Since June 2000 the Ada Tepe deposit hasbeen included in the Krumovgrad license areaand is explored by Balkan Mineral and MiningAD Based on the results of regional mapping andsurface sampling BMM conducted two drillingprograms with 145 drill holes using both diamondand reverse circulation drilling techniques for atotal length of 12 440 m A total of 10 218 m ofsurface channel samples were also collected dur-ing 2001ndash2002 Resource calculations by RSGGlobal of Perth Australia show that Ada Tepe isa high-grade AundashAg deposit containing 615 Mt ofore at 46 gt Au for 910 000 ounces of Au using a10 gt cutoff grade

On the basis of initial results from the explora-tion program and the geographical proximity ofthe Ada Tepe deposit to the base metal mineraliza-tion at the Oligocene Zvezdel volcano Crummy(2002) suggested a direct genetic link between thebase and precious metal districts Marchev et al(2003) favoured an Upper Eocene age based onprecise 40Ar39Ar data and a detachment faultingmodel for the origin of the Ada Tepe deposit

3 Geologic setting

31 District geology

Basement metamorphic rocks in the Krumovgradarea build up the elongated Kessebir dome in NndashNE direction (Kozhoukharov et al 1988 Ricou etal 1998 Bonev 2002) (Figs 1 2) Two major tec-tonostratigraphic units (complexes) have beenrecognized on the basis of composition and tec-tonic setting of the metamorphic rocks a Gneiss-migmatite complex and a Variegated Complex(Kozhoukharov et al 1988 Haydoutov et al2001) which correspond to the Continental andMixed units of Ricou et al (1998) respectivelyThe structurally lower Gneiss-migmatite complexcrops out in the core of the Kessebir metamorphicdome It is dominated by igneous protoliths in-cluding metagranites migmatites and migmatizedgneisses overlain by a series of pelitic gneissesand rare amphibolites Eclogites and eclogitic am-phibolites have been described in the Greek partof the Kessebir dome (Mposkos and Krohe2000) RbndashSr ages of 3346 plusmn 34 Ma (Mposkosand Wawrzenitz 1995) from metapegmatites inGreece and RbndashSr (328 plusmn 25 Ma) and UndashPb zir-con ages of 310 plusmn 55 Ma and 319 plusmn 9 Ma (Pey-tcheva et al 1995 1998) from metagranites of theKessebir dome in Bulgaria indicate that theGneiss-migmatite complex is formed of Variscanor older continental basement

The overlaying Variegated complex consists ofa heterogeneous assemblage of pelitic schists para-gneisses amphibolites marbles and ophiolitebodies (Kozhoukharova 1984 Kolcheva and Es-kenazi 1998) Metamorphosed ophiolitic peridot-ites and amphibolitized eclogites are intruded bygabbros gabbronorites plagiogranites and dio-rites UndashPb zircon dating of a gabbro from theneighbouring Biala Reka dome (Fig 1) yields aLate Neoproterozoic age of 572 plusmn 5 Ma for thecore and outer zone recording a Hercynian meta-morphic event at ~300ndash350 Ma (Carrigan et al2003) Volumetrically minor plutonic bodies ofpresumably Upper Cretaceous age intrude theVariegated complex and represent an importantphase of igneous activity in the area (Belmustako-va et al 1995) The rocks are medium to fine-grained light grey two-mica granites of massiveto slightly foliated structure

The rocks of the Variegated complex in theKrumovgrad area are overlain by the Maastrich-tianndashPaleocene syndetachment Shavarovo For-mation (Goranov and Atanasov 1992 see be-low) which is in turn overlain by Upper EocenendashLower Oligocene coal-bearing-sandstone andmarl-limestone formations Lava flows anddomes of the ~35 Ma andesites (Lilov et al1987) of the Iran Tepe volcano are exposednortheast of Krumovgrad (Fig 1) This magmaticactivity was followed by scarce dykes of latitic torhyolitic composition in the northern part of theKessebir dome dated at 3182 plusmn 020 Ma and fi-nally by 26ndash28 Ma old intra-plate basaltic mag-matism in the southern part of the dome(Marchev et al 1997 1998)

32 Deposit geology

The Ada Tepe deposit is hosted by the syndetach-ment MaastrichtianndashPaleocene Shavarovo For-mation overlaying the northeastern closure ofthe Kessebir metamorphic core complex (Fig 2)The contact between poorly consolidated sedi-mentary rocks and underlying metamorphicrocks corresponds to the location of a regionallydeveloped low-angle normal fault named theTockachka detachment fault (TDF) by Bonev(1996) The fault dips 10ndash15deg to the north-north-east It can be traced further southwest for morethan 40 km from Krumovgrad to the Bulgarian-Greek frontier and is well exposed in the areasouthwest of Synap village about 2 km west ofAda Tepe In the latter area the rocks in the hang-ing wall of the TDF consist of metamorphicblocks breccia conglomerates sandstone marlsand argillaceous limestone This unit was first rec-ognized by Atanasov and Goranov (1984) in the

P Marchev et al64

Fig 4 Mineralization of The Wall showing multiple periods of silicification and veining (a) Brecciated early micro-crystalline silica (early stage) with a single gneiss fragment (b) Breccia of the microcrystalline silica (early stage)cemented by calcite (third stage) (c) Subvertical colloform finely banded vein Early stage gray (g) microcrystallinesilica on the margins of the veins followed by the massive microcrystalline (mq) cross-cut by the third stage sugaryquartz and massive and bladed calcite (qc) Late veinlets of dolomite-ankerite (da) cut early vein mineralization(d) Transition of subhorizontal opaline carbonate vein into subvertical step-like vein in the overlying sediments

65Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

Krumovgrad area and described as the ShavarovoFormation of the Krumovgrad Group in a laterpaper (Goranov and Atanasov 1992) The totalthickness of the type section of the ShavarovoFormation (~350 m) was measured along the Kal-djic river near the Shavar village Blocks andclasts of breccia and conglomerate are mixtures ofgneiss amphibolite and marble derived primarilyfrom the Variegated complex

4 Structure

The dominant structure in the Ada Tepe prospectis the TDF It is a mapable tectonic contact bestexpressed along the western slope of Ada Tepehill (Fig 3a) In outcrop and drill cores the fault isrecognized as a contact between Paleozoic meta-morphic rocks and MaastrichtianndashPaleocene sed-imentary rocks of the Shavarovo Formation Im-mediately above the detachment fault there is a05 to 20 m thick zone of massive silicification Indrill core the lower limit of the structure ismarked by a 10ndash20 cm-thick layer of tectonic clayGneiss migmatite marble and the amphibolite ofthe metamorphic basement beneath the detach-ment surface are transformed into a cataclasticrock over a thickness of several meters Thesouthwestern slope of the Ada Tepe hill is cut by aNWndashSE-oriented high-angle fault (Fig 2) whichformed a small half graben fill by older Maas-trichtianndashPaleocene syndetachment sedimentaryrocks overlain in the eastern part by younger Pria-bonianndashOligocene sedimentary rocks

Major structures cutting the basement rocks ofthe Ada Tepe deposit are rare although severalEndashW and NndashS-oriented subvertical faults markedby a zone of silicification occur to the south eastand west of the Ada Tepe deposit The longest NndashS-striking zone has been traced for almost 200 mThe EndashW and NndashS faults coincide with the majordirections of the mineralized veins in the sedi-mentary cover and can be interpreted as feederstructures for the mineralizing fluids

5 Mineralization

51 Ore geometry and ore bodies

The Ada Tepe deposit is 600 m along strike and300ndash350 m wide Gold mineralization occurs as(1) a massive tabular body located immediatelyabove the detachment fault (Figs 3b c) and (2)open space-filling within the breccia-conglomer-ate and sandstone along predominantly east-westoriented subvertical listric faults within the hang-

ing wall (Figs 3b d) In cross section the tabularbody dips (10ndash15deg) to the north-northeast whichis consistent with the low angle TDF The orebody is so strongly silicified company geologistsrefer to it as ldquoThe Wallrdquo Detailed observations onThe Wall indicate a complicated sequence ofevents with multiple periods of silicification andveining Silicification began with the deposition ofmassive white to light grey silica above the de-tachment fault partly or totally replacing the orig-inal rocks Original rock textures were destroyedwith the exception of ghosted gneiss fragments(Fig 4a) The latter are rounded with veins ofquartz or carbonate cutting or surrounding theclasts Early microcrystalline silica was commonlyveined by subvertical banded veins or brecciatedby later fault movements and cemented by calcitesugary quartz (Fig 5b) or by pyrite

In the less silicified parts The Wall consists ofnumerous differently oriented veins ranging inwidth from less than 1 cm up to 30 cm (Fig 4c)Subvertical veins which were formed by deposi-tion of colloform silica layers in open space canbe traced from The Wall into the listric faults inthe overlying sedimentary rocks (Fig 4d) On thesurface they form EndashW-trending ridges separatedby argillic zones (Figs 3a b d)

52 Ore and gangue mineralogy andstages of mineralization

The ore mineralogy of the Ada Tepe deposit is sim-ple consisting mainly of electrum and subordinatepyrite with traces of galena and gold-silver tellu-rides Gangue minerals consist of silica polymorphs(microcrystalline fine-grained sugary quartz andopaline silica) various carbonates (calcite dolo-mite ankerite and siderite) and adularia Telluridesinclude hessite (Ag2Te) and petzite (Ag3AuTe2) Allsamples with Au content higher than 1 gt have al-most constant AuAg ratios ~ 3 reflecting the com-position of electrum (76ndash73 wt Au)

Unlike the massive silicification of The Wallbanded veins in and above it provide an excellentframework for identifying the different stages ofmineralization Mineralization in Ada Tepe canbe subdivided into 4 broad generally overlappingstages of quartz and carbonate veins on the basisof cross-cutting relationships changes in mineral-ogy texture and gold grades

The earliest stage is characterized by deposi-tion of microcrystalline to fine-grained grayish towhite massive or banded quartz with rare pyriteand adularia (Figs 5andashd) In some veins silica depo-sition starts with a thin band of almost isotropicopaline (Fig 4 c) followed by thin band of quartzwith rare bladed calcite Breccia texture is typical

P Marchev et al66

for the massive silicification of The Wall Theproducts of the early stage form microfracturefilling in the host rocks near the margins of the

Fig 5 Microphotographs and SEM images of silica and bonanza ore textures (a) Bands of early stage microcrystal-line (mq) and jigsaw quartz (jq) (b) Electrum band (e) deposited on the interface between early microcrystallinequartz (mq) and third stage sugary quartz and radiating bladed calcite (qc) (c) Band of electrum (e) and isotropicopaline Electrum is located on the bottom of the band (d) Third stage coarse crystalline quartz (ccq) cutting theopaline (o) and microcrystalline quartz (mq) which in turn is cut by calcite (c) (e) Subparallel bands of electrum (e)and microcrystalline adularia and silica separated by a calcite vein (c) The arrow shows the vertical direction (f)SEM image of the outlined area

veins or rock fragments in The Wall causing thesilica-adularia-pyrite replacement in them Thegold grade of this stage is unknown

67Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

The economic gold is related to the secondstage of mineralization It starts at the end of themicrocrystalline quartz and the beginning of thecoarser grained (sugary) quartz of the third stagebelow (Fig 5b) Usually it forms black millimeter-scale bands of opaline or microcrystalline chal-cedony-adularia and electrum (Figs 5cndashf) Locallyopaline silica crosscuts the microcrystallinequartz (Figs 5cndashd) of the early stage and in turn iscut by the coarser grained quartz and carbonatesof the third and forth stages (Figs 5dndashe) Petro-graphic examination of the electrum-rich bandsreveals that electrum commonly occurs towardsthe bottom of the opaline or microcrystallinebands (Figs 5c e) Where opaline bands are notpresent electrum precipitates at the boundary offirst and third stages (Fig 5b) The electrum hasirregular forms as the result of crystallization inthe quartz interstices being relatively well-shapedin the coarser-grained quartz crystals The size ofelectrum grains ranges from less than 1 to 20 micromHowever the single crystals form chains reachingup to 650 microm (Figs 5b 6a) Electrum in gold-richbands etched with HF is characterized by botryoi-dal surfaces that appear to have formed by theaggregation of spherical electrum particles (Figs6andashb) as observed in the Sleeper deposit Nevada(Saunders 1990 1994) Electrum in bonanza veinscan occupy 50 vol (Figs 5f 6c) accompanied byrare As-bearing pyrite Occasionally micron orsubmicron size electrum inclusions form zoneswithin the As-bearing pyrite (Figs 6d) Tiny AundashAg tellurides have been observed at the contactsbetween electrum grains and small galena crystals(Fig 6f)

The third stage starts with deposition of sugaryquartz (up to 15 mm large Figs 5d 7a) followedby bands of quartz and bladed calcite (Figs 5b 7b)and finally by massive calcite with small amountof quartz (Fig 5d) Locally calcite fills quartz ge-odes or forms small veins (Fig 7a) Microscopical-ly the late calcite corrodes and replaces quartzcrystals Small amount of adularia and pyrite arealso typical for this stage

The hydrothermal mineralization ended withthe deposition of pinkish ankerite-dolomite andsiderite which was preceded by precipitation ofmassive pyrite in some parts of the deposit (Figs7cndashd)

The mineralization of Ada Tepe exhibits sever-al important spatial and textural features thatprovide insights to the physical environment ofore deposition In its upper part quartz becomesthe predominant gangue mineral whereas car-bonates are subordinate This is particularly obvi-ous in the last stage where in the surface samplesdolomite and ankerite are absent in the geodes

and bladed minerals consist of microcrystallinequartz only Another spatial variation is the di-minishing of banding textures with the decreasein grade of Au mineralization to the north Animportant feature is the absence of bladed car-bonates in the veinlets beneath The Wall in con-trast to the entire extent of the deposit above thedetachment surface

53 Oxidized zone

The uppermost part of the Ada Tepe deposit hasbeen oxidized by supergene processes with thedepth of oxidation being dependent upon inten-sity of faulting Due to extensive silicification TheWall itself is generally not oxidized however a 2to 5 m thick zone of oxidation invariably occursabove The Wall Narrow zones of oxidation occurdown to this zone utilizing high angle listric faultsThese faults and fractures most likely acted asconduits for the surface waters to reach the im-permeable silicified zone forming the oxidizedzone above The Wall

Most of the pyrite in the oxidized zones hasbeen converted into limonite identified asgoethite Kaolinite is widely distributed in thiszone forming locally small veinlets Electrum inthe oxidized zone however shows no evidence ofAu enrichment

6 Alteration

In the poorly mineralized zones hydrothermal al-teration is weak with most of the detrital minerals(quartz muscovite and feldspars) remaining unal-tered Alteration consists of chlorite calcite kaoli-nite and subordinate albite pyrite and ankerite-dolomite It seems that calcite and ankerite-dolo-mite are late however it is difficult to distinguishreplacement calcite from vein calcite Quartzadularia calcite pyrite and dolomite-ankerite-siderite plusmn sericite and clay minerals in variableproportions occur in The Wall and immediateproximity to the subvertical quartz and carbonateveins in the strongly mineralized zones Theyformed through replacement of the original meta-morphic minerals included in the sandstone andconglomerate and are associated with relict de-trital quartz and muscovite This complex mineralassemblage is the result of the evolution of thehydrothermal activity

Hydrothermal alteration of the basementmetamorphic rocks is developed within a haloseveral meters in extent beneath The Wall and ismore pronounced around the subvertical veinsThe alteration assemblage includes quartz adu-

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

P Marchev et al64

Fig 4 Mineralization of The Wall showing multiple periods of silicification and veining (a) Brecciated early micro-crystalline silica (early stage) with a single gneiss fragment (b) Breccia of the microcrystalline silica (early stage)cemented by calcite (third stage) (c) Subvertical colloform finely banded vein Early stage gray (g) microcrystallinesilica on the margins of the veins followed by the massive microcrystalline (mq) cross-cut by the third stage sugaryquartz and massive and bladed calcite (qc) Late veinlets of dolomite-ankerite (da) cut early vein mineralization(d) Transition of subhorizontal opaline carbonate vein into subvertical step-like vein in the overlying sediments

65Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

Krumovgrad area and described as the ShavarovoFormation of the Krumovgrad Group in a laterpaper (Goranov and Atanasov 1992) The totalthickness of the type section of the ShavarovoFormation (~350 m) was measured along the Kal-djic river near the Shavar village Blocks andclasts of breccia and conglomerate are mixtures ofgneiss amphibolite and marble derived primarilyfrom the Variegated complex

4 Structure

The dominant structure in the Ada Tepe prospectis the TDF It is a mapable tectonic contact bestexpressed along the western slope of Ada Tepehill (Fig 3a) In outcrop and drill cores the fault isrecognized as a contact between Paleozoic meta-morphic rocks and MaastrichtianndashPaleocene sed-imentary rocks of the Shavarovo Formation Im-mediately above the detachment fault there is a05 to 20 m thick zone of massive silicification Indrill core the lower limit of the structure ismarked by a 10ndash20 cm-thick layer of tectonic clayGneiss migmatite marble and the amphibolite ofthe metamorphic basement beneath the detach-ment surface are transformed into a cataclasticrock over a thickness of several meters Thesouthwestern slope of the Ada Tepe hill is cut by aNWndashSE-oriented high-angle fault (Fig 2) whichformed a small half graben fill by older Maas-trichtianndashPaleocene syndetachment sedimentaryrocks overlain in the eastern part by younger Pria-bonianndashOligocene sedimentary rocks

Major structures cutting the basement rocks ofthe Ada Tepe deposit are rare although severalEndashW and NndashS-oriented subvertical faults markedby a zone of silicification occur to the south eastand west of the Ada Tepe deposit The longest NndashS-striking zone has been traced for almost 200 mThe EndashW and NndashS faults coincide with the majordirections of the mineralized veins in the sedi-mentary cover and can be interpreted as feederstructures for the mineralizing fluids

5 Mineralization

51 Ore geometry and ore bodies

The Ada Tepe deposit is 600 m along strike and300ndash350 m wide Gold mineralization occurs as(1) a massive tabular body located immediatelyabove the detachment fault (Figs 3b c) and (2)open space-filling within the breccia-conglomer-ate and sandstone along predominantly east-westoriented subvertical listric faults within the hang-

ing wall (Figs 3b d) In cross section the tabularbody dips (10ndash15deg) to the north-northeast whichis consistent with the low angle TDF The orebody is so strongly silicified company geologistsrefer to it as ldquoThe Wallrdquo Detailed observations onThe Wall indicate a complicated sequence ofevents with multiple periods of silicification andveining Silicification began with the deposition ofmassive white to light grey silica above the de-tachment fault partly or totally replacing the orig-inal rocks Original rock textures were destroyedwith the exception of ghosted gneiss fragments(Fig 4a) The latter are rounded with veins ofquartz or carbonate cutting or surrounding theclasts Early microcrystalline silica was commonlyveined by subvertical banded veins or brecciatedby later fault movements and cemented by calcitesugary quartz (Fig 5b) or by pyrite

In the less silicified parts The Wall consists ofnumerous differently oriented veins ranging inwidth from less than 1 cm up to 30 cm (Fig 4c)Subvertical veins which were formed by deposi-tion of colloform silica layers in open space canbe traced from The Wall into the listric faults inthe overlying sedimentary rocks (Fig 4d) On thesurface they form EndashW-trending ridges separatedby argillic zones (Figs 3a b d)

52 Ore and gangue mineralogy andstages of mineralization

The ore mineralogy of the Ada Tepe deposit is sim-ple consisting mainly of electrum and subordinatepyrite with traces of galena and gold-silver tellu-rides Gangue minerals consist of silica polymorphs(microcrystalline fine-grained sugary quartz andopaline silica) various carbonates (calcite dolo-mite ankerite and siderite) and adularia Telluridesinclude hessite (Ag2Te) and petzite (Ag3AuTe2) Allsamples with Au content higher than 1 gt have al-most constant AuAg ratios ~ 3 reflecting the com-position of electrum (76ndash73 wt Au)

Unlike the massive silicification of The Wallbanded veins in and above it provide an excellentframework for identifying the different stages ofmineralization Mineralization in Ada Tepe canbe subdivided into 4 broad generally overlappingstages of quartz and carbonate veins on the basisof cross-cutting relationships changes in mineral-ogy texture and gold grades

The earliest stage is characterized by deposi-tion of microcrystalline to fine-grained grayish towhite massive or banded quartz with rare pyriteand adularia (Figs 5andashd) In some veins silica depo-sition starts with a thin band of almost isotropicopaline (Fig 4 c) followed by thin band of quartzwith rare bladed calcite Breccia texture is typical

P Marchev et al66

for the massive silicification of The Wall Theproducts of the early stage form microfracturefilling in the host rocks near the margins of the

Fig 5 Microphotographs and SEM images of silica and bonanza ore textures (a) Bands of early stage microcrystal-line (mq) and jigsaw quartz (jq) (b) Electrum band (e) deposited on the interface between early microcrystallinequartz (mq) and third stage sugary quartz and radiating bladed calcite (qc) (c) Band of electrum (e) and isotropicopaline Electrum is located on the bottom of the band (d) Third stage coarse crystalline quartz (ccq) cutting theopaline (o) and microcrystalline quartz (mq) which in turn is cut by calcite (c) (e) Subparallel bands of electrum (e)and microcrystalline adularia and silica separated by a calcite vein (c) The arrow shows the vertical direction (f)SEM image of the outlined area

veins or rock fragments in The Wall causing thesilica-adularia-pyrite replacement in them Thegold grade of this stage is unknown

67Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

The economic gold is related to the secondstage of mineralization It starts at the end of themicrocrystalline quartz and the beginning of thecoarser grained (sugary) quartz of the third stagebelow (Fig 5b) Usually it forms black millimeter-scale bands of opaline or microcrystalline chal-cedony-adularia and electrum (Figs 5cndashf) Locallyopaline silica crosscuts the microcrystallinequartz (Figs 5cndashd) of the early stage and in turn iscut by the coarser grained quartz and carbonatesof the third and forth stages (Figs 5dndashe) Petro-graphic examination of the electrum-rich bandsreveals that electrum commonly occurs towardsthe bottom of the opaline or microcrystallinebands (Figs 5c e) Where opaline bands are notpresent electrum precipitates at the boundary offirst and third stages (Fig 5b) The electrum hasirregular forms as the result of crystallization inthe quartz interstices being relatively well-shapedin the coarser-grained quartz crystals The size ofelectrum grains ranges from less than 1 to 20 micromHowever the single crystals form chains reachingup to 650 microm (Figs 5b 6a) Electrum in gold-richbands etched with HF is characterized by botryoi-dal surfaces that appear to have formed by theaggregation of spherical electrum particles (Figs6andashb) as observed in the Sleeper deposit Nevada(Saunders 1990 1994) Electrum in bonanza veinscan occupy 50 vol (Figs 5f 6c) accompanied byrare As-bearing pyrite Occasionally micron orsubmicron size electrum inclusions form zoneswithin the As-bearing pyrite (Figs 6d) Tiny AundashAg tellurides have been observed at the contactsbetween electrum grains and small galena crystals(Fig 6f)

The third stage starts with deposition of sugaryquartz (up to 15 mm large Figs 5d 7a) followedby bands of quartz and bladed calcite (Figs 5b 7b)and finally by massive calcite with small amountof quartz (Fig 5d) Locally calcite fills quartz ge-odes or forms small veins (Fig 7a) Microscopical-ly the late calcite corrodes and replaces quartzcrystals Small amount of adularia and pyrite arealso typical for this stage

The hydrothermal mineralization ended withthe deposition of pinkish ankerite-dolomite andsiderite which was preceded by precipitation ofmassive pyrite in some parts of the deposit (Figs7cndashd)

The mineralization of Ada Tepe exhibits sever-al important spatial and textural features thatprovide insights to the physical environment ofore deposition In its upper part quartz becomesthe predominant gangue mineral whereas car-bonates are subordinate This is particularly obvi-ous in the last stage where in the surface samplesdolomite and ankerite are absent in the geodes

and bladed minerals consist of microcrystallinequartz only Another spatial variation is the di-minishing of banding textures with the decreasein grade of Au mineralization to the north Animportant feature is the absence of bladed car-bonates in the veinlets beneath The Wall in con-trast to the entire extent of the deposit above thedetachment surface

53 Oxidized zone

The uppermost part of the Ada Tepe deposit hasbeen oxidized by supergene processes with thedepth of oxidation being dependent upon inten-sity of faulting Due to extensive silicification TheWall itself is generally not oxidized however a 2to 5 m thick zone of oxidation invariably occursabove The Wall Narrow zones of oxidation occurdown to this zone utilizing high angle listric faultsThese faults and fractures most likely acted asconduits for the surface waters to reach the im-permeable silicified zone forming the oxidizedzone above The Wall

Most of the pyrite in the oxidized zones hasbeen converted into limonite identified asgoethite Kaolinite is widely distributed in thiszone forming locally small veinlets Electrum inthe oxidized zone however shows no evidence ofAu enrichment

6 Alteration

In the poorly mineralized zones hydrothermal al-teration is weak with most of the detrital minerals(quartz muscovite and feldspars) remaining unal-tered Alteration consists of chlorite calcite kaoli-nite and subordinate albite pyrite and ankerite-dolomite It seems that calcite and ankerite-dolo-mite are late however it is difficult to distinguishreplacement calcite from vein calcite Quartzadularia calcite pyrite and dolomite-ankerite-siderite plusmn sericite and clay minerals in variableproportions occur in The Wall and immediateproximity to the subvertical quartz and carbonateveins in the strongly mineralized zones Theyformed through replacement of the original meta-morphic minerals included in the sandstone andconglomerate and are associated with relict de-trital quartz and muscovite This complex mineralassemblage is the result of the evolution of thehydrothermal activity

Hydrothermal alteration of the basementmetamorphic rocks is developed within a haloseveral meters in extent beneath The Wall and ismore pronounced around the subvertical veinsThe alteration assemblage includes quartz adu-

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

65Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

Krumovgrad area and described as the ShavarovoFormation of the Krumovgrad Group in a laterpaper (Goranov and Atanasov 1992) The totalthickness of the type section of the ShavarovoFormation (~350 m) was measured along the Kal-djic river near the Shavar village Blocks andclasts of breccia and conglomerate are mixtures ofgneiss amphibolite and marble derived primarilyfrom the Variegated complex

4 Structure

The dominant structure in the Ada Tepe prospectis the TDF It is a mapable tectonic contact bestexpressed along the western slope of Ada Tepehill (Fig 3a) In outcrop and drill cores the fault isrecognized as a contact between Paleozoic meta-morphic rocks and MaastrichtianndashPaleocene sed-imentary rocks of the Shavarovo Formation Im-mediately above the detachment fault there is a05 to 20 m thick zone of massive silicification Indrill core the lower limit of the structure ismarked by a 10ndash20 cm-thick layer of tectonic clayGneiss migmatite marble and the amphibolite ofthe metamorphic basement beneath the detach-ment surface are transformed into a cataclasticrock over a thickness of several meters Thesouthwestern slope of the Ada Tepe hill is cut by aNWndashSE-oriented high-angle fault (Fig 2) whichformed a small half graben fill by older Maas-trichtianndashPaleocene syndetachment sedimentaryrocks overlain in the eastern part by younger Pria-bonianndashOligocene sedimentary rocks

Major structures cutting the basement rocks ofthe Ada Tepe deposit are rare although severalEndashW and NndashS-oriented subvertical faults markedby a zone of silicification occur to the south eastand west of the Ada Tepe deposit The longest NndashS-striking zone has been traced for almost 200 mThe EndashW and NndashS faults coincide with the majordirections of the mineralized veins in the sedi-mentary cover and can be interpreted as feederstructures for the mineralizing fluids

5 Mineralization

51 Ore geometry and ore bodies

The Ada Tepe deposit is 600 m along strike and300ndash350 m wide Gold mineralization occurs as(1) a massive tabular body located immediatelyabove the detachment fault (Figs 3b c) and (2)open space-filling within the breccia-conglomer-ate and sandstone along predominantly east-westoriented subvertical listric faults within the hang-

ing wall (Figs 3b d) In cross section the tabularbody dips (10ndash15deg) to the north-northeast whichis consistent with the low angle TDF The orebody is so strongly silicified company geologistsrefer to it as ldquoThe Wallrdquo Detailed observations onThe Wall indicate a complicated sequence ofevents with multiple periods of silicification andveining Silicification began with the deposition ofmassive white to light grey silica above the de-tachment fault partly or totally replacing the orig-inal rocks Original rock textures were destroyedwith the exception of ghosted gneiss fragments(Fig 4a) The latter are rounded with veins ofquartz or carbonate cutting or surrounding theclasts Early microcrystalline silica was commonlyveined by subvertical banded veins or brecciatedby later fault movements and cemented by calcitesugary quartz (Fig 5b) or by pyrite

In the less silicified parts The Wall consists ofnumerous differently oriented veins ranging inwidth from less than 1 cm up to 30 cm (Fig 4c)Subvertical veins which were formed by deposi-tion of colloform silica layers in open space canbe traced from The Wall into the listric faults inthe overlying sedimentary rocks (Fig 4d) On thesurface they form EndashW-trending ridges separatedby argillic zones (Figs 3a b d)

52 Ore and gangue mineralogy andstages of mineralization

The ore mineralogy of the Ada Tepe deposit is sim-ple consisting mainly of electrum and subordinatepyrite with traces of galena and gold-silver tellu-rides Gangue minerals consist of silica polymorphs(microcrystalline fine-grained sugary quartz andopaline silica) various carbonates (calcite dolo-mite ankerite and siderite) and adularia Telluridesinclude hessite (Ag2Te) and petzite (Ag3AuTe2) Allsamples with Au content higher than 1 gt have al-most constant AuAg ratios ~ 3 reflecting the com-position of electrum (76ndash73 wt Au)

Unlike the massive silicification of The Wallbanded veins in and above it provide an excellentframework for identifying the different stages ofmineralization Mineralization in Ada Tepe canbe subdivided into 4 broad generally overlappingstages of quartz and carbonate veins on the basisof cross-cutting relationships changes in mineral-ogy texture and gold grades

The earliest stage is characterized by deposi-tion of microcrystalline to fine-grained grayish towhite massive or banded quartz with rare pyriteand adularia (Figs 5andashd) In some veins silica depo-sition starts with a thin band of almost isotropicopaline (Fig 4 c) followed by thin band of quartzwith rare bladed calcite Breccia texture is typical

P Marchev et al66

for the massive silicification of The Wall Theproducts of the early stage form microfracturefilling in the host rocks near the margins of the

Fig 5 Microphotographs and SEM images of silica and bonanza ore textures (a) Bands of early stage microcrystal-line (mq) and jigsaw quartz (jq) (b) Electrum band (e) deposited on the interface between early microcrystallinequartz (mq) and third stage sugary quartz and radiating bladed calcite (qc) (c) Band of electrum (e) and isotropicopaline Electrum is located on the bottom of the band (d) Third stage coarse crystalline quartz (ccq) cutting theopaline (o) and microcrystalline quartz (mq) which in turn is cut by calcite (c) (e) Subparallel bands of electrum (e)and microcrystalline adularia and silica separated by a calcite vein (c) The arrow shows the vertical direction (f)SEM image of the outlined area

veins or rock fragments in The Wall causing thesilica-adularia-pyrite replacement in them Thegold grade of this stage is unknown

67Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

The economic gold is related to the secondstage of mineralization It starts at the end of themicrocrystalline quartz and the beginning of thecoarser grained (sugary) quartz of the third stagebelow (Fig 5b) Usually it forms black millimeter-scale bands of opaline or microcrystalline chal-cedony-adularia and electrum (Figs 5cndashf) Locallyopaline silica crosscuts the microcrystallinequartz (Figs 5cndashd) of the early stage and in turn iscut by the coarser grained quartz and carbonatesof the third and forth stages (Figs 5dndashe) Petro-graphic examination of the electrum-rich bandsreveals that electrum commonly occurs towardsthe bottom of the opaline or microcrystallinebands (Figs 5c e) Where opaline bands are notpresent electrum precipitates at the boundary offirst and third stages (Fig 5b) The electrum hasirregular forms as the result of crystallization inthe quartz interstices being relatively well-shapedin the coarser-grained quartz crystals The size ofelectrum grains ranges from less than 1 to 20 micromHowever the single crystals form chains reachingup to 650 microm (Figs 5b 6a) Electrum in gold-richbands etched with HF is characterized by botryoi-dal surfaces that appear to have formed by theaggregation of spherical electrum particles (Figs6andashb) as observed in the Sleeper deposit Nevada(Saunders 1990 1994) Electrum in bonanza veinscan occupy 50 vol (Figs 5f 6c) accompanied byrare As-bearing pyrite Occasionally micron orsubmicron size electrum inclusions form zoneswithin the As-bearing pyrite (Figs 6d) Tiny AundashAg tellurides have been observed at the contactsbetween electrum grains and small galena crystals(Fig 6f)

The third stage starts with deposition of sugaryquartz (up to 15 mm large Figs 5d 7a) followedby bands of quartz and bladed calcite (Figs 5b 7b)and finally by massive calcite with small amountof quartz (Fig 5d) Locally calcite fills quartz ge-odes or forms small veins (Fig 7a) Microscopical-ly the late calcite corrodes and replaces quartzcrystals Small amount of adularia and pyrite arealso typical for this stage

The hydrothermal mineralization ended withthe deposition of pinkish ankerite-dolomite andsiderite which was preceded by precipitation ofmassive pyrite in some parts of the deposit (Figs7cndashd)

The mineralization of Ada Tepe exhibits sever-al important spatial and textural features thatprovide insights to the physical environment ofore deposition In its upper part quartz becomesthe predominant gangue mineral whereas car-bonates are subordinate This is particularly obvi-ous in the last stage where in the surface samplesdolomite and ankerite are absent in the geodes

and bladed minerals consist of microcrystallinequartz only Another spatial variation is the di-minishing of banding textures with the decreasein grade of Au mineralization to the north Animportant feature is the absence of bladed car-bonates in the veinlets beneath The Wall in con-trast to the entire extent of the deposit above thedetachment surface

53 Oxidized zone

The uppermost part of the Ada Tepe deposit hasbeen oxidized by supergene processes with thedepth of oxidation being dependent upon inten-sity of faulting Due to extensive silicification TheWall itself is generally not oxidized however a 2to 5 m thick zone of oxidation invariably occursabove The Wall Narrow zones of oxidation occurdown to this zone utilizing high angle listric faultsThese faults and fractures most likely acted asconduits for the surface waters to reach the im-permeable silicified zone forming the oxidizedzone above The Wall

Most of the pyrite in the oxidized zones hasbeen converted into limonite identified asgoethite Kaolinite is widely distributed in thiszone forming locally small veinlets Electrum inthe oxidized zone however shows no evidence ofAu enrichment

6 Alteration

In the poorly mineralized zones hydrothermal al-teration is weak with most of the detrital minerals(quartz muscovite and feldspars) remaining unal-tered Alteration consists of chlorite calcite kaoli-nite and subordinate albite pyrite and ankerite-dolomite It seems that calcite and ankerite-dolo-mite are late however it is difficult to distinguishreplacement calcite from vein calcite Quartzadularia calcite pyrite and dolomite-ankerite-siderite plusmn sericite and clay minerals in variableproportions occur in The Wall and immediateproximity to the subvertical quartz and carbonateveins in the strongly mineralized zones Theyformed through replacement of the original meta-morphic minerals included in the sandstone andconglomerate and are associated with relict de-trital quartz and muscovite This complex mineralassemblage is the result of the evolution of thehydrothermal activity

Hydrothermal alteration of the basementmetamorphic rocks is developed within a haloseveral meters in extent beneath The Wall and ismore pronounced around the subvertical veinsThe alteration assemblage includes quartz adu-

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

P Marchev et al66

for the massive silicification of The Wall Theproducts of the early stage form microfracturefilling in the host rocks near the margins of the

Fig 5 Microphotographs and SEM images of silica and bonanza ore textures (a) Bands of early stage microcrystal-line (mq) and jigsaw quartz (jq) (b) Electrum band (e) deposited on the interface between early microcrystallinequartz (mq) and third stage sugary quartz and radiating bladed calcite (qc) (c) Band of electrum (e) and isotropicopaline Electrum is located on the bottom of the band (d) Third stage coarse crystalline quartz (ccq) cutting theopaline (o) and microcrystalline quartz (mq) which in turn is cut by calcite (c) (e) Subparallel bands of electrum (e)and microcrystalline adularia and silica separated by a calcite vein (c) The arrow shows the vertical direction (f)SEM image of the outlined area

veins or rock fragments in The Wall causing thesilica-adularia-pyrite replacement in them Thegold grade of this stage is unknown

67Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

The economic gold is related to the secondstage of mineralization It starts at the end of themicrocrystalline quartz and the beginning of thecoarser grained (sugary) quartz of the third stagebelow (Fig 5b) Usually it forms black millimeter-scale bands of opaline or microcrystalline chal-cedony-adularia and electrum (Figs 5cndashf) Locallyopaline silica crosscuts the microcrystallinequartz (Figs 5cndashd) of the early stage and in turn iscut by the coarser grained quartz and carbonatesof the third and forth stages (Figs 5dndashe) Petro-graphic examination of the electrum-rich bandsreveals that electrum commonly occurs towardsthe bottom of the opaline or microcrystallinebands (Figs 5c e) Where opaline bands are notpresent electrum precipitates at the boundary offirst and third stages (Fig 5b) The electrum hasirregular forms as the result of crystallization inthe quartz interstices being relatively well-shapedin the coarser-grained quartz crystals The size ofelectrum grains ranges from less than 1 to 20 micromHowever the single crystals form chains reachingup to 650 microm (Figs 5b 6a) Electrum in gold-richbands etched with HF is characterized by botryoi-dal surfaces that appear to have formed by theaggregation of spherical electrum particles (Figs6andashb) as observed in the Sleeper deposit Nevada(Saunders 1990 1994) Electrum in bonanza veinscan occupy 50 vol (Figs 5f 6c) accompanied byrare As-bearing pyrite Occasionally micron orsubmicron size electrum inclusions form zoneswithin the As-bearing pyrite (Figs 6d) Tiny AundashAg tellurides have been observed at the contactsbetween electrum grains and small galena crystals(Fig 6f)

The third stage starts with deposition of sugaryquartz (up to 15 mm large Figs 5d 7a) followedby bands of quartz and bladed calcite (Figs 5b 7b)and finally by massive calcite with small amountof quartz (Fig 5d) Locally calcite fills quartz ge-odes or forms small veins (Fig 7a) Microscopical-ly the late calcite corrodes and replaces quartzcrystals Small amount of adularia and pyrite arealso typical for this stage

The hydrothermal mineralization ended withthe deposition of pinkish ankerite-dolomite andsiderite which was preceded by precipitation ofmassive pyrite in some parts of the deposit (Figs7cndashd)

The mineralization of Ada Tepe exhibits sever-al important spatial and textural features thatprovide insights to the physical environment ofore deposition In its upper part quartz becomesthe predominant gangue mineral whereas car-bonates are subordinate This is particularly obvi-ous in the last stage where in the surface samplesdolomite and ankerite are absent in the geodes

and bladed minerals consist of microcrystallinequartz only Another spatial variation is the di-minishing of banding textures with the decreasein grade of Au mineralization to the north Animportant feature is the absence of bladed car-bonates in the veinlets beneath The Wall in con-trast to the entire extent of the deposit above thedetachment surface

53 Oxidized zone

The uppermost part of the Ada Tepe deposit hasbeen oxidized by supergene processes with thedepth of oxidation being dependent upon inten-sity of faulting Due to extensive silicification TheWall itself is generally not oxidized however a 2to 5 m thick zone of oxidation invariably occursabove The Wall Narrow zones of oxidation occurdown to this zone utilizing high angle listric faultsThese faults and fractures most likely acted asconduits for the surface waters to reach the im-permeable silicified zone forming the oxidizedzone above The Wall

Most of the pyrite in the oxidized zones hasbeen converted into limonite identified asgoethite Kaolinite is widely distributed in thiszone forming locally small veinlets Electrum inthe oxidized zone however shows no evidence ofAu enrichment

6 Alteration

In the poorly mineralized zones hydrothermal al-teration is weak with most of the detrital minerals(quartz muscovite and feldspars) remaining unal-tered Alteration consists of chlorite calcite kaoli-nite and subordinate albite pyrite and ankerite-dolomite It seems that calcite and ankerite-dolo-mite are late however it is difficult to distinguishreplacement calcite from vein calcite Quartzadularia calcite pyrite and dolomite-ankerite-siderite plusmn sericite and clay minerals in variableproportions occur in The Wall and immediateproximity to the subvertical quartz and carbonateveins in the strongly mineralized zones Theyformed through replacement of the original meta-morphic minerals included in the sandstone andconglomerate and are associated with relict de-trital quartz and muscovite This complex mineralassemblage is the result of the evolution of thehydrothermal activity

Hydrothermal alteration of the basementmetamorphic rocks is developed within a haloseveral meters in extent beneath The Wall and ismore pronounced around the subvertical veinsThe alteration assemblage includes quartz adu-

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

67Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

The economic gold is related to the secondstage of mineralization It starts at the end of themicrocrystalline quartz and the beginning of thecoarser grained (sugary) quartz of the third stagebelow (Fig 5b) Usually it forms black millimeter-scale bands of opaline or microcrystalline chal-cedony-adularia and electrum (Figs 5cndashf) Locallyopaline silica crosscuts the microcrystallinequartz (Figs 5cndashd) of the early stage and in turn iscut by the coarser grained quartz and carbonatesof the third and forth stages (Figs 5dndashe) Petro-graphic examination of the electrum-rich bandsreveals that electrum commonly occurs towardsthe bottom of the opaline or microcrystallinebands (Figs 5c e) Where opaline bands are notpresent electrum precipitates at the boundary offirst and third stages (Fig 5b) The electrum hasirregular forms as the result of crystallization inthe quartz interstices being relatively well-shapedin the coarser-grained quartz crystals The size ofelectrum grains ranges from less than 1 to 20 micromHowever the single crystals form chains reachingup to 650 microm (Figs 5b 6a) Electrum in gold-richbands etched with HF is characterized by botryoi-dal surfaces that appear to have formed by theaggregation of spherical electrum particles (Figs6andashb) as observed in the Sleeper deposit Nevada(Saunders 1990 1994) Electrum in bonanza veinscan occupy 50 vol (Figs 5f 6c) accompanied byrare As-bearing pyrite Occasionally micron orsubmicron size electrum inclusions form zoneswithin the As-bearing pyrite (Figs 6d) Tiny AundashAg tellurides have been observed at the contactsbetween electrum grains and small galena crystals(Fig 6f)

The third stage starts with deposition of sugaryquartz (up to 15 mm large Figs 5d 7a) followedby bands of quartz and bladed calcite (Figs 5b 7b)and finally by massive calcite with small amountof quartz (Fig 5d) Locally calcite fills quartz ge-odes or forms small veins (Fig 7a) Microscopical-ly the late calcite corrodes and replaces quartzcrystals Small amount of adularia and pyrite arealso typical for this stage

The hydrothermal mineralization ended withthe deposition of pinkish ankerite-dolomite andsiderite which was preceded by precipitation ofmassive pyrite in some parts of the deposit (Figs7cndashd)

The mineralization of Ada Tepe exhibits sever-al important spatial and textural features thatprovide insights to the physical environment ofore deposition In its upper part quartz becomesthe predominant gangue mineral whereas car-bonates are subordinate This is particularly obvi-ous in the last stage where in the surface samplesdolomite and ankerite are absent in the geodes

and bladed minerals consist of microcrystallinequartz only Another spatial variation is the di-minishing of banding textures with the decreasein grade of Au mineralization to the north Animportant feature is the absence of bladed car-bonates in the veinlets beneath The Wall in con-trast to the entire extent of the deposit above thedetachment surface

53 Oxidized zone

The uppermost part of the Ada Tepe deposit hasbeen oxidized by supergene processes with thedepth of oxidation being dependent upon inten-sity of faulting Due to extensive silicification TheWall itself is generally not oxidized however a 2to 5 m thick zone of oxidation invariably occursabove The Wall Narrow zones of oxidation occurdown to this zone utilizing high angle listric faultsThese faults and fractures most likely acted asconduits for the surface waters to reach the im-permeable silicified zone forming the oxidizedzone above The Wall

Most of the pyrite in the oxidized zones hasbeen converted into limonite identified asgoethite Kaolinite is widely distributed in thiszone forming locally small veinlets Electrum inthe oxidized zone however shows no evidence ofAu enrichment

6 Alteration

In the poorly mineralized zones hydrothermal al-teration is weak with most of the detrital minerals(quartz muscovite and feldspars) remaining unal-tered Alteration consists of chlorite calcite kaoli-nite and subordinate albite pyrite and ankerite-dolomite It seems that calcite and ankerite-dolo-mite are late however it is difficult to distinguishreplacement calcite from vein calcite Quartzadularia calcite pyrite and dolomite-ankerite-siderite plusmn sericite and clay minerals in variableproportions occur in The Wall and immediateproximity to the subvertical quartz and carbonateveins in the strongly mineralized zones Theyformed through replacement of the original meta-morphic minerals included in the sandstone andconglomerate and are associated with relict de-trital quartz and muscovite This complex mineralassemblage is the result of the evolution of thehydrothermal activity

Hydrothermal alteration of the basementmetamorphic rocks is developed within a haloseveral meters in extent beneath The Wall and ismore pronounced around the subvertical veinsThe alteration assemblage includes quartz adu-

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

P Marchev et al68

Fig 6 (a) SEM images of electrum dendrite etched with HF (b) Negative forms of quartz removed by HFand spherical colloidal particles (c) Opal-electrum band containing ~ 50 electrum and small amount ofpyrite Dark crystals between electrum are adularia and silica (d) Pyrite with a zone of submicron-size elec-trum (e) Former Py converted to goethite with inclusions of Au (f) AundashAg tellurides between larger elec-trum and galena crystals

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

69Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

laria illite or sericite pyrite kaolinite and latestage ankerite-dolomite-siderite

Clay minerals are more abundant and betterdeveloped in the alteration close to the surfacewhere they form argillic zones between the EndashW-oriented swarms of veins Kaolinite and quartzpredominate in the argillic zones and illite chlo-rite adularia and carbonate are subordinate ac-companied locally by illite-montmorillonite Partof the shallow-level kaolinite may be supergenebut its association with unoxidized pyrite beneathand within The Wall indicates that a hypogene ori-gin is likely at least for the deeper kaolinite

7 Geochronology

40Ar39Ar ages from adularia and sanidine sam-ples were determined to constrain the timing ofmineralization and volcanic activity in the Kesse-bir dome Sample AT01-2 from which adulariawas extracted for dating is from a drill-hole at725 m depth It represents a quartz-adularia-car-bonate banded vein filled with finely veinedgneiss clasts It is a mixture of completely replaceddetrital K-feldspar and directly deposited adular-ia crystals The rhyolite dyke (sample AT01-17) isemplaced within the core orthogneisses of theKessebir dome located 4 km southwest from AdaTepe (Fig 2) It comprises 2 cm diameter sanidinephenocrysts that were separated for dating plusplagioclase biotite and quartz

40Ar39Ar experiments were conducted at UW-Madison Rare Gas Geochronology LaboratoryMinerals were separated from 100ndash250 micronsieve fractions using standard magnetic densityand handpicking methods Five milligrams of eachmineral were irradiated at Oregon State Univer-sity USA for 12 hours along with 2834 Ma sani-dine from the Taylor Creek rhyolite as the neu-tron fluence monitor Analytical procedures in-cluding mass spectrometry procedural blanks re-actor corrections and estimation of uncertaintiesare given in Singer and Brown (2002) Isotopicmeasurements were made of the gas extracted byincrementally heating of 2ndash3 mg multi-crystal ali-

quots using a defocused CO2 laser beam Resultsof incremental heating experiments along withthe 40Ar39Ar ages of muscovite and biotite fromthe lower plate of the metamorphic basementrocks taken from Bonev et al (accepted) are inTable 1 Incremental heating experiments and agespectra of the sanidine and adularia are presentedin Table 2 and Fig 8 Uncertainties in the agesinclude analytical contributions only at plusmn2

The plateau age of the adularia based on793 of the gas released is 3499 plusmn 023 Ma (Fig8) and is indistinguishable from the total fusionage of 3505 plusmn 023 Ma Sanidine from the rhyolitedyke yielded a plateau age of 3182 plusmn 020 Mabased on 60 percent of the gas released which isindistinguishable from the total fusion age of3188 plusmn 020 Ma As is the case for the adulariathis indicates that argon loss due to weathering oralteration is negligible Moreover the inverse iso-chron ages of 3511 plusmn 037 Ma for the adularaiaand 3191 plusmn 027 Ma for the sanidine are indistin-guishable from the plateau ages and the 40Ar36Arintercept values indicate that there is virtually noexcess argon present

Lavas from the Iran Tepe volcano four kmnortheast of Ada Tepe give KndashAr ages of ~35 Mabut their stratigraphic position above the UpperPriabonianndashLower Oligocene marl-limestone for-mation suggests that they are younger than 34Ma

40Ar39Ar measurements on muscovite and bi-otite (Bonev et al accepted) from the Gneiss-migmatite complex record ages of 3813 plusmn 036and 3773 plusmn 023 Ma respectively (Table 1) Theseages are broadly in agreement with the published40Ar39Ar ages for muscovite (422 plusmn 50 and 361 plusmn50 Ma Lips et al 2000) and muscovite and am-phibole (39 plusmn 1 and 45 plusmn 2 Ma respectively Muka-sa et al 2003) from the Biala Reka metamorphiccore complex Similar ages (39ndash35 Ma) have beenobtained by Peytcheva et al (1992) for the closingof Sr isotope system in the metagranites from theBiala Reka dome

The relative age of the mineralization was de-termined based on its relationship with the hostKrumovgrad group and the overlying Upper

Sample Mineral Weighted Plateau Total Fusion Sourceage (Ma) age (Ma)

AT01-2 adularia 3499 plusmn 023 3505 plusmn 023 this studyAT01-17 sanidine 3182 plusmn 020 3188 plusmn 020 this studyH6 biotite 3773 plusmn 025 3756 plusmn 024 Bonev et al (accepted)H706 muscovite 3813 plusmn 036 3834 plusmn 031 Bonev et al (accepted)

Table 1 Summary of 40Ar39Ar incremental heating experiments on adularia sanidinebiotite and muscovite from Ada Tepe deposit and Kessebir dome

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

P Marchev et al70

Exp

erim

ent

Las

er p

ower

36A

r37

Ara

38A

r39

Ara

40A

r40

Ar

39A

rK

Ca

Age

plusmn 2

num

ber

(Wat

ts)

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

(Vol

ts)

plusmn 1

()

()

(Ma)

b

AT

01-2

adu

lari

a J

= 0

002

987

plusmn 0

0000

09 m

ass

disc

rim

inat

ion

per

amu

10

022

plusmn 0

0004

A

B21

21 0

04

W 0

012

604

00

0003

0 0

000

159

00

0000

5 0

002

772

00

0002

1 0

029

850

00

0007

7 3

871

564

00

0155

74

70

46

132

72

plusmn 3

84A

B21

22 0

10

W 0

009

696

00

0002

2 0

004

566

00

0004

6 0

004

738

00

0002

0 0

239

380

00

0013

1 4

348

793

00

0312

034

92

91

633

85

plusmn 0

38A

B21

23 0

16

W 0

008

804

00

0001

5 0

005

107

00

0003

2 0

006

792

00

0003

0 0

427

709

00

0033

7 5

368

227

00

0278

052

25

32

634

93

plusmn 0

18A

B21

24 0

22

W 0

003

814

00

0001

1 0

000

564

00

0001

0 0

003

949

00

0001

2 0

268

571

00

0011

2 2

865

288

00

0078

961

33

315

034

83

plusmn 0

16A

B21

26 0

30

W 0

008

692

00

0001

1 0

000

261

00

0000

2 0

009

565

00

0002

0 0

650

692

00

0034

7 6

809

601

00

0275

562

78

079

935

01

plusmn 0

11A

B21

27 0

45

W 0

011

765

00

0001

5 0

000

226

00

0000

6 0

017

464

00

0003

5 1

258

986

00

0031

211

697

117

00

0648

770

524

817

70

350

0plusmn

009

AB

2128

06

9 W

00

1320

9 0

000

022

00

0019

9 0

000

003

00

1740

1 0

000

014

12

3066

3 0

001

217

119

6079

5 0

012

113

676

379

195

935

11

plusmn 0

15A

B21

29 1

25

W 0

012

274

00

0002

1 0

000

249

00

0000

3 0

017

405

00

0003

7 1

255

200

00

0080

311

896

055

00

0449

369

817

316

11

353

2plusmn

010

Pla

teau

age

349

9plusmn

023

MSW

D (

n =

5 o

f 8)

186

Tota

l Fus

ion

Age

350

5plusmn

023

AT

01-1

7a s

anid

ine

J =

00

0297

6 plusmn

000

0009

mas

s di

scri

min

atio

n pe

r am

u 1

002

2 plusmn

000

04

AB

2132

00

4 W

00

0055

1 0

000

005

00

0001

4 0

000

006

00

0017

5 0

000

005

00

0362

1 0

000

024

01

8252

2 0

000

076

118

00

456

313

1plusmn

456

AB

2133

00

6 W

00

0019

3 0

000

004

00

0002

6 0

000

003

00

0034

3 0

000

006

00

2579

1 0

000

046

02

1431

6 0

000

110

745

03

447

326

0plusmn

051

AB

2134

01

2 W

00

0018

8 0

000

004

00

0011

6 0

000

007

00

0251

1 0

000

023

02

0468

8 0

000

253

12

8821

1 0

000

967

959

20

568

320

8plusmn

012

AB

2136

01

8 W

00

0050

3 0

000

004

00

0056

8 0

000

012

00

1520

7 0

000

032

12

5179

3 0

000

404

76

2511

3 0

002

510

980

124

678

318

1plusmn

004

AB

2137

02

4 W

00

0056

0 0

000

005

00

0084

3 0

000

018

00

2351

2 0

000

025

19

2193

9 0

000

651

116

3788

5 0

004

597

985

190

700

317

9plusmn

004

AB

2138

02

7 W

00

0036

1 0

000

008

00

0070

7 0

000

010

00

1915

2 0

000

024

15

7872

8 0

000

654

95

5607

9 0

004

091

988

156

686

318

7plusmn

005

AB

2139

03

1 W

00

0036

3 0

000

001

00

0052

2 0

000

008

00

1472

7 0

000

035

12

0293

6 0

000

478

73

0039

4 0

003

784

985

119

709

318

4plusmn

005

AB

2141

03

5 W

00

0031

1 0

000

007

00

0037

7 0

000

008

00

1054

1 0

000

029

08

7899

6 0

000

319

53

5364

2 0

002

077

983

87

716

318

8plusmn

005

AB

2142

04

5 W

00

0032

8 0

000

004

00

0055

3 0

000

015

00

1592

7 0

000

034

13

0785

9 0

000

548

79

4703

7 0

003

239

987

129

726

319

6plusmn

005

AB

2143

05

9 W

00

0028

5 0

000

005

00

0039

4 0

000

012

00

1085

1 0

000

045

08

9161

3 0

000

557

54

3036

7 0

002

402

984

88

696

319

3plusmn

006

AB

2145

08

4 W

00

0016

0 0

000

004

00

0019

6 0

000

007

00

0560

3 0

000

028

04

5183

5 0

000

332

27

6009

9 0

001

984

983

45

706

319

7plusmn

008

AB

2146

13

5 W

00

0016

6 0

000

005

00

0018

3 0

000

004

00

0462

0 0

000

018

03

8294

7 0

000

252

23

5385

3 0

001

397

980

38

647

320

5plusmn

008

Pla

teau

age

318

2plusmn

020

MSW

D (

4 of

12)

279

Tota

l Fus

ion

Age

318

8plusmn

020

aC

orre

cted

for

37A

r an

d 39

Ar

deca

yb

Age

s ca

lcul

ated

rel

ativ

e to

28

34 M

a Ta

ylor

Cre

ek R

hyol

ite

stan

dard

ita

lics

indi

cate

ste

ps o

mit

ted

from

the

plat

eau

age

calc

ulat

ion

see

text

for

deta

ils

Tabl

e 2

40A

r39

Ar

incr

emen

tal-

heat

ing

anal

yses

of a

dula

ria

and

sani

dine

from

Ada

Tep

e

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

71Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

EocenendashOligocene sedimentary rocks The latteris represented by sandstones unaffected by hy-drothermal alteration which cover the northernpart of the deposit

8 Sr and Pb isotope compositionsof carbonates and pyrite

Sr and Pb isotopes were measured by standardTIMS procedures at the University of GenevaSwitzerland following the methods of Valenza etal (2000) and Chiaradia and Fontboteacute (2002) Srisotope ratios on carbonate and Pb isotope ratiosfrom pyrite at Ada Tepe along with whole rock Srand Pb isotope ratios of the leachate fraction of asample from Synap prospect 2 km west of AdaTepe are presented in Table 3 and Figs 9 and 10The Ada Tepe carbonate is from a late stage an-kerite-dolomite vein taken from a drill-core sam-ple (AT01-22) at 996 m depth The bulk rock(sample Sin2000-28) from Synap is from sericite-adularia alteration from the surface Pyrite fromAda Tepe is from a strongly silicified sample(Km2000-6) Pyrite from Synap is from the samehand sample Sin2000-28 on which the Sr wasmeasured Both samples belong to the early alter-

ation The initial 87Sr86Sr ratio of the whole rockcorrected for 35 Ma is 070809 The pyrite sampleand whole rock leachate show identical 207Pb204Pb (1568) and 208Pb204Pb (3885) ratios andslightly different 206Pb204Pb ratios (1871 in AdaTepe and 1866 in Synap) The Sr and Pb isotopedata are compared with those of local Paleogenevolcanic rocks and dykes and the Rhodope meta-morphic rocks in Figs 9 and 10

9 Discussion

91 Physical environment

Epithermal quartz from the Ada Tepe deposit ispoor in fluid inclusions particularly the early mi-crocrystalline quartz and opal Euhedral late sug-ary quartz is the only phase which contains largerfluid inclusions but with no reliable characteris-tics that may allow to classify them as primary in-clusions For the similar Sleeper deposit USASaunders (1994) suggested that quartz formed bytransformation of poorly ordered silica may ex-clude trapped fluids thus precluding the use ofmicrothermometry to constrain the physicochem-ical conditions of silica and gold deposition

Fig 7 (a) Geode of coarse-crystalline quartz with carbonate infill (third stage) (b) Third stage coarse-crystallinequartz (ccq) and bladed parallel type calcite (c) (c) Late vein (forth stage) composed by pyrite on its margins fol-lowed by dolomite-ankerite (da) (d) Same fracture Dolomite-ankerite forms a halo overlapping quartz-adularia-pyrite alteration

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

P Marchev et al72

The presence of adularia that spans the time ofhydrothermal mineralization from the early mi-crocrystalline stage to deposition of bonanza elec-trum bands accompanied by early bladed calciteand followed by late bladed CandashMgndashFe carbon-ates is consistent with boiling during most of thehydrothermal process (Izava et al 1990 Sim-mons and Christenson 1994) Evidence for boil-ing occurs throughout the Ada Tepe deposit fromthe present-day surface down to the detachmentsurface The absence of bladed calcite in the vein-lets beneath the detachment however suggeststhat boiling started as the fluid reached the sedi-mentary sequence Boiling cools the fluid andconcentrates dissolved silica leading to precipita-tion of silica colloids (Saunders 1994) Thus muchof the mineralization in The Wall appears to coin-cide with the onset of boiling

Thermal conditions can be roughly evaluatedon the basis of the distribution of temperature-sensitive minerals The absence of epidote in thehost-rock alteration can be interpreted in favor ofa low-temperature (lt240 degC) for the initial fluids(Bird et al 1984 Rayes 1990) This is confirmedby the large distribution of kaolinite crystallizingat temperatures lt200 degC (Simmons and Browne2000) and deposition of microcrystalline silica(chalcedony) or its transformation to jigsawquartz in the stage of massive silicification whichoccurs at temperature of gt180 degC (Fournier1985) Thus the temperature of the early fluidseems to have been 200ndash180 degC Deposition ofbonanza opal-electrum bands however requiresconsiderably lower temperatures (100ndash150 degCHedenquist et al 2000) suggesting a significantdrop of temperature (see also Saunders 1994)Such cooling can be attributed to simple boilingcausing cooling during decompression and in-crease in pH and ƒO2 which accompanies the lossof dissolved gasses Alternatively it can be the re-

sult of mixing with cooler meteoric groundwaterin a manner similar to that described for theHishikari low-sulfidation deposit (Izawa et al1990) Mixing of cool oxidized meteoric waterwith a hot reduced deep crustal fluid has beenproposed for low-angle faults bounding severalmetamorphic core complexes (Kerrich and Reh-ring 1987 Kerrich and Hyndman 1987 Spencerand Welty 1986 Beaudoin et al 1991)

92 Paleodepth

Mineralogical and textural characteristics such asbladed carbonates and crustiform and colloformbanded chalcedony and opaline silica plus adular-ia (Izawa et al 1990 Cook and Simmons 2000)suggest that the depth for the formation of miner-alization at Ada Tepe deposit was shallow In theabsence of precise temperature and salinity datafrom fluid inclusion studies the paleodepth of theAda Tepe deposit cannot be precisely con-strained A minimum depth can be roughly esti-mated assuming boiling of low salinity solutions(lt1 wt NaCl equivalent) at a temperature of200 degC The depth of boiling of such a fluid belowthe water table is about 160ndash170 m (Haas 1971)

Fig 8 40Ar39Ar incremental-heating age spectra for adularia from Ada Tepe and sanidine from rhyolite dyke fromKessebir dome

Sample AT 01-22 Sin2000-28 KM 2000-6carbonate bulk rock pyrite

Rb 218Sr 8287Rb86Sr 7695087Sr86Sr 0709270plusmn14 0711911plusmn10(87Sr86Sr)i 070809206Pb204Pb 18663 18707207Pb204Pb 15682 15684208Pb204Pb 38849 38843

Table 3 Sr and Pb isotope compositions of carbonatepyrite and whole rock from Ada Tepe deposit and Synapmineralization

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

73Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

For fluids under a hydrodynamic gradient andcontaining dissolved CO2 the depth of boilingwould be somewhat deeper (Hedenquist andHenley 1985) This is consistent with the 350 mthickness of the host Shavarovo Formation(Goranov and Atanasov 1992) which implies thatthe overlying coal-bearing-sandstone and marl-limestone formations had not yet been depositedwhen the ore formed

93 Sources and relationships among minerali-zation extension and magmatism

The new 40Ar39Ar ages from adularia and sani-dine have been compared with those from musco-vite and biotite in the Kessebir dome (Bonev etal accepted) KndashAr ages of nearby Iran Tepe vol-cano (Lilov et al 1987) and intra-plate alkalinebasalts (Marchev et al 1997 1998) to clarify rela-tionships between volcanism thermal events inthe Kessebir dome and mineralization in the AdaTepe deposit As discussed above the 40Ar39Artotal fusion age of adularia (Marchev et al 2003)shows that the mineralization at Ada Tepe formedat 35 Ma Mineralization may have been coevalwith 35 Ma lavas of the Iran Tepe paleovolcano(Lilov et al 1987 Pecskay pers comm) howeverthe uncertainties of the dating methods do notrule out differences in age on the order of several100 kyr Notwithstanding field relationships indi-cate that volcanic activity of the Iran Tepe volca-no is younger than the Krumovgrad group

Another possible source of fluids and metals isthe magmatism of the Kessebir dome itself Com-parison of the timing of Ada Tepe mineralization(35 Ma) the Kessebir rhyolite dyke (3182 plusmn 020Ma) and KndashAr ages of intra-plate alkaline basalts(28ndash26 Ma) show that mineralization preceded byat least 3 Ma the rhyolitic magmatism and at least6 Ma the intra-plate basalts Thus it is highly un-

likely that this younger igneous activity was thesource of Ada Tepe mineralizing fluids or heat

Mineralization at Ada Tepe is ca 3 Ma young-er than the 40Ar39Ar ages of the muscovite (3813plusmn 036) and biotite (3773 plusmn 025 Ma) from thelower plate of the Kessebir dome Recently ob-tained older 40Ar39Ar age for amphibole (45 plusmn 2Ma) and similar age for the muscovite (39 plusmn 1 Ma)in the Biala Reka dome (Mukasa et al 2003)have been interpreted as cooling ages for the up-per plate Therefore muscovite and biotite agesfrom Kessebir can be interpreted as cooling of thelower plate rocks to below 350ndash300 degC followingthe MaastrichtianndashPaleocene metamorphism(Marchev et al 2004) The latter ages are general-ly consistent with a 42ndash35 Ma range of ages ob-tained earlier by Peytcheva (1997) for the closingof RbndashSr system in the granitoids in Biala Rekaand Kessebir after a presumable amphibolitefacies metamorphism Ore deposit formation at~35 Ma in the hanging wall of the Tokachka de-tachment coincides with late stage brittle exten-sion after cooling of the basement rocks at tem-peratures lt 200 degC (Bonev et al accepted) Thehighest grade portion of the ore mineralization(The Wall) is dominated by multiple phases ofveining and brecciation suggesting a syndetach-ment evolution of the hydrothermal process Thepresence of several unconformities in the overly-ing Upper EocenendashLower Oligocene coal-bear-ing-sandstone and marl-limestone formations(Goranov and Atanassov 1992 Boyanov andGoranov 1994 2001) suggests a continuation ofthe uplift and exhumation of the Kessebir domeeven several million years after the formation ofthe Ada Tepe deposit

Probable source rocks for the Sr (Ca) and Pbcan be constrained by comparison of the Sr andPb isotope ratios of carbonates and pyrites fromAda Tepe with the major source rocks (Figs 9 and

Fig 9 Sr isotope composition of calcite from Ada Tepe and whole rock sample from Synap compared withKrumovgrad area igneous and metamorphic rocks Data for the intra-plate basalts ndash Marchev et al 1998 for theZvezdel and Iran Tepe lavas ndash Marchev unpublished data for the metamorphic rocks ndash Peytcheva et al 1992 andPljusnin et al 1988)

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

P Marchev et al74

10) Bulk rock and carbonate Sr isotope data(070809ndash070927) lie within the lower range ofpresent-day 87Sr86Sr isotopic ratios between0708ndash0737 for the prevailing gneissic metamor-phic rock within the Eastern Rhodopes (Peytch-eva et al 1992) and are higher than the Sr isoto-pic values of the nearby younger igneous rocksfrom Iran Tepe and Zvezdel volcanoes (070641ndash070741 Marchev et al unpubl data) It appearslikely that the hydrothermal Sr and by inferenceCa is mostly of metamorphic origin or represent amixture of metamorphic and small amount of oldmagmatic source isotopically similar to Iran Tepeand Zvezdel magmas (see below)

Pyrite Pb isotope ratios overlap with the fieldof feldspars from Zvezdel igneous rocks howeverthe leachate fraction of the whole rock sample

from Synap has a slightly lower 206Pb204Pb ratiothan the feldspars Collectively the two samplesplot in the fields of the Rhodope metamorphicrocks From these data it is not possible to dis-criminate between a metamorphic and an igneousorigin of Pb in the hydrothermal fluid This is notsurprising as orogenic igneous rocks of the East-ern Rhodopes were strongly contaminated withcrustal Pb (Marchev et al 1998 and 2004) and dif-fer isotopically from the asthenospheric-derivedKrumovgrad intra-plate basalts (Fig 10) There-fore isotopic homogenization of the orogenic ig-neous and metamorphic rocks in the region limitsthe possibility to discriminate between these twosources Nevertheless it may be suggested thatfluid circulation through the metamorphic base-ment rocks and metamorphic-derived sediments

Fig 10 Pb isotope composition of pyrite (Ada Tepe) and whole rock alteration (Synap) compared with feldsparsand whole rocks from Zvezdel Oligocene volcano (Marchev et al unpubl data) and whole rocks from Krumovgradintra-plate basalts (Marchev et al (1998) and Rhodope metamorphic rocks from Frei (1995)

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

75Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

would cause a large contribution from metamor-phic lead

Thus the simplest genetic model to explain thegeology geochronology and Sr and Pb isotopesinvolves hydrothermal convection through thePaleozoic metamorphic basement The close tem-poral association of mineralization at Ada Tepewith the exhumation processes and formation ofthe Kessebir dome favours a genetic link betweenthem We propose that the metamorphic base-ment rocks provided the heat to drive the hydro-thermal system During exhumation however thelower plate of the Kessebir metamorphic corecomplex experienced cooling and decompressionsince the Upper Cretaceous (Marchev et al2004) This has been used in other metamorphiccomplexes (Smith et al 1991) to argue that pro-grade metamorphic devolatilization of the lowerplate is unlikely to generate fluid responsible formineralization during extension These authorssuggested that igneous rather than metamorphicfluid source is more likely for the metamorphiccore complex-related mineralization This appar-ent contradiction can be explained by astheno-spheric upwelling that was operating in the East-ern Rhodope area (Marchev et al 2004) causingretrograde metamorphism at shallow levels andat the same time resulting in prograde metamor-phism at deeper levels observed by the high heatflow in the region Although magmatic activityseems to be younger than the ore mineralizationthe first erupted lavas of Iran Tepe were close intime The beginning of this volcanism was proba-bly preceded by accumulation of magma at great-er depths Under such conditions fluids liberatedby deep prograde devolatilization reactions andor mantle degassing could ascend to form lowtemperature fluids

94 Gold deposits associatedwith detachment faults

Different types of mineralization in low-angle de-tachment faults associated with metamorphiccore complexes are described mostly in the south-western United States These include Picacho(Drobeck et al 1986 Liebler 1988) and BullardPeak (Roddy et al 1988) Southwestern Arizonaand Riverside Pass (Wilkinson et al 1988) andWhipple-Buckskin-Rawhide (Spencer and Welty1986) Southeastern California Another exampleis in the Kokanee Range Southern British Co-lumbia Canada (Beaudoin et al 1991)

These kinds of deposits are divided by Eng etal (1988) into two categories base-metal en-riched (Whipple-Buckskin-Rawhide and Koka-nee deposits) and base-metal poor (Picacho and

Riverside Pass) In the base-metal poor depositsgold is typically associated with silicification andhematite (Picacho) or hematite and chrisocolla(Riverside Pass) Pyrite which is entirely oxidizedin Riverside Pass is the principal sulfide mineralQuartz and lesser calcite are the main gangueminerals

Mineralization at Ada Tepe deposit sharesmany characteristics with Picacho and RiversidePass including association of gold with multipleepisodes of silicification and veining within shal-lowly dipping detachment faults and a low con-tent of base-metals and As However comparisonof the characteristics of Ada Tepe to other base-metal poor detachment-related mineralizationsreveals several notable differences These includelack of copper and specular hematite a relativelyhigh content of adularia and various silica poly-morphs and carbonates These features may re-flect differences in crustal compositions and his-tory the nature of the ore-forming hydrothermalfluids and physico-chemical conditions at the fo-cus of ore deposition

10 Conclusions

(1) Based on alteration mineralization andtextures we classify Ada Tepe as a low-sulfidationepithermal gold deposit hosted in sedimentaryrocks

(2) The Tokachka low-angle detachment faultand the steep listric faults in the Maastrichtian-Paleocene sedimentary sequence were the majorsites of gold mineralization There are many indi-cators including breccias that synmineralizationmovement took place along the detachment sur-face

(3) The deposit formed at shallow depth lessthan 200ndash250 m below the paleosurface Micro-crystalline quartz and opal which are the main sil-ica polymorphs crystallizing with the electrumbands preclude the use of fluid inclusion to con-strain the conditions of gold deposition Fromother alteration minerals temperature of forma-tion is evaluated to lt200 degC The mineralogy sug-gests that the fluid boiled throughout the deposit

(4) Mineralization is an exclusively Au systemwith traces of As and with no base metals Themajority of the gold occurs as electrum with 73ndash76 Au which is reflected in the average AuAgratio (~3) of the deposit

(5) The 40Ar39Ar age of adularia indicates thatthe ore was deposited at Ada Tepe at 350 plusmn 02Ma This age is ~3 Ma younger than the closure ofmetamorphic muscovite and biotite at ~350 degC inthe underlying basement and 3 Ma older than sa-

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

P Marchev et al76

nidine in the nearest rhyolite thereby precludinglocal magmatism as a source of fluids or heat

(6) Sr and Pb isotope ratios are consistent withthe idea that metals and carbonates were proba-bly derived from the metamorphic basementrocks with a possible contribution from an igne-ous source

(7) Collectively these data suggest an intimateassociation of Ada Tepe Au mineralization to themetamorphic core-complex formation rather thanto the local magmatism

Acknowledgments

This is a contribution to the ABCD-GEODE programof the European Science Foundation Supported by theSwiss National Science Foundation Joint ResearchProject 7BUPJ062276 and grants 21-5904199 and200020-101853 40Ar39Ar analyses at UW-Madison sup-ported by US NSF grants EAR-10114055 and EAR-0337667 to Singer BMM gave permission to publishthese results and provided financial support The au-thors appreciate the comments and helpful reviews ofthe journal reviewers D Altherton N Skarpelis and Al-brecht von Quadt Denis Fontignie and Massimo Chi-aradia (Universities of Geneva Switzerland and LeedsUK) are thanked for the Sr and Pb isotope analysesThanks are owed to Albrecht von Quadt and IvanBonev for etching of gold samples and the SEM picturesof the etched samples respectively

References

Atanasov G and Goranov A (1984) On the Palaeo-geography of the East Rhodopes C R Acad bulgSci 37(6) 783ndash784

Beaudoin G Taylor BE and Sangster DF (1991) Sil-ver-lead-zinc veins metamorphic core complexesand hydraulic regimes during crustal extensionGeology 19 1217ndash1220

Belmustakova H Boyanov I Ivanov I and Lilov P(1995) Granitoid bodies in the Byala Reka dome CR Acad bulg Sci 48 (4) 37ndash40 (in Russian)

Bird DK Schiffman P Elders WA Williams AEand McDowell SD (1984) Calc-silicate mineraliza-tion in active geothermal systems Econ Geol 79671ndash695

Bonev N (1996) Tokachka shear zone southwest ofKrumovgrad in Eastern Rhodopes Bulgaria an ex-tensional detachment Ann Univ Sofia 89 97ndash106

Bonev N (2002) Structure and evolution of the Kesse-bir gneiss dome Eastern Rodopes PhD thesis Univof Sofia unpubl 282 pp (in Bulgarian)

Bonev N Marchev P and Singer B (accepted) Inter-ference between tectonic ore-forming and magmat-ic processes during the Tertiary extensional exhuma-tion in the Eastern Rhodope (Bulgaria) 40Ar39Argeochronology constraints Geodin Acta

Boyanov I and Goranov A (1994) PaleocenendashEocenesediments from the Northern periphery of theBorovica depression and their correlation with simi-lar sediments in the East Rhodopean Paleogene de-pression Rev Bulg Geol Soc 55(1) 83ndash102 (in Bul-garian with English abstract)

Boyanov I and Goranov A (2001) Late Alpine (Paleo-gene) superimposed depressions in parts of South-east Bulgaria Geol Balc 34(3ndash4) 3ndash36

Carrigan C Mukasa S Haydoutov I and Kolcheva K(2003) Ion microprobe UndashPb zircon ages of pre-Al-pine rocks in the Balkan Sredna Gora and Rho-dope terranes of Bulgaria Constraints on Neopro-terozoic and Variscan tectonic evolution J CzechGeol Soc Abstract volume 481ndash2 32ndash33

Chiaradia M and Fontboteacute L (2002) Separate leadisotope analyses of leachate and residue rock frac-tions implications for metal source tracing in oredeposit studies Mineral Deposita 38 185ndash195

Cook DR and Simmons SF (2000) Characteristics andgenesis of epithermal gold deposits Reviews EconGeol 13 221ndash224

Crummy J (2002) A speculative genetic model for thelocation of gold mineralization on the Krumovgradlicense SE Rhodope Bulgaria Geol Mineral Res1 20ndash24

Duffield W and Dalrymple GB (1990) The TaylorCreek Rhyolite of New Mexico a rapidly emplacedfield of lava domes and flows Bull Volcanol 52475ndash487

Drobeck PA Hillemeyer JL Frost EG and LieblerGS (1986) The Picacho Mine a gold mineralizeddetachment in southeastern California In Beatty Band Wilkinson PAK (eds) Frontiers in geologyand ore deposits of Arizona and the Southwest Ari-zona Geol Soc Digest XVI Tucson 187ndash221

Eng T Boden DR Reischman MR and Biggs JO(1995) Geology and mineralization of the BullfrogMine and vicinity Nye County Nevada In CoynerAR and Fahey PL (eds) Geology and ore depos-its of the American Cordillera Symposium Proceed-ings Geol Soc Nevada RenoSparks Nevada Vol1353ndash400

Fournier RO (1985) Silica minerals as indicators ofconditions during gold deposition US Geol SurveyBull 1646 15ndash26

Frei R (1995) Evolution of mineralizing fluid in theporphyry copper system of the Scouries depositNortheast Chalkidiki (Greece) Evidence from com-bined PbndashSr and stable isotope data Econ Geol 90746ndash762

Goranov A and Atanasov G (1992) Lithostratigraphyand formation conditions of Maastrichtian-Paleo-cene deposit in Krumovgrad District Geol Balc22(3) 71ndash82

Haas JL Jr (1971) The effect of salinity on the maxi-mum thermal gradient of a hydrothermal system athydrostatic pressure Econ Geol 66 940ndash946

Hedenquist JW and Henley RW (1985) The impor-tance of CO2 on freezing point measurment of fluidinclusions evidence from active geothermal systemsand implications for epithermal ore depositionEcon Geol 80 1379ndash1399

Hedenquist JW Arribas AR and Gonzales-Urien E(2000) Exploration for epithermal gold depositsSoc Econ Geol Reviews 13 245ndash277

Izawa E Urashima Y Ibaraki K Suzuki R Yokoya-ma T Kawasaki K Koga A and Taguchi S (1990)The Hishikari gold deposit High grade epithermalveins in Quaternary volcanics of southern KyushuJapan J Geochem Explor 36 1ndash56

Kerrich R and Rehring W (1987) Fluid motion associ-ated with Tertiary mylonitization and detachmentfaulting 18O16O evidence from the Picacho meta-morphic core complex Arizona Geology 15 58ndash62

Kerrich R and Hyndman D (1987) Thermal and fluidregimes in the Bitteroot lobe-Sapphire block de-tachment zone Montana Evidence from 18O16Oand geologic relations Geol Soc Am Bull 97 147ndash155

77Ada Tepe gold deposit in the Eastern Rhodopes SE Bulgaria

Kolcheva K and Eskenazy G (1988) Geochemistry ofmetaeclogites from the Central and Eastern Rho-dope Mts (Bulgaria) Geol Balc 18 61ndash78

Kozhoukharov D Kozhoukharova E and Papaniko-laou D (1988) Precambrian in the Rhodope massifIn Zoubek V (ed) Precambrian in Younger FoldBelts Chichester 723ndash778

Kozhoukharova E (1984) Origin and structural posi-tion of the serpentinized ultrabasic rocks of the Pre-cambrian ophiolitic association in the RhodopeMassif I Geologic position and composition ofophiolite association Geol Balc 14 9ndash36 (in Rus-sian)

Kunov A Stamatova V Atanasova R and Petrova P(2001) The Ada Tepe Au-Ag-polymetallic occur-rence of low-sulfidation (adularia-sericite) type inKrumovgrad district Mining and Geol 2001(4) 16ndash20 (in Bulgarian)

Liebler GS (1988) Geology and gold mineralization atthe Picacho mine Imperial County California InSchafer RW Cooper JJ and Vikre PG (eds) Bulkmineable precious metal deposits of the WesternUnited States Symposium Proceedings Geol SocNevada RenoSparks Nevada Part IV Gold-silverdeposits associated with detachment faults 453ndash504

Lilov P Yanev Y and Marchev P (1987) KAr dating ofthe Eastern Rhodopes Paleogene magmatism GeolBalc 17(6) 49ndash58

Lips ALW White SH and Wijbrans JR (2000) Mid-dle-Late Alpine thermotectonic evolution of thesouthern Rhodope Massif Greece Geodin Acta 13281ndash292

Marchev P Harkovska A Pecskay Z Vaselli O andDownes H (1997) Nature and age of the alkalinebasaltic magmatism south-east of Krumovgrad SE-Bulgaria C R Acad bulg Sci 50(4) 77ndash80

Marchev P Vaselli O Downes H Pinarelli L IngramG Rogers G and Raicheva R (1998) Petrologyand geochemistry of alkaline basalts and lampro-phyres implications for the chemical composition ofthe upper mantle beneath the Eastern Rhodopes(Bulgaria) In Christofides G Marchev P and SerriG (eds) Tertiary magmatism of the Rhodopian re-gion Acta Vulcanologica 102 233ndash242

Marchev P and Singer B (2002) 40Ar39Ar geochronol-ogy of magmatism and hydrothermal activity of theMadjarovo base-precious metal ore district easternRhodopes Bulgaria In Blundell D Neubauer Fand von Quadt A (eds) The timing and location ofmajor ore deposits in an evolving orogen Geol SocLondon Spec Publ 204 137ndash150

Marchev P Singer B Andrew C Hasson A MoritzR and Bonev N (2003) Characteristics and prelim-inary 40Ar39Ar and 87Sr86Sr data of the UpperEocene sedimentary-hosted low-sulfidation golddeposits Ada Tepe and Rosino SE Bulgaria possi-ble relation with core complex formation In Elio-poulos et al (eds) Mineral Exploration and Sus-tainable Development Millpress Rotterdam 1193ndash1196

Marchev P Raicheva R Downes H Vaselli O Massi-mo C and Moritz R (2004) Compositional diversi-ty of Eocene-Oligocene basaltic magmatism in theEastern Rhodopes SE Bulgaria implications for itsgenesis and geological setting Tectonophysics 393301ndash328

Marchev P Raicheva R Singer B Downes H AmovB and Moritz R (2000) Isotopic evidence for theorigin of Paleogene magmatism and epithermal oredeposits of the Rhodope Massif In Geodynamicsand Ore Deposit Evolution of the Alpine-Balkan-Carpathian-Dinaride-Province Borovets Abstracts

ABCD-GEODE 2000 workshop Borovets Bulga-ria p 47

Mposkos E and Krohe A (2000) Petrological andstructural evolution of continental high pressure(HP) metamorphic rocks in the Alpine RhodopeDomain (N Greece) In Panayides I XenopontosC and Malpas J (eds) Proceedings of the 3rd Inter-national Conf on the Geology of the Eastern Medi-terranean (Nicosia Cyprus) Geol Survey NicosiaCyprus 221ndash232

Mposkos E and Wawrzenitz N (1995) Metapegmatitesand pegmatites bracketing the time of high P meta-morphism in polymetamorphic rocks of the E-Rho-dope N Greece Petrological and geochronologicalconstraints Geol Soc Greece Spec Publ 4(2) 602ndash608

Mukasa S Haydoutov I Carrigan C and Kolcheva K(2003) Thermobarometry and 40Ar39Ar ages of ec-logitic and gneissic rocks in the Sredna Gora andRhodope terranes of Bulgaria J Czech Geol SocAbstract volume 481ndash2 94ndash95

Ovtcharova M Quadt A von Heinrich CA FrankM and Kaiser-Rohrmeier M (2003) Triggering ofhydrothermal ore mineralization in the CentralRhodopean Core Complex (Bulgaria) ndash Insightfrom isotope and geochronological studies on Terti-ary magmatism and migmatization In Eliopoulos etal (eds) Mineral Exploration and Sustainable De-velopment Millpress Rotterdam 367ndash370

Peytcheva I (1997) The Alpine metamorphism in East-ern Rhodopes ndash RbndashSr isotope data Rev Bulg GeolSoc 58(3) 157ndash165 (in Bulgarian with English ab-stract)

Peytcheva I Kostitsin Y Salnikova E OvtcharovaM and von Quadt A (1999) The Variscan orogen inBulgaria ndash new isotope-geochemical data RomJour tectonics and regional geology 77 Abstract vol-ume p 72

Peytcheva I Kostitsin JA and Shukolyukov JA(1992) RbndashSr isotope system of gneisses in theSouth-Eastern Rhodopes (Bulgaria) C R Acadbulg Sci 45(10) 65ndash68 (in Russian)

Peytcheva I Ovcharova M Sarov S and Kostitsin Y(1998) Age and metamorphic evolution of meta-granites from Kessebir reka region Eastern Rho-dopes ndash RbndashSr isotope data Abstracts XVI Con-gress CBGA Austria Vienna

Peytcheva I and von Quadt A (1995) UndashPb dating ofmetagranites from Byala-Reka region in the EastRhodopes Bulgaria Geol Soc Greece Spec Publ4(2) 637ndash642

Plyusnin GS Marchev PG and Antipin VS (1988)Rubidium-Strontium age and genesis of the sho-shonite-latite series in the Eastern-Rhodope Bul-garia Dokl Akad Nauk SSSR 303(3) 719ndash724

Reyes AG (1990) Petrology of Philippine geothermalsystems and the application of alteration mineralogyto their assessment J Volcanol Geotherm Res 43279ndash309

Ricou LE Burg JP Godfriaux I and Ivanov Z (1998)Rhodope and Vardar the metamorphic and the olis-tostromic paired belts related to the Cretaceous sub-duction under Europe Geodin Acta 11 285ndash309

Roddy MS Reynolds SJ Smith BM and Ruiz J(1988) K metasomatism and detachment-relatedmineralization Harcuvar Mountains Arizona GeolSoc Am Bull 100 1627ndash1639

Rohrmeier M Quadt Avon Handler R OvtcharovaM Ivanov Z and Heinrich C (2002) The geody-namic evolution of hydrothermal vein deposits inthe Madan metamorphic core complex BulgariaGeochim Cosmochim Acta Special Supplement

P Marchev et al78

Abstracts of the 12th Ann Goldschmidt Confer-ence Davos Switzerland 66 S1 A645

Sarov S and twelve others (1996) Report for the resultsof execution of the geological task ldquoGeological map-ping in scale 125 000 and geomorphologic mappingin scale 150 000 with complex prognostic evaluationof the mineral resources in the area of Krumovgradand villages Gornoseltsi Popsko Djanka and Nano-vitsa (Eastern Rhodopes) on area of 485 km2 Na-tional geofond IV-438

Saunders JA (1990) Colloidal transport of gold and sil-ica in epithermal precious-metal systems Evidencefrom the Sleeper deposit Nevada Geology 18 757ndash760

Shabatov Y and eight others (1965) Report for the geo-logical mapping accompanied with prospecting forore mineralizations in scale 125 000 in 1963ndash1964 inpart of the SE Rhodopes Geofond of Committee ofGeology IV-217

Saunders JA (1994) Silica and gold texture in bonanzaores of the Sleeper deposit Humbolt county Ne-vada Evidence for colloids and implications for epi-thermal ore-forming processes Econ Geol 89 628ndash638

Simmons SF and Browne PRL (2000) Hydrothermalminerals and precious metals in the Broadlands-Ohaaki geothermal system Implications for under-standing low-sulfidation epithermal environmentsEcon Geol 95 971ndash999

Simmons SF and Christenson BW (1994) Origins ofcalcite in a boiling geothermal system Am J Sci294 361ndash400

Singer B and Brown LL (2002) The Santa Rosa Event40Ar39Ar and paleomagnetic results from the Valesrhyolite near Jaramillo Creek Jemez Mountains NewMexico Earth Planet Sci Lett 197 51ndash64

Singer B and Marchev P (2000) Temporal evolution ofarc magmatism and hydrothermal activity includingepithermal gold veins Borovitsa caldera southernBulgaria Econ Geol 95 1155ndash1164

Smith BM Reynolds SJ Day HW and Bodnar RJ(1991) Deep-seated fluid involvement in ductile-brit-tle deformation and mineralization South Mountainmetamorphic core complex Arizona Geol Soc AmBull 103 559ndash569

Spencer JE and Welty JW (1986) Possible controls ofbase- and precious-metal mineralization associatedwith Tertiary detachment faults in the lower Colora-do River trough Arizona and California Geology14 195ndash198

Stoyanov R (1979) Metallogeny of the Rhodope Cen-tral Massif Nedra Moscow 180 pp (in Russian)

Valenza K Moritz R Mouttaqi A Fontignie D andSharp Z (2000) Vein and karstic barite deposits inthe Western Jebilet of Morocco Fluid inclusion andisotope (SOSr) evidence for regional fluid mixingrelated to Central Atlantic rifting Econ Geol 95587ndash605

Wilkinson WH Wendt CJ and Dennis MD (1988)Gold mineralization along Riverside Mountains De-tachment Fault Riverside County California InSchafer RW Cooper JJ and Vikre PG (eds)Bulk mineable precious metal deposits of the West-ern United States Symposium Proceedings GeolSoc Nevada RenoSparks Nevada Part IV Gold-silver deposits associated with detachment faults487ndash504

Received 24 March 2003Accepted in revised form 23 July 2004Editorial handling A von Quadt