Tectonic evolution, petrochemistry, geochronology and palaeohydrology of the Tampakan porphyry and high sulphidation epithermal Cu-Au deposit Mindanao, Phillipines
Magmatic-related porphyry copper and high-sulphidation epithermal copper-gold ore deposits in continent-margin and intra-oceanic arcs of the Pacific Rim are spatially clustered in discrete volcanic arc segments. Ore-forming episodes likewise occur within discrete time intervals and are temporally associated with intervals of compression within arcs. Understanding the spatial and temporal controls on fertility of magmas requires detailed multidisciplinary studies of young, ore-productive...[Show more] districts where the tectonic evolution of the arc can be closely integrated with the evolution of crustal stress, with the timing of mineralisation and with evolution of petrological, petrochemical and magmatic physico-chemical properties. The late Miocene to Recent magmatism of the Tampakan ore district of southern Mindanao, Philippines, provides this opportunity. Porphyry copper and high sulphidation epithermal mineralisation within the giant Tampakan Cu-Au deposit (2500 Mt@ 0.48% Cu) are hosted by a polygenetic volcanic complex that was constructed over the past 7 Myr. This interval spans the pre-, syn- and late-collision stages of arc-arc collision in the southern Mindanao segment of the Sangihe arc. Synthesis of tectonic reconstructions and of plate motions from GPS data reveal that crustal compression in southern Mindanao commenced at ~7 Ma and peaked at ~4-3 Ma as subduction waned at the divergent Sangihe and Halmahera subduction systems, and during establishment of the nascent Philippine Trench and Cotabato Trench subduction systems. Porphyry Cu mineralisation at 4.24-4.26 Ma (⁴⁰Ar- ³⁹Ar) and high-sulphidation Cu-Au mineralisation at 3.24-3.28 Ma (⁴⁰Ar- ³⁹ Ar; K/Ar) formed during peak compression. Crustal deformation was manifested by regional folding and thrust faulting. Laser-ablation ICPMS ²³⁸U-²⁰⁶Pb dating of detrital and rock-hosted zircon grains, together with ⁴⁰Ar-³⁹Ar and K/Ar radiometric dating and whole-rock chemistry define five magmatic cycles which extended from the late Miocene to the present. These produced four stratovolcanoes that were built and eroded in successive eruptive and erosional cycles. These semi-continuous magmatic products recorded the covariation of magmatic physico-chemical variables as the arc underwent a transition from normal subduction to cessation of subduction associated with compressive crustal stress during arc-arc collision. The volcanic series evolved from water-poor pyroxene-hornblende-phyric basaltic andesites to hornblende-pyroxene andesites and to waterrich hornblende-biotite dacites. Major and trace element chemical data reveal progressive advance of hornblende saturation, and retreat of plagioclase saturation in the crystallisation sequence of successive magmatic cycles. High Sr/Y ratios are commonly attributed as a chemical feature of adakites, which are ascribed an origin by partial melting of an eclogitic source where refractory garnet retains Y. In contrast, high Sr/Y ratios in the Tampakan district are caused by hydrous, high-pressure crystal fractionation in the lower crust, where suppression of plagioclase crystallisation by high magmatic water activities allows Sr to accumulate during crystal fractionation, and crystallisation of hornblende to form lower-crustal hornblende-augitemagnetite cumulates depletes Y from the residual melt. This element ratio (Sr/Y) is identified as a qualitative indicator of the magmatic water activity, and is a robust guide to the ore-forming potential of a magmatic series. Magmatic temperatures for the series range from 765°C to 909°C and volcanic log /02 varies between NNO+l.53 and NN0+2.50. Sulphur abundances in the melt are low (312 to 57 ppm) and decrease systematically with temperature. Data points for the series plot around the anhydrite-saturation curve in :ΣSmelt-temperature coordinates, consistent with sulphur speciation calculations based on measured oxygen fugacity that indicate SO₃:H₂S abundance proportions between ~200: 1 and ~6,000: 1 in the melt. The magmatic series were saturated with anhydrite during much of their evolution. Magmatic water contents were calculated for the successive magmatic cycles that erupted over the past 7 Myr. Magmatic water contents calculated using the Housh and Luhr (1991) plagioclase-melt Na-Ca exchange geohygrometer reveal water contents that increase from 4.1 % in the late Miocene to up to 8.2 % in the Pleistocene. ²³⁸U-²⁰⁶Pb geochronology on 471 zircon samples from the Tampakan volcanic succession were used to parameterise time series in chemical compositions of volcanic rocks and phenocrysts, and time-series in magmatic temperature, oxygen fugacity and wt.% H₂0 in the pre-eruptive magma over the past 7 Myr. U/Ti, U/Ge and Th/Ti ratios in dated detrital zircon grains resolve multiple million-year-scale magma recharge-and-crystallisation cycles within a long-lived lowercrustal chamber. This deep reservoir resides at 18-22 km depth (~5-6 kbars; Al-in-hornblende geobarometry). The cyclic ramp-up and drop of these element ratios coincides with a 7 Myr-long "sawtooth" cyclic ramp-up in concentrations of volatiles and crystal incompatible trace elements in erupted andesites and dacites. Water contents climbed from 4.1 wt.% to 8.2 wt.% as SiO₂ evolved from 57 to 67 wt.%, because the accumulation of volatiles in residual melt was passed down through multiple cycles of magma-chamber replenishment, magma mixing and crystallisation. The lower-crustal chamber was periodically tapped to form overlying subvolcanic chambers and four overprinting stratovolcanoes within the late Miocene to Recent Tampakan polygenetic volcanic complex. A lower-crustal magma chamber having a long lifespan and slow crystallisation rate relative to the frequency of recharge is required in order to generate the observed petrochemical trends and cyclic climb magmatic water content relative to SiO₂ content of the melts. This longevity is thermally and physically permissible where magma entrapment in the lower crust occurs in compressive stress regimes beneath volcanic arc segments that undergo transient collision, or are under-thrust by buoyant segments of the subducting plate. Calculated buoyancy forces of 1-3 km thick basaltic to andesitic melt columns in the ductile lower crust are comparable to horizontal tectonic stresses in orogenic zones, indicating that melt entrapment can be modulated by an ambient stress regime that inhibits magma ascent by dyke propagation. Numerical thermal models created using the 2-D graphical, user-interactive, heat flow program KWare HEAT predict that lower-crustal sills that are entrapped in the lower crust cool extremely slowly, with residual melt fractions remaining above the wet solidus for several million years, so intermittently erupted magmas exhibit chemical continuity over the ~3-10 Myr period of crustal compression in collisional volcanic arcs. The results from this integrated study of the Tampakan district indicate that the spatial and temporal clustering of magmatic Cu-Au porphyry ores in volcanic arcs is a product of shared regional compressive stress which inhibits magma ascent by sub-vertical dyke propagation and promotes development of sub-horizontal magma chambers in the lower crust, where the trapped magma proceeds to crystallise cumulates until the residual melt evolves to sufficient buoyancy to propagate sub-vertical dykes. Volcanics and epizonal plutons related to porphyry-Cu ore in the Tampakan district display trace-element evidence that the melts segregated from high pressure (lower-crustal) cumulates consisting largely of Al-rich augite and hornblende, but little or no plagioclase. Magma chambers in hot lower crust cool very slowly and live long enough to undergo multiple, million-year-scale cycles of magma replenishment and fractional crystallisation and tapping, over the course of which concentrations of "incompatible components" such as H₂O, Cl and SO₃ are passed on through multiple cycles of chamber replenishment and crystallisation and minor discharge and so accumulate to exceptional concentrations relative to major elements (Si₂, AliO₃, Na₂O etc). Successive batches of increasingly H₂O-rich melt leaving the lower crustal chamber began to exsolve a hydrothermal fluid at successively greater depths. Hydrothermal fluids that exsolve at greater depths are denser and more efficient in scavenging Cu from the melt, because the fluid-melt partition coefficient of Cu is extremely pressure sensitive. This study suggests that the transition to metallogenic fertility of magmas at convergent margins is ultimately modulated by compressional stress that induces deep entrapment, build-up to anomalously high water contents and consequent magmatichydrothermal fluid exsolution at deep mid-crustal depths in ascending magmas, and segregation of Cu-rich brines to apical parts of the ascending magma body. The superposition of both porphyry Cu and high-sulphidation-epithermal Cu-Au mineralisation in the Tampakan deposit, and the partial preservation of the host stratovolcanic edifice, allows investigation of the genetic relationship between these two deposit styles and study of the uppercrustal palaeohydrology of a stratocone-centred, ore-forming magmatic-hydrothermal system. ⁴⁰Ar-³⁹Ar dating reveals that the porphyry Cu mineralisation formed during the early Pliocene (4.24 ± 0.02 Ma, 4.26 ± 0.02 Ma), whereas high-sulphidation-epithermal mineralisation formed during the middle Pliocene (3.23 ± 0.03 Ma, 3.34 ± 0.05 Ma, 3.28 ± 0.06 Ma). The ~1 Myr age difference requires their formation from separate magmatic-hydrothermal systems that were established in the upper crust from different batches of melt. Petrochemical trends indicate that both hydrothermal systems emanated from epizonal magma chambers fed from a shared, longlived lower crustal magma reservoir. Erosion of Cycle 3 andesites during collisional uplift exposed porphyry-Cu-stage quartz stockwork veins at the palaeosurface in less than ~ 350 Kyr after porphyry mineralisation. After unroofing of the porphyry system, construction of the Cycle 4a middle Pliocene volcanic centre commenced at 3.93 Ma. A cryptic unconformity between Cycle 3 and Cycle 4a andesites became the principal surface for a stratigraphic groundwater aquifer that acted as a condensor for high-sulphidation-stage magmatic volatiles. Three aspects of the Tampakan high-sulphidation-epithermal palaeohydrological system are investigated: 1) the physical properties and hydrological transport mechanics of the magmatic supercritical fluids along the "magmatic vapor plume" from the site of accumulation in the carapace beneath the deposit to the meteoric- and magmatic-fluid mixing environment within the deposit; 2) identification of the composition and thermal properties of the fluid end-members and the geometry of mixing paths within the deposit; 3) the geometry and relative mixing ratios of magmatic and meteoric groundwater in various regional alteration zones of the district and the effect of topographic forcing of hybrid hydrothermal fluids along the western flank of the volcanic complex. The Tampakan high-sulphidation epithermal mineralisation formed from a dense vapor-like supercritical fluid with a density of ~ 0 .15 to 0 .25 g/ cc that exsolved from a relatively mafic andesitic melt emplaced at shallow depths of 2.6 km to 4 km. These melts had significantly less magmatic water (~3-4 wt.% H20) than the more evolved andesitic melts associated with precursor porphyry Cu mineralisation (~ 6.0 wt.% H20). The lower water content of highsulphidation- stage melts allowed shallower crustal emplacement and fluid exsolution as a low-density vapor, whereas more water-rich porphyry-stage melts from the preceding cycle exsolved a dense supercritical brine phase (0.3 to 0.45 g/cc) at deeper (~ 6 to 8 km) crustal levels. Pressure and enthalpy constraints calculated for the high-sulphidation-stage magmatic fluid at several points along its flow path provide substantial insights into the magmatic fluid transport process. The magmatic vapor ascended along a nearly isochoric decompression path from the site of exsolution to the site of fluid mixing. The density of the vapor increased from ~0.2 g/cc to ~0.3 g/cc over a vertical ascent distance of ~1.2 km. During transit, the vapor cooled conductively by ~350°C. The nearly isochoric vapor transport mechanism through the lithostatically pressured, ductile rock column requires propagation of fluid-filled, fine-scale, migratory hydrofractures, with intimate contact between the vapor and the ductile wall rocks during vapor ascent. This ensured substantial conductive cooling (~875°C to ~525°C) along the ascent path and that thermal contraction of the vapor balanced the tendency to expand with decompression. Instantaneous isoenthalpic decompression of the magmatic-vapor-charged mobile hydrofractures at the lithostatic-hydrostatic interface (brittle/ductile transition) near the base of the deposit, was associated with "instantaneous" cooling of the supercritical vapor from ~525°C to ~375°C. This pressure-temperature quenching efficiently condensed magmatic vapor to a modestly saline (5 wt.% NaCl equivalent) condensate that concurrently mixed with ambient meteoric water within a palaeo-aquifer at the base of the hydrostatic regime. Cooling of the dense magmatic condensate liquid (~0.62 g/cc) by dilution in the mixing column was associated with hydrolysis of SO₂ to H₂SO₄ , HSO₄ , SO₄ and to H₂S which in turn produced a vertical pH gradient and a vertical textural zonation in alteration facies in the advanced-argillic lithocap. Oxygen-isotope and enthalpy balances indicate that sericite in the deep portions of the deposit and pyrophyllite at higher and peripheral regions precipitated from hybrid magmatic-meteoric waters which comprised ~50% magmatic condensate. The hot, hybrid fluids formed a thermally buoyant plume due to transfer of heat from the high-enthalpy magmatic vapor into the meteoric water regime. The plume ascended and became entrained into a stratabound aquifer system on the west slope of the volcano. A substantial hydraulic head in the aquifer is implied by downstratigraphic- slope deflections in the time-integrated proxy fluid isotherms identified by calibration of PIMA II™ infrared spectral parameters with the chemical composition of potassic white mica. These calibrations reveal chemical trends in the composition of potassic white micas that can be tracked across several alteration environments. A central, and deep-seated, hightemperature zone of nearly stoichiometric muscovite coincides with the locus of the inferred magmatic vapour plume. This zone is transitional to shallower and peripheral regions where there is an increasing replacement of K ions by neutral H₂O molecules in the potassic white mica crystal structure, and decreasing Cu and Au grades. These trends reflect a central, deep-level zone of high fluid temperatures, with cooling paths deflected down-palaeo-slope at shallower levels in the volcanic edifice. Substantial magmatic fluid ascended into the hydrostatic regime along a 5 km by 1.5 km wide NNE-trending fault zone that partly controlled mineralisation. Lateral outflow of the hybrid fluids was controlled by regional dilational faults that transect the volcanic centre. Zoning of hydrothermal mineral compositions and assemblages reveal a superb example of hydrothermal plume-groundwater interaction and downslope dispersion. The plume of heated meteoric water and admixed magmatic condensate in the hydrostatic environment was centred within the Tampakan deposit. The deposit is located where gradients in the hybrid fluid's temperature proportion of magmatic fluid are greatest. Mineralisation was localised in the zone of steep temperature and pressure gradients associated with the interface between a deep lithostatic-pressured plume and a shallow hydrostatic-pressured plume.
|Collections||Open Access Theses|
|b21810059-Rohrlach, B.pdf||361.15 MB||Adobe PDF|