- © The Mineralogical Society Of America
Thermochronology finds many applications in economic geology. Utilizing temperature-sensitive radiometric dating techniques to reveal low-temperature, upper crustal processes can elucidate many aspects of deposit genesis, including timing and duration of mineralization processes, rate of exhumation and erosion of intrusive ore deposits and comparative preservation potential. The tools are utilized to best advantage when combined with other thermochronometry techniques that provide complementary information. In addition, when thermochronometers are combined with geochronometers (e.g., zircon U/Pb), over 800 °C of thermal history is revealed from emplacement to erosion (Fig. 1⇓). With the advent of computational algorithms that provide more accurate and detailed models of thermochronology results, the economic geologist has a powerful tool to use when assessing economic favorability of a region or prospect.
This chapter summarizes the various ways that low-temperature thermochronometers have been utilized in studies of economic mineralization and primarily focuses on studies of porphyry ore deposits. These deposits are well characterized and provide a good platform from which to demonstrate the extended understanding of ore formation processes that is provided by thermochronology studies.
THERMOCHRONOLOGY AND MINERALIZED SYSTEMS – AN INTRODUCTION
In this section we review the fundamental application of (U-Th)/He, fission track and 40Ar/39Ar thermochronometry methods to mineralized systems. We then present a synopsis of how these techniques can be used in thermal history analysis, either alone or in combination with other chronometers, in order to provide more information on deposit genesis. The discussion is presented from low through to higher temperature thermochronometers, corresponding to their presentation on Figure 1⇓.
As documented earlier in the volume, (U-Th)/He thermochronology is based on measuring the accumulation of radiogenic 4He produced from U and Th decay. The daughter helium is retained until the mineral is heated to a temperature at which its structure and helium retentivity change. This “closure temperature” varies from mineral to mineral, providing a powerful way to track the low temperature thermal history of an ore deposit if more than one mineral phase from a single sample can be dated (Fig. 1⇓). However, the closure temperature for a given mineral is dependant on a series of assumptions regarding grain size and shape, chemical zonation and cooling rate (e.g., apatite has a closure temperature of 75 °C assuming a cooling rate of 10 °C/m.y.) and interpretation of the age depends on the complexity of the cooling history (how long the mineral was held at a given temperature). This caveat on thermochronometry interpretation is particularly relevant to apatite (U-Th)/He ages and apatite fission track ages where low temperature processes (tectonic or erosional) can result in samples spending long periods in the helium partial retention (Wolf et al. 1998) or fission track partial annealing zone (Fleischer et al. 1975). Careful interpretation of thermochronology ages is critical (Wolf et al. 1998) as an age may reflect: (i) rapid monotonic cooling of the sample over a very short interval on a geological timescale. In this case, the (U-Th)/He age reflects cooling of the sample through its closure temperature and can often be related to a particular geological event (Dodson 1973); or (ii) slow monotonic cooling (or more complex nonmonotonic net cooling) over a long time interval with the final age not necessarily related to any specific geological event. A good example is the use of apatite fission-track and (U-Th)/He techniques in areas with low denudation rates (as is the case for some porphyry deposits in South America and Carlin deposits in Nevada) where rocks may spend prolonged periods in a partial annealing/retention zone. Interpretation of thermochronometry ages is reviewed in more detail in other chapters of this volume (Donelick et al. 2005; Dunai 2005; Gallagher et al. 2005; Harrison and Zeitler 2005; Ketchum 2005).
The most frequently utilized mineral in (U-Th)/He thermochronometry is the common accessory phase, apatite (see comprehensive reviews by Farley 2002 and Ehlers and Farley 2003). Apatite makes an ideal (U-Th)/He chronometer because it is usually well crystallized, enriched in U and Th, commonly >60 μm in diameter and transparent (making it relatively easy to detect the presence of fluid and mineral inclusions). Apatite is also ideal for other thermochronometry applications (e.g., fission track dating) and is therefore useful for cross-calibration purposes and complementary low-T history studies. Due to the low closure temperature for He in apatite (75–100 °C; Zeitler et al. 1987; Wolf et al. 1996; Farley et al. 1998; House et al. 1998; Farley 2002 and references therein), it is generally assumed (within limits which will be discussed in a later section) that the age documents the time when rocks pass through the upper 1–3 km of the crust. It follows that the apatite (U-Th)/He age can help constrain the post-mineralization uplift (Fig. 1⇓) and exhumation history of a deposit (McInnes et al. 1999) with implications for ore preservation and supergene enrichment processes (McInnes et al. 2003). Another potential use for (U-Th)/He in apatite is to define the timing and scale of low temperature hydrothermal systems like the Carlin-type gold deposits.
While the application of (U-Th)/He thermochronometry to fluorite is in its preliminary stages, the closure temperature of vein fluorite from Yucca Mountain, Nevada has been determined to be 80–100 °C based on a 10 °C/m.y. cooling rate and a 200–300 μm diameter (see discussion in Evans et al. 2005). This suggests potential applications for constraining the low-temperature history of hydrothermal ore deposits. Fission-track ages of fluorite mineralization from late hydrothermal veins in Norway have been used to constrain the minimum age of Late Cretaceous/Early Tertiary hydrothermal activity (Grønlie et al. 1990) although the uncertainty on the ages ranges from 35–55%.
While the first attempts at zircon (U-Th)/He thermochronometry were performed early last century (Strutt 1910a,b), more recent studies (Reiners et al. 2002; Tagami et al. 2003; Reiners et al. 2004) have recognized the value of thermally-dependant helium diffusion from zircon to thermal history studies. Reiners et al. (2004) have identified the closure temperature as being between 171–196 °C. Because zircon is resistant to weathering and is found in magmatic, metamorphic and pegmatitic settings, zircon (U-Th)/He data has wide-ranging application in mineralized systems. For example, phenocrystic zircon (U-Th)/He dating of a porphyry system records the cooling of the rock below 200 °C which corresponds to lower temperature alteration (e.g., argillic alteration stage). As the solubility of Cu in hydrothermal fluids is limited below 200 °C, high temperature Cu transport and mineralization can be constrained as occurring between the magmatic (U/Pb and 40Ar/39Ar) and zircon (U-Th)/He ages.
Oxide (magnetite, hematite, rutile) (U-Th)/He.
Since Fanale and Kulp’s (1962) early work on magnetite, the development of (U-Th)/He methods for Fe and Mn-oxides has been primarily aimed at dating magnetite-hematite and base metal vein mineralization (Lippolt and Weigel 1988; Wernicke and Lippolt 1992, 1994a, 1997; Lippolt et al. 1993). While the effect of grain size and compositional variations on diffusion have not been investigated fully, closure temperatures in the range of 180–250 °C for large (5mm) specularite grains and >90–160 °C for botryoidal hematite (>10 μm diameter) have been suggested (Bähr et al. 1994). If the diffusion characteristics of iron oxides can be better understood, numerous applications can be envisioned in iron ore exploration and in understanding the thermal and depositional histories of weathered and altered deposits.
Preliminary development of rutile (U-Th)/He thermochronometry (Crowhurst et al. 2002) has suggested a closure temperature of >180–200 °C. Further development of this chronometer is desirable as it presents yet another resistate mineral with a high closure temperature that can be utilized in metamorphic regimes. In mineralized systems, rutile is a common hydrothermal alteration product of titanite. Preliminary work dating rutile and zircon from the Darrezhar porphyry Cu prospect in Iran yielded identical (U-Th)/He ages within error, (17.0 ± 0.63 Ma and 16.0 ± 0.68 Ma) supporting the predicted similarity in rutile and zircon He closure temperature (B. McInnes, unpublished data).
As described in earlier chapters (Donelick et al. 2005; Tagami and O’Sullivan 2005), fission track dating is based on the accumulation of radiation damage tracks from spontaneous nuclear fission of 238U within mineral lattices (Gleadow et al. 2002). The thermally induced annealing of the tracks in minerals such as apatite and zircon provides the basis of the use of fission track dating as a mineral thermochronometer and quantitative predictive models of the temperature dependence of annealing significantly advanced the field (e.g., Crowley 1985; Gleadow et al. 1986; Laslett et al 1987; Duddy et al. 1988; Corrigan 1991; Lutz and Omar 1991; Gallagher 1995; Yamada et al. 1995; Ketcham et al. 1999, 2000; Green et al. 1999). In apatite, the process of fission track annealing is thermally resilient relative to He diffusivity (helium partial retention zone lies at temperatures 35 °C cooler than the analogous fission track partial annealing zone; Wolf et al. 1998) and therefore a slowly cooled sample ( 10 °C/m.y.) should have a fission track age older than its (U-Th)/He age. One method is often used to corroborate the other. For example, excess helium might be an issue if the apatite (U-Th)/He age is greater than the apatite fission track age. Because fission track lengths in apatite are shortened within the helium partial retention zone, apatite (U-Th)/He ages can be used to test track-length thermal models.
Fission tracks in apatite will anneal above about 125 °C (about 4–5 km depth under a normal geothermal gradient), producing an altered distribution of track lengths. A field based estimate of the zircon fission track closure temperature for a cooling rate of 15 °C/m.y. is 240 °C (zircon from Gold Butte, Nevada; Bernet et al. 2004). Whereas absolute ages can be obtained from fission track analysis, this is only possible for samples that have cooled rapidly and have remained undisturbed at, or close to the surface since formation (Gleadow et al. 2002). More commonly, apparent ages are obtained which reflect aspects of the thermal history (uplift, denudation). In terms of mineralized systems, fission track dating has long been used in conjunction with other techniques to resolve the timing of mineralization (Fig. 1⇓) and geodynamic setting of ore deposits (Banks and Stuckless 1973; Lipman et al. 1976; Shawe et al. 1986; Koski et al. 1990; Naeser et al. 1990; Arne 1992; Hill et al. 2002; Arehart et al. 2003; Suzuki et al. 2004), identify potential buried stocks and potential associated ore deposits (Naeser et al. 1980; Cunningham et al. 1984; Steven et al. 1984; Beaty et al. 1987), constrain the timing and amount of exhumation and erosion (Maksaev and Zentilli 1999), aid in resolution of the sense and amount of fault offset (McInnes et al. 1999) and reveal the thermotectonic history of petroleum basins (Sutriyono 1998; Osadetz et al. 2002).
Based on the natural decay of 40K to 40Ar and the induced decay of 39K to 39Ar, 40Ar/39Ar thermochronometry assumes that in the case of partial radiogenic Ar loss, some domains in mineral grains remain unaffected and that excess Ar incorporated during mineral formation, has a different distribution in the crystal than the Ar produced by in situ 40K decay (McDougall and Harrison 1999; Williams 2004). Similar to the (U-Th)/He system, the highly predictable time dependant nature of radioactive decay, in conjunction with the temperature dependence of Ar diffusion can be used to reveal the thermal history of the sample (McDougall and Harrison 1999). Unique to 40Ar/39Ar is the step heating technique, where the irradiated mineral is progressively and incrementally heated in order to release the contained Ar. This process accounts for both excess Ar and Ar loss and can yield both the minimum age of crystallization (high temperature steps) and the timing of thermal events (lower temperature steps) on a single sample (e.g., Tomkins et al. 2004).
The recent improvement in precision of isotope ratio measurement and low sample size requirements permit 40Ar/39Ar techniques to decipher overprinted hydrothermal and igneous events that are not typically resolved by K-Ar methods (Reynolds et al. 1998). By analyzing a suite of commonly occurring K-bearing minerals (hornblende, micas, potassic alkali feldspars) with a range of diffusion properties (closure temperatures), it is possible to elucidate the timing of thermal events and cooling rates (e.g., Snee 1999) over a higher temperature range than the (U-Th)/He system alone (>200 °C). In combination with other thermochronometry methods (commonly fission track and apatite (U-Th)/He), even more information is obtained. Some applications of 40Ar/39Ar thermochronometry in mineralized systems include resolving the timing of mineralization (Fig. 1⇓) and alteration/metamorphic events (e.g., Fullagar et al. 1980; Goldfarb et al. 1991; Arehart et al. 1993, 2003; Dammer et al. 1996; Bierlein et al. 1999; Snee et al. 1999; Chan et al. 2001; Mote et al. 2001; Smith et al. 2001; Hill et al. 2002; Lund et al. 2002; Wilson et al. 2003; Mauk and Hall 2004) and documenting the duration of hydrothermal systems (e.g., Silberman et al. 1979; Arribas et al. 1995; Henry et al. 1995; Marsh et al. 1997; Maksaev and Zentilli 1999; Garwin 2002).
Using thermochronometry in thermal history studies
There are several ways that thermochronometry has been applied to exploring the thermal history of mineralized systems. The first involves use of a single chronometer and can provide meaningful data under some conditions. Of more value is the use of multiple chronometers that can potentially reveal the full thermal history of the region. Finally, modeling of thermochronometry data adds value to thermal history analysis by refining the interpretation of the age data.
A single thermochronometer will yield an apparent age but interpretation of the significance of that age will require independent geological evidence of the cooling history. For example, an apatite (U-Th)/He age of 60 Ma might record a short-term, low temperature hydrothermal pulse which reset the “helium clock” 60 m.y. ago, or it may suggest that the apatite has experienced a more complex history and has been held within the helium partial retention zone for some extended period. More evidence would be required to ascertain which history is more likely (see Wolf et al. 1998 and Farley 2002 for more discussion). Similar scenarios have been discussed for fission track ages (e.g., Green et al. 1989; Gallagher 1995).
Diffusion experiments on fluorite, apatite, titanite, zircon and rutile have determined a progressive increase in minimum He closure temperature (Tc) from ~75 °C to ~200 °C (Crowhurst et al. 2002; Farley 2002; Reiners et al. 2002; Evans et al. 2005). By combining U/Pb (Tc >900 °C; Lee et al. 1997; Cherniak and Watson 2000), K-Ar, Re-Os (Tc range from 300–500 °C; McDougall and Harrison 1999; Suzuki et al. 1996), fission track and (U-Th)/He techniques, an interval of over 800 °C of thermal history of an ore deposit or mineral district can be elucidated. It should be noted that one may still have a problem deciding on likely cooling paths at low temperature and hence what event is represented by the age determined using low temperature thermochronometers.
Rahl et al (2003) demonstrated for the first time that a single grain of zircon could be analyzed by excimer laser ablation inductively coupled mass spectrometry (ELA ICP-MS) to obtain a U/Pb age and then analyzed by conventional (U-Th)/He methods to yield two ages on a single crystal. McInnes et al. (2003a, b, 2004, 2005) have performed apatite (U-Th)/He, zircon (U-Th)/He and U/Pb dating on samples from porphyry deposits in Iran, Chile and Indonesia in order to investigate their genesis, exhumation and preservation potential. Advantages of this approach are that:
a single radioactive decay scheme is utilized U + Th → Pb + He,
apatite and zircon are both usually obtainable in significant quantities for analysis from a 1 kg sample of igneous rock,
apatite and zircon are stable in all hypogene alteration assemblages in porphyry deposits,
the coupled use of zircon U/Pb and zircon (U-Th)/He dating determines both the emplacement age of the porphyry deposit and the interval where the bulk of base metal transport and deposition occurs in magmatic-hydrothermal systems (750–200 °C),
the coupled use of zircon and apatite (U-Th)/He dating determines the post-mineralization uplift and exhumation history of the deposit with implications for ore preservation and supergene remobilization (200–90 °C).
Time-temperature curves are one way of graphically displaying the data outputs from multiple thermochronometers because all the radiometric ages are tied to nominal closure temperatures. Multiple age determinations produce time-temperature curves delineating the thermal history of a porphyry deposit from the time of its emplacement to the time of its thermal decline to ambient conditions. Graphical analysis of combined age-temperature data show a range of cooling profiles varying from a subvertical line for a rapidly cooled intrusion emplaced in the upper 1–2 km of the crust (see right inset in Fig. 2⇓) to a “hockey stick” pattern for a slowly cooled pluton emplaced at depths greater than 2–3 km in the crust (see left inset in Fig. 2⇓). Using the “hockey stick” as a process model for the thermal histories of porphyry copper systems, the handle of the stick (represented by higher temperature chronometers like zircon U/Pb and K-Ar) constrains the high-temperature history of the pluton and its associated Cu ore shell, while the blade of the stick (zircon (U-Th)/He and apatite (U-Th)/He chronometers) constrain the post-mineralization uplift and exhumation history of the deposit with implications for ore preservation and supergene remobilization.
Modeling of thermochronometry data.
Using numerical models to derive potential cooling histories directly from thermochronometry data provides researchers with a powerful tool, particularly when modeling is based on multiple chronometers. As discussed earlier, modeling of apatite and zircon fission track data is particularly advanced (e.g., Crowley 1985; Gleadow et al. 1986; Laslett et al 1987; Duddy et al. 1988; Corrigan 1991; Lutz and Omar 1991; Gallagher 1995; Yamada et al. 1995; Green et al. 1999; Ketcham et al. 1999, 2000) and provides a means to assess denudation histories, timing of hydrothermal activity and the low temperature cooling history of igneous intrusions. Modeling of (U-Th)/He ages has progressed in the past few years (e.g., Wolf et al. 1998; Ehlers et al. 2003 and references therein) and now allows quantification of a number of parameters related to the dynamic processes of magmatic-hydrothermal cooling, exhumation and erosion of igneous intrusions (e.g., Fu et al. 2005). Further detail on forward and inverse modeling is provided in other chapters (Dunai et al. 2005; Gallagher et al. 2005; Ketchum 2005).
One such inverse modeling package (4DTherm v.1.1; Fu et al. 2005) is used later in this work to analyze multiple age data (primarily (U-Th)/He and U/Pb) obtained on selected Indonesian and Iranian porphyry deposits. Physical parameters and assumed initial conditions for porphyry deposits are listed in Table 1⇓ and as the detailed description of the model will be published after this volume goes to press, background information on the modeling technique is provided in Appendix II. Although future versions of 4DTherm will include advective and convective cooling scenarios, the current version inverts thermochronometry data assuming conductive heat transfer, which is the least efficient heat transfer mechanism. The outputs of the modeling runs provided in this paper should, therefore, be considered as end members. For example, if the depth of porphyry emplacement is constrained by the model to be 5 km, then that result is a minimum depth.
4DTherm v.1.1 treats the cooling history of igneous bodies from their assumed emplacement temperature at 1000 °C to an ambient surface temperature of 10 °C. Throughout the cooling history, two distinct phases were defined (Table 2⇓): (i) Magmatic-hydrothermal cooling begins at intrusion emplacement and continues until both igneous and country rocks reach a final thermal equilibrium under a steady-state geothermal gradient. This phase is shown as R1 on the schematic time-temperature plot (Fig. 3⇓). Initial cooling is rapid but towards the end of the magmatic-hydrothermal cooling stage, the igneous and country rocks cool more slowly until both reach a final thermal equilibration and the geothermal gradient returns to pre-intrusion thermal conditions (defined as the “cooled” state), (ii) Exhumation cooling (R2 in Fig. 3⇓) begins when the intrusion reaches the “cooled” state and continues until the body reaches the surface (at 10 °C). The rate of cooling through this stage is primarily controlled by exhumation and erosion processes. The hypogene deposition of most economic minerals (e.g., Cu, Au) mainly occurs during the early magmatic-hydrothermal cooling stage while supergene remobilization and precipitation processes occur mainly during the late exhumation cooling stage.
APPLICATIONS OF THERMOCHRONOMETRY TO GOLD MINERALIZATION
Carlin-type gold deposits
In order to better understand the timing of mineralization and hydrothermal events in Carlin deposits, fission track, (U-Th)/He and 40Ar/39Ar methods have all been applied to studies of the sediment-hosted gold deposits (Carlin-type) in the Great Basin of western North America (Arehart et al. 1993, 1995, 2003; Ilchik 1995; Chakurian et al. 2003; Hickey et al. 2003). These giant deposits typically consist of finely (submicron) disseminated gold within hydrothermal arsenian pyrite, hosted in altered, silty carbonate rocks. Ascertaining the timing of gold mineralization is critical to the development of genetic models that aid exploration targeting.
Rb/Sr dates on galkhaite and 40Ar/39Ar on hydrothermal illite in dykes and on magmatic biotite in syn-mineral dikes establish that the Carlin-type deposits of the Carlin Trend, Jerritt Canyon district, Getchell district are Eocene in age (Hofstra et al. 1999; Hofstra and Cline 2000; Ressel et al. 2000; Hickey et al. 2003). The reliability of 40Ar/39Ar methods for determining the mineralization age of altered rocks associated with sediment-hosted gold deposits has been questioned due to problems of partial resetting of sericite ages after heating by hydrothermal fluids below the sericite closure temperature (Arehart et al. 2003; Chakurian et al. 2003). The most reliable apatite fission track data on several deposits in the Carlin Trend indicate mineralization ages of between 33–42 Ma (Hofstra et al. 1999; Ressel et al. 2000; Arehart et al. 2003; Hickey et al. 2003). Two apatite (U-Th)/He ages on the Carlin East deposit were significantly younger than the fission track ages (31.0 ± 1.9 Ma and 21.4 ± 1.3 Ma) and may be the result of thermal resetting associated with Miocene volcanism. Landscape restoration based on thermal modeling of apatite fission track data suggests that Carlin deposits in the northern Carlin trend were emplaced at paleodepths of between <2–3km for geothermal gradients of 20–30 °C/km, (Cline et al. 2005).
Thermochronology has helped resolve the genetic relationship between the Barney’s Canyon and Melco disseminated gold deposits and the giant Bingham Canyon porphyry Cu-Mo-Au deposit in Utah. K-Ar and 40Ar/39Ar ages for Bingham Canyon mineralization range from 37 to 40 Ma (Warnaars et al. 1978, Kendrick et al. 2001; and Parry et al. 2001), whereas Arehart et al. (2003) reported an apatite (U-Th)/He age of 34.9 ± 1.8 Ma, indicating rapid cooling and relatively shallow emplacement (<4 km). Apatite from sandstone formations hosting gold at the Barney’s Canyon deposit, 8 km distal from Bingham, was found to have a (U-Th)/He age of 33.1 ± 1.6 Ma. This suggests a genetic relationship whereby auriferous low temperature fluids driven by the Bingham Canyon thermal system precipitated gold in sedimentary host rocks at Barney’s Canyon. This was confirmed in a major paleothermal study (Cunningham et al. 2004) that also demonstrated that apatite fission track ages in Permian sediments proximal to Bingham Canyon were completely reset whereas zircon tracks were not significantly annealed.
Epithermal gold deposits
Thermochronology studies can be used in conjunction with geochronology methods (refer to discussion in Reiners et al. 2005 to understand the distinction between the two) to determine the temporal relationship between mineralization and magmatism and to explore the thermal history of epithermal deposits. For example, the Porgera epithermal gold deposit in Papua New Guinea was studied by hornblende 40Ar/39Ar, roscoelite 40Ar/39Ar and biotite K-Ar methods to constrain the timing of mineralization to within 0.1 m.y. of emplacement of associated igneous intrusions at around 6 Ma (Richards and McDougall 1991; Ronacher et al. 2002).
Belhadi et al. (1999) combined zircon fission track ages with previously published whole rock, biotite and hornblende K-Ar thermochronometry (Sawai et al. 1998) at quartz vein-hosted gold deposits in the Hoshino gold region, Japan. The results indicate that volcanic activity began at 4.3 Ma with a second phase of activity extending from 3.5 Ma to 2.6 Ma. The zircon fission track ages (2.8 Ma) from the youngest volcanic unit (Takeyama andesite) were within analytical uncertainty of whole rock K-Ar ages from hydrothermally altered samples associated with gold deposits (Sawai et al 1998). The authors postulate that the eruption of the Takeyama andesite was responsible for the hydrothermal alteration and that field relationships also suggest a contemporaneous relationship between the Takeyama andesite and gold metallogenesis.
Archean lode gold deposits
Archean lode gold deposits occur in complex metamorphosed terranes, and the majority of geochronology research has been focused on unraveling the timing of mineralization and deposit genesis. The application of low temperature thermochronometers to the study of the exhumation history of these systems is still in the early stages. Kent and McDougall (1995) applied 40Ar/39Ar to hydrothermal muscovite samples from the Kalgoorlie gold field (located in the Archean Yilgarn block, Western Australia) in order to determine the timing of various styles of gold mineralization (see also Witt et al. 1996 and Kent and McDougall 1996). The gold stockwork system at the Mount Charlotte deposit was formed at least 10 m.y. before that at Golden Mile suggesting separate hydrothermal episodes formed the stockwork and shear-hosted mineralization. However, Kent and McCuaig (1997) highlight the need for caution when interpreting 40Ar/39Ar ages in hydrothermal ore deposits where regional post-mineralization fluid movement may result in Ar loss. In another study (Napier et al. 1998), amphibole and biotite 40Ar/39Ar step-heating was utilized to reveal the complex, post-metamorphic thermal history of the Southern Cross area of the Yilgarn. The results suggest that after main gold mineralization (~2620 Ma), temperatures remained at 500 °C for 20–70 m.y. suggesting longer-lived regional tectonic activity than previously predicted. Tomkins et al. (2004) provide a time-temperature history for the Challenger Deposit, Gawler Craton, South Australia based on an integrated study of U/Pb, Sm-Nd, Rb-Sr and 40Ar/39Ar ages. Three thermal events spanning over 1800 Ma were identified including a very low temperature event (~150–200 °C) at about 1531 Ma.
Shale-hosted lode gold deposits
Muscovite, illite and amphibole from argillite-hosted vein gold deposits of the Meguma Terrane, Nova Scotia were studied by 40Ar/39Ar to reveal the timing of vein formation (ca. 380–405 Ma), which is about 10–15 m.y. after the peak of regional metamorphism (Kontak et al. 1998). Combined fluid inclusion and thermochronology data led the authors to conclude that the gold was precipitated from metamorphic hydrothermal fluids at 400–450 °C in response to cooling. The auriferous fluids are interpreted to have originated from a reservoir at 18–20 km depth, and gold precipitated at depths of around 12 km as the fluids hydrofractured their way through the crust. Studies of similar deposits in the Lachlan Fold Belt of Australia used 40Ar/39Ar to resolve two distinct phases of gold mineralization at about 460–440 Ma and 380–360 Ma (Arne et al. 1998; Foster et al. 1998; Bierlein et al. 1999).
APPLICATION OF THERMOCHRONOMETRY TO PORPHYRY COPPER-MOLYBDENUM-GOLD MINERALIZATION
Reynolds et al (1998) define porphyry deposits as “large (1–5km diameter), low grade (typically <1% Cu) copper (± Mo, Au, Ag) concentrations in which sulfide minerals occur disseminated in a network of veinlets and breccias distributed within, and more or less concentrically around supracrustal porphyritic stocks of intermediate to felsic composition”. The cooling rate in the porphyry system changes with time in response to the various processes operating over the life of the system. For example, initial rapid cooling (800 °C → 350 °C) occurs during post-intrusion rupture of the rock column and is typically followed by more moderate cooling and thermal collapse of the system (see “hockey-stick” pattern in left inset, Fig. 2⇓). The timing of this cooling is constrained by the U/Pb and K-Ar or 40Ar/39Ar ages. Early cooling of the magmatic fluid plume is largely a result of decompression (from lithostatic to hydrostatic conditions) and phase separation (producing coexisting brine and vapor-rich fluid inclusions) (e.g., Ulrich et al. 2001; Harris et al. 2003). These cooling fluids are responsible for the formation of potassic alteration assemblages and the deposition of the bulk of the mineralization in these deposits. Phyllic and argillic alteration results from magmatic and/or mixed magmatic-meteoric fluid circulation at temperatures ranging from 350 °C to ~200 °C (Hedenquist and Richards 1998) and may develop episodically over a period of a million years or more (e.g., Chuquicamata; Reynolds et al. 1998). The zircon (U-Th)/He age probably records the timing of the lowest temperature hydrothermal event in the deposit whereas the apatite (U-Th)/He age records the thermal collapse of the hydrothermal system and/or unroofing of the system.
Porphyry deposits are found worldwide and have been the focus of numerous thermal history studies employing multiple chronometers, originally based on K-Ar and fission track dating and more recently evolving to (U-Th)/He, U/Pb and 40Ar/39Ar methods. This multiple chronometer approach allows discrimination between the intrusion age and the age of later hydrothermal alteration. The (U-Th)/He method is particularly amenable to application in porphyry deposit research, because many of the minerals suited to radiometric dating occur as accessory mineral phases in intrusions known to host disseminated mineralization. In addition, precise low temperature thermochronometers can better resolve the timing of different alteration stages (e.g., potassic versus phyllic) and the duration of the magmatic-hydrothermal system.
Pioneering porphyry thermochronology work was conducted in the southwest U.S.A. by Lipman et al. (1976) and Naeser et al. (1980). Their studies employed K-Ar and fission track dating to demonstrate that mineralization in the San Juan Mountains in the southern Rocky Mountains was episodic, spanning a period of at least 25 m.y.. Cunningham et al. (1987, 2004) suggested that low temperature thermochronometry methods such as fission track dating could be used as a potential exploration tool for porphyry systems under cover. Other work by Naeser et al. (1980) and Larson et al. (1994) predicted the presence of a buried mineralized porphyry system before it was drilled and discovered based on apatite and zircon paleothermal anomalies.
Despite the fact that porphyry deposits have been exploited for 100 years (open pit mining at Bingham began in 1905), a number of time- and temperature-related variables involved in their genesis remain poorly understood:
longevity of the ore precipitation event during the thermal decline of the magmatic-hydrothermal system,
depth of emplacement,
preservation potential of hypogene ores during orogenic uplift and exhumation, and
formation potential of supergene ores from eroded hypogene precursors.
In the following discussion, we review the available data for the world’s largest Cu-Mo porphyry deposit, Chuquicamata, Chile. We present new thermochronology data (Table 3⇓, Fig. 4⇓) for samples from the potassic alteration zone of selected porphyries in Indonesia and Iran where multiple chronometers have been applied to yield complete thermal histories. We utilize inverse modeling of the thermal history data for these deposits to demonstrate how such analysis can enhance our understanding of deposit genesis and conclude this section with a summary of how thermochronology studies can address the issues listed above.
Selected porphyry deposits
The thermal history of the world’s largest porphyry Cu-Mo deposit (15,000 Mt @ 0.71% Cu and 0.01% Mo) at Chuquicamata, northern Chile has been studied extensively and from various perspectives. Ballard et al. (2001) obtained zircon U/Pb ages (excimer laser ablation-inductively coupled plasma-mass spectrometry (ELA-ICP-MS) and sensitive high-resolution ion microprobe (SHRIMP)) and identified two temporally distinct porphyry intrusions, one at 34.6 ± 0.2 Ma and a second emplaced 0.9–1.5 m.y. later (33.3 ± 0.3 Ma and 33.5 ± 0.2 Ma;). The age of the first intrusion (East Porphyry) correlates well with the 40Ar/39Ar and Re-Os ages determined for the earliest alteration (Reynolds et al. 1998; Mathur et al. 2000a). Subsequent intrusions (referred to as the Bench and West porphyries) can be correlated with Cu deposition associated with potassic alteration that has a K-feldspar 40Ar/39Ar age of 33.4 ± 0.3 Ma (Reynolds et al. 1998). The Cu-enriched phyllic alteration zone associated with the West Fault structure was developed at least 2 m.y. after the emplacement of the last igneous intrusion at Chuquicamata (sericite 40Ar/39Ar age of 31.1 ± 0.3 Ma, Reynolds et al. 1998; pyrite Re-Os age of 31.0 ± 0.3 Ma, Mathur et al. 2000) and cooled rapidly below ~200°C as recorded by the apatite fission track and (U-Th)/He ages of near surface samples (about 31 Ma; McInnes et al. 1999).
A unique feature of the Chuquicamata Cu-Mo deposit is that a significant portion of the western side of the ore body was dislocated by post-mineralization motion along the West Fault, a major N-S strike-slip feature that cuts through Northern Chile. The vertical displacement along the fault was determined by a low temperature thermochronology (apatite (U-Th)/He and fission track) study based on comparative age dating of samples taken from both sides of the fault (McInnes et al. 1999). The study revealed a 600 m vertical offset of the western Fortuna block and determined that the “missing” portion of Chuquicamata was probably eroded during tectonic uplift and denudation of the western crustal block.
Grasberg porphyry Cu-Au deposit, Indonesia.
The Grasberg Cu-Au deposit is a giant ore system in Irian Jaya with proven and probable reserve estimates of around 2700 Mt @ 1.08% Cu and 0.98 g/t Au. The economic porphyry is the Main Grasberg Intrusion (MGI), a quartz monzodiorite that intruded the core of the Dalam Diatreme. The final stage of magmatic activity at Grasberg was the emplacement of the dyke-like Kali intrusion. Although U/Pb dating has not been carried out, combined 40Ar/39Ar and (U-Th/He) dating on the same samples from the MGI and Kali intrusions shows extremely rapid cooling with apatite (U-Th)/He ages (2.9 to 3.1 (± 0.1 Ma); McInnes et al. 2004) nearly identical to biotite 40Ar/39Ar ages (from 2.7–3.3 Ma; Pollard et al. 2005). These ages overlap with a sulfide Re-Os age of 2.9 ± 0.3 Ma (Mathur et al. 2000b) and a more recent molybdenite Re-Os age of 2.88 ± 0.01 Ma (Mathur et al. 2005) indicating that ore-related intrusions in the district underwent extremely rapid monotonic cooling as a consequence of shallow emplacement. These studies corroborate earlier interpretations based on apatite fission track dating that the MGI was emplaced within ~1 km of the paleosurface (Weiland and Cloos 1996). Zircon U/Pb analysis at Grasberg is underway but for the purposes of inverse modeling, the zircon U/Pb age (2.97 ± 0.57 Ma; Gibbins et al. 2003) from the nearby Ertsberg intrusive (1 km to the SE of Grasberg) is adopted as a minimum U/Pb age for the MGI.
Batu Hijau porphyry Cu-Au deposit, Indonesia.
The Batu Hijau porphyry Cu-Au deposit is located in SW Sumbawa Island, Nusa Tenggara with proven and probable reserve estimates of around 920 Mt @ 0.55% Cu and 0.41 g/t Au. Mineralization is associated with a multiphase tonalite porphyry complex hosted in quartz-dioritic and andesitic wallrocks (Garwin 2002). Zircon U/Pb and apatite (U-Th)/He ages for the late-mineralization Young Tonalite (collected at 150 m ASL) are 3.74 ± 0.14 (2σ) Ma and 2.23 ± 0.09 (2σ) Ma, respectively (Garwin 2000). Hydrothermal biotite from the Young Tonalite (3 samples collected from 150–350 m ASL) yields a mean plateau 40Ar/39Ar age of 3.73 ± 0.08 (2σ) Ma, which is indistinguishable from the zircon U/Pb age (Garwin 2002). Previous studies using amphibole-plagioclase thermobarometry indicate that the tonalitic magmas began to crystallize at 9km depth with final crystallization and stock emplacement occurring at <2 km (+ 0.5 km) depth and between 710–780 °C (Garwin 2000, 2002).
Ciemas Cu-Au porphyry prospect, Indonesia.
The Ciemas porphyry Cu-Au prospect is located near the southern coast of western Java, about 150 km south of Jakarta. The Ciemas prospect has seen limited drilling (8 diamond holes) and development, but the results to date suggest extensive zones of sub-economic metal grades (~0.2% Cu, ~0.2 g/t Au) associated with a quartz diorite porphyry intrusion hosted principally by andesitic volcanic rocks. Zircon U/Pb, sulfide Re-Os (McInnes et al. 2000) and apatite (U-Th)/He dating (McInnes et al. 2004) has returned ages of 17.8, 15.2 and 7.2 Ma, respectively. It should be noted that there is a large error bar associated with the Re-Os age (Fig. 4⇓) which places some uncertainty on the thermal history. The geochronology and thermchronology of the Ciemas prospect is included to facilitate thermal history comparison between an apparent “failed” porphyry and the other giant porphyry deposits of Indonesia.
Sar Cheshmeh, Iran.
The Kerman Belt, located in southeastern Iran, is an elongated NNW-SSE mountain belt, 500 km long and 100 km wide. It is principally composed of a folded and faulted early Tertiary volcano-sedimentary complex and is bordered to the southwest by a major thrust zone and the Tertiary and Paleozoic sedimentary rocks of the Zagros Mountains (Waterman and Hamilton 1975). The Sar Cheshmeh and Miduk copper deposits (mined by the National Iranian Copper Industries Company, NICICO) are two of the largest known porphyry Cu deposits in the Kerman district. Sar Cheshmeh and Miduk are associated with a high-K calc-alkaline Eocene volcanic arc formed after cessation of subduction of Tethyan oceanic lithosphere at the Zagros suture zone. (Sengor and Kidd 1979). Post-collisional compression and mantle buoyancy forces led to the uplift of the Iranian plateau, with middle Miocene marine sediments occurring at elevations greater than 3000 m in the Kerman belt (Hassanzadeh 1993). The preservation of porphyry Cu deposits in the Kerman belt is therefore dependent on the original depth of emplacement and the rate of exhumation in response to this tectonically driven uplift and erosion.
No previous thermochronology studies have been carried out in the Kerman District. We have determined the zircon U/Pb, zircon (U-Th)/He and apatite (U-Th)/He ages of igneous units from the potassic alteration zone of three Iranian porphyry Cu systems (Table 3⇓, Fig. 4⇓) and have integrated this data with pre-existing whole rock Rb-Sr, biotite K-Ar and 40Ar/39Ar geochronology ages (Shahabpour 1982; Hassanzadeh 1993). These ages constrain the maximum period of longevity of hydrothermal mineralization in porphyry-epithermal environments. The thermal histories for Sar Cheshmeh and Miduk are compared to that of the Abdar Cu-Au prospect hosted within the collapsed and partially eroded caldera of the Kuh-e-Masahim stratovolcano located between the two porphyry copper deposits.
The Sar Cheshmeh Cu deposit (1100 Mt @ 0.64% Cu, 0.03% Mo) is contained within an ovoid (2.5 km × 1 km) Cu shell surrounding a Cu-poor granodiorite to quartz monzonite intrusion known as the Sar Cheshmeh porphyry (SCP) that was emplaced within Eocene to Oligocene volcanic rocks of andesitic composition. The alteration halo and satellite intrusions extend for 7 km. During Cu ore formation the deposit was intruded by 3 intrusions interpreted by Ghorashi-Zadeh (1979) to be fractional crystallization products of the magma that produced the SCP. Cross-cutting relationships define the order of emplacement of the intramineral intrusions as: (1) Late fine porphyry (quartz monzonite), (2) Early hornblende porphyry (dacite) and (3) Late hornblende porphyry dyke swarms (latite). Although these intrusions played a role in redistributing Cu throughout the deposit, their net contribution to ore genesis has been to dilute the initial Cu content of the SCP.
Shahabpour (1982) determined whole rock Rb-Sr and biotite K-Ar ages for the Sar Cheshmeh porphyry of 12.2 ± 1.2 Ma and 12.5 ± 0.5 Ma, respectively. Zircon U/Pb/(U-Th)/He and apatite (U-Th)/He dating was conducted on samples of the SCP and from the open pit in August 2002 (McInnes et al. 2003a,b).
The Miduk Cu deposit (>170 Mt @ 0.82% Cu) is a circular body about 400m in diameter centered over an intrusive quartz diorite stock known as the Miduk porphyry. The main intrusion hosts 90% of the Cu mineralization and is cross-cut by multiple NNE trending dykes called the Miduk fine porphyry. The dykes are of similar composition to the main intrusion and are interpreted as comagmatic (Hassanzadeh 1993).
Zircon U/Pb and zircon (U-Th)/He ages of 12.5 Ma for the Miduk porphyry (McInnes et al. 2003a,b) are essentially identical to mineral-whole rock Rb-Sr ages (12.4 ± 0.5 Ma) reported in Hassanzadeh (1993). However, Hassanzadeh (1993) notes that the Rb-Sr age is considered unreliable due to Rb loss and Sr addition during alteration. Hassanzadeh (1993) also determined 40Ar/39Ar isochron ages of 11.2 ± 0.5 Ma for biotite in potassic alteration assemblages and 10.8 ± 0.4 Ma for sericite in phyllic alteration zones. Similar to Sar Cheshmeh, the progressively decreasing ages for the Miduk porphyry in the U/Pb and 40Ar/39Ar systems reflects the cooling history of the deposit through the temperature interval 750 °C to 300–350 °C, yielding a cooling rate of 250–350 °C/m.y.. Surprisingly, the zircon (U-Th)/He age is identical to the zircon U/Pb age for the same sample of Miduk porphyry and the 40Ar/39Ar isochron ages are younger than the zircon (U-Th)/He age, implying that either: (1) the Miduk sample dated for zircon (U-Th)/He has been emplaced near a contact with cool country rock and therefore has had a more rapid cooling history than the samples in the 40Ar/39Ar study, or (2) that zircon (U-Th)/He in rapidly cooled, high level intrusions acts as a geochronometer, rather than a thermochronometer (see Reiners et al. 2005 for definition). No apatite was recovered from the Miduk samples and in order to facilitate modeling, a value of 9.5 Ma was assigned as the apatite (U-Th)/He age. As Abdar is located only 18km away and the erosion rate in the area is thought to be similar to that at Miduk, the difference between the zircon and apatite (U-Th)/He ages at Abdar was subtracted from the zircon (U-Th)/He age for Miduk resulting in a proxy apatite (U-Th)/He age of 9.5 Ma.
Abdar Cu-Au prospect, Kuh-e-Masahim volcano, Iran.
The Abdar Cu-Au prospect, located approximately 15 km SE of Miduk, is associated with a subvolcanic dioritic intrusion hosted within the partially eroded caldera of the Kuh-e-Masahim stratovolcano (35 km basal diameter, 3500 m asl total elevation, 1500 m above surrounding plateau). Epithermal high-sulfidation Au-Ag-base metal veins are exposed in the caldera peripheral to the Abdar Cu prospect. Reconnaissance scale drilling of the prospect has detected anomalous yet uneconomic concentrations of Cu (values ranging from 0.1–0.25% Cu). Samples for geochronology investigation were taken from potassic alteration zones from mineralized drill core. Similar to Miduk, the zircon U/Pb and zircon (U-Th)/He age data for the Abdar diorite (Table 3⇓) are identical within error. This further supports the suggestion that the zircon (U-Th)/He system acts as a geochronometer for shallow, rapidly cooled subvolcanic intrusions. The relatively young 40Ar/39Ar ages for lava flows on the flanks of the volcano indicate that the feeder conduits did not thermally reset the zircon U/Pb and zircon (U-Th)/He ages of the diorite intrusion. The apatite (U-Th)/He age of 4.9 Ma for the Abdar diorite indicates the time when the subvolcanic intrusion cooled below 90 °C due to caldera collapse and rapid erosion of the overlying volcanic pile.
Duration of hypogene ore formation: measured vs. modeled
It has been shown that the overall duration of magmatic-hydrothermal activity in some porphyry ore deposits (e.g., Divide, Silberman et al. 1979; Far Southeast-Lepanto, Arribas et al. 1995; Potrerillos, Marsh et al. 1997) is within the resolution of K/Ar and 40Ar/39Ar techniques (0.1–0.3 Ma; Sillitoe 2000). Through the use of multiple geochronology methods and the age dating of intrusions that cross-cut mineralization, some studies have shown that intrusion-related hydrothermal mineralization takes place within hundred thousand year time frames: Sar Cheshmeh, ~160 Ka (McInnes et al. 2003); Batu Hijau, ~80 Ka (Garwin 2002); Grasberg, ~100 Ka (Pollard et al. 2005); Lepanto-Far South East, 100–300 Ka. (Arribas et al. 1995); Round Mountain, ~100 Ka (Henry et al. 1997). These findings are consistent with heat flow modeling of cooling in and around the small, high-level intrusions typically found associated with porphyry copper deposits (e.g., Norton and Knight 1977; Norton and Cathles 1979; Smith and Shaw 1979; Cathles 1981; Cathles et al. 1997). It is probable that these magmatic-hydrothermal activity duration estimates are maximum values, taking into account the resolution achievable using radiometric dating methods. Zircon U/Pb and zircon (U-Th)/He ages of intrusion-related ore deposits can potentially be used to constrain the maximum duration of hypogene ore formation because their closure temperatures bracket the magmatic-hydrothermal temperature interval of 750° to 200 °C. Taking into account the uncertainties on the determined ages, the measured age differentials suggest a maximum period of ore deposition of 3 Ma for Sar Cheshmeh, and around 0.5 Ma for Abdar and Miduk (Table 3⇓).
The measured geochronology data (Table 3⇓, Fig. 4⇓), intrusion size estimates and a range of emplacement depths (Table 4⇓) were input into 4DTherm v.1.1 to iteratively reproduce an idealized time-temperature history for each deposit (Fig. 5⇓). Because the solubility of chalcopyrite, the main hypogene ore mineral in porphyry deposits, decreases by over two orders of magnitude in hydrothermal fluids during a temperature reduction from 500–300 °C (McPhail and Liu 2002; Liu and McPhail 2005), it is possible to more precisely estimate the duration of hypogene ore formation for each deposit studied by extracting inverse thermal modeling outputs through this narrow temperature interval (Table 4⇓). The modeled ore formation duration results are consistent with intervals previously determined for deposits for which we have comparative data: Sar Cheshmeh 270 k.y. (this study) vs. ~160 k.y. (McInnes et al. 2003a,b); Batu Hijau 10.5 k.y. (this study) vs. ~80 k.y. (Garwin 2002); Grasberg 15 k.y. (this study, assuming the zircon U/Pb age is similar to the biotite 40Ar/39Ar age) vs. ~100 k.y. (Pollard et al. 2005). Without wishing to put too fine a point on the modeling outputs, those deposits with ore formation durations less than 100 k.y. scale are Grasberg, Batu Hijau, Abdar and Miduk, whereas Sar Cheshmeh falls into an intermediate category (100–1,000 k.y.), and Ciemas took more than 1,000 k.y. to cool through the 500–300 °C interval (Fig. 5⇓).
The main parameters controlling the cooling rate of an intrusion are size; emplacement depth and heat transfer efficiency (conductive vs. advective cooling regime). For similar sized intrusions, deep emplacement and conductive thermal regimes will produce the lowest overall cooling rates whereas shallow emplacement and advective regimes will produce the fastest cooling rates. Raw average cooling rates for the porphyry deposits were calculated (Table 4⇓) based on the age data in Table 3⇓. Of the deposits studied, Grasberg has the fastest (>1,000 °C/m.y.) and Ciemas the slowest (<100 °C/m.y.) raw average cooling rate, with the remaining deposits falling within 100–300 °C/m.y. range (Table 4⇓). Overall average cooling rates calculated by 4DTherm v.1.1 (Table 4⇓) are consistent with the raw results.
As discussed in an earlier section, 4DTherm v.1.1 solves for emplacement depth during cooling in a conductive thermal regime, and therefore the minimum depths of emplacement can be determined for intrusions of a given size. A range of possible emplacement depths along with a best-fit model depth are provided in Table 4⇓, and the results are schematically presented in Figure 6⇓. The model depth of emplacement for the Iranian porphyry deposits indicate that Sar Cheshmeh was emplaced at deeper levels than the Miduk and Abdar intrusions (Table 4⇓, Fig. 6⇓). Geological reconstruction of the Kuh-e-Masahim volcano indicates that approximately 2 km of volcanic cover overlying the Abdar intrusion has been eroded (McInnes, unpublished data), consistent with the minimum possible emplacement depth generated by the model (Table 4⇓). Similarly in Indonesia, model depths of emplacement for Grasberg of 700 m are supported by independently determined geological data that the Main Grasberg Intrusion was emplaced into a volcanic edifice within 1 km of the paleosurface (MacDonald and Arnold 1994; Weiland and Cloos 1996). At Batu Hijau, the modeled depth of emplacement of 2400 m (−400/+600 m) is consistent with paleodepth reconstruction by Garwin (2002) of 2000 ± 500 m. The depth estimate for the Ciemas intrusion is 5500 m below the paleosurface, however, due to the paucity of information for this prospect, this depth cannot be corroborated by other geological data.
Hypogene copper grade as a function of cooling rate
Temperature is one of the fundamental variables controlling the solubility of copper in magmatic-hydrothermal systems (McPhail and Liu 2002; Liu and McPhail 2005) and therefore thermal history analysis may prove useful in understanding processes that produce high-grade hypogene ores. One way to rapidly deposit Cu-sulfide minerals within small rock volumes is to pass a hydrothermal fluid through a steeply declining thermal gradient. In contrast, weak thermal gradients should generate more diffuse haloes of Cu mineralization. A schematic representation of this concept is provided in Figure 2⇓ where Cu transport and heat transfer are treated as diffusive processes. Under these conditions, intrusions emplaced within the uppermost crust should experience greater thermal gradients than those emplaced in mid-crustal regions where temperature regimes are moderated by the Earth’s geotherm. Support for the cooling rate hypothesis can be found in the Indonesian study set. Grasberg, the most shallowly emplaced intrusion-related hydrothermal system with a high average Cu grade of 1.08% experienced the highest rate of cooling of any porphyry system studied. In contrast, the Ciemas porphyry with an average grade of 0.2% Cu experienced the slowest cooling rate over the hypogene temperature interval (almost 200 °C/m.y.) and is interpreted to be the most deeply emplaced intrusion (5.5 km). It took over 1 m.y. for Ciemas to cool through the hypogene Cu window, and it is possible that the low Cu grades for the deposit might be explained by the fact that the original magmatic Cu was distributed over a larger volume of country rock. It is suggested therefore that in the Indonesian examples, Grasberg experienced the thermal gradient conditions of the pluton depicted in the right of Figure 2⇓ and Ciemas represents the pluton on the left. This is also supported by their contrasting patterns in Figures 4⇓ and 5⇓.
Although these preliminary data suggest a correlation might exist between cooling rate and hypogene copper grade in porphyry Cu deposits, it is not known whether thermal history analysis can be applied successfully to mineral exploration. Other factors such as total metal availability, reactivity of wallrocks (e.g., carbonates at Grasberg) and periodic pulses of metal emplacement cannot be assessed through thermal history analysis. More studies of intrusions on a regional scale are needed to increase the data density and definitively assess the relationships between cooling rate and hypogene Cu grade.
Preservation potential of hypogene ores and potential formation of supergene ores
As discussed earlier, 4DTherm v.1.1 treats the cooling of intrusive bodies from magmatic temperatures to surface temperatures (in this study nominated as 10 °C) as a two-stage process: (i) magmatic-hydrothermal cooling (R1 in Fig. 3⇓) and (ii) exhumation cooling (R2 in Fig. 3⇓). Exhumation cooling is controlled by the rate of removal of cover material overlying the sample by tectonic (e.g., extension) and/or erosional processes (e.g., glaciation). If a post-emplacement exhumation rate can be determined for an intrusion, then an assessment of the preservation potential of associated hypogene mineralization can be made. Exhumation rates for the porphyry deposits studied range from 0.26 to 0.72 km/m.y. (Table 4⇓). The highest overall exhumation rate was found at Batu Hijau (0.72 km/m.y.), where a combination of collision-driven Pliocene uplift (Garwin 2002 and references therein) and pluvial processes occurred. Collision-driven uplift is also a feature of the Kerman Belt where the majority of uplift is occurring along the actively deforming Zagros Thrust Zone. The exhumation rates determined for the Sar Cheshmeh, Miduk and Abdar deposits are within a narrow range of 0.3–0.4 km/m.y., presumably because they are located co-parallel and 80 km distant from the Zagros Thrust Zone. Grasberg has a lower calculated exhumation rate (0.37 km/m.y.). Other workers (Weiland and Cloos 1996; Hill et al. 2002) have argued that although rapid denudation (0.7–1.0 km/m.y.) is occurring along major thrust fronts actively forming the New Guinea Fold Belt, these fronts are 50 km distant from the Grasberg deposit and the peak uplift forces have not yet transitioned to the Grasberg area. However, it should be noted that adding the effects of advective and convective cooling to 4DTherm will likely result in increased emplacement depths and hence increased erosion rates, so the 0.37 km/m.y. exhumation rate for Grasberg may be a minimum.
Understanding the rate of exhumation for a mineral district or a metallogenic belt has implications for area selection during mineral exploration. The determination of an exhumation rate permits an assessment of the erosion potential of hypogene ore deposits and an evaluation of the potential for supergene ore formation from eroded hypogene ores. Figure 6⇓ portrays the amount of erosion experienced by each intrusion as a dashed cylinder. In Indonesia, Batu Hijau has experienced the least amount of erosion since exposure whereas Ciemas has experienced the most. Although the deposits of the Kerman belt have similar exhumation rates, differences in their exposure age indicate substantially different amounts of potential hypogene mineralization have been eroded. Taking Sar Cheshmeh as an example, the porphyry copper deposit was exposed 5.2 million years ago and was exhumed at a rate of 0.39 km/m.y. (Table 4⇓), while the erosion rate for the porphyry since exposure is estimated to be about 0.06 km/m.y. (see notes, Table 4⇓). These calculations infer that 312 m of porphyry Cu mineralization have eroded since exposure of the porphyry at surface. Assuming an average copper shell thickness of 100 m and a rock density of 2.7 g/cm3, the total amount of rock eroded from the Sar Cheshmeh porphyry system was about 810 million tons (Mt). The Sar Cheshmeh porphyry was intruded by three copper-poor igneous dike units that account for about 1/3 of the deposit volume, so the total amount of eroded ore is reduced to 540 Mt. At a minimum copper grade of 0.64%, the total amount of copper eroded equates to around 3.5 Mt, which is nearly half of the remaining reserve as estimated in 1998. If the amount of supergene Cu contained at Sar Cheshmeh is less than the amount of Cu eroded, then the unaccounted Cu may be contained in Exotica-type deposits below sedimentary and volcanic cover in the region. Similar calculations can be performed for the other deposits to assess both the preservation potential of hypogene shells and the potential for formation of supergene ore deposits.
CURRENT TRENDS, FUTURE DIRECTIONS
Thermochronology has made a fundamental contribution to economic geology through the provision of data needed to construct genetic models for hydrothermal ore deposits. Mineral exploration has traditionally involved the search for the measurable anomalies in the Earth’s crust produced by hydrothermal systems. Although detection of visual (e.g., Fe-oxide staining or white mica formation), chemical (e.g., trace element enrichments) and physical (e.g., magnetite formation or destruction) anomalies will always be important, economic geologists are increasingly turning to thermochronology data to reveal the paleothermal anomalies produced by hydrothermal systems (e.g., Naeser et al. 1980; Cunningham et al. 1987). Because the search for new ore deposits involves looking deeper below the surface, thermochronology will potentially serve a greater role in testing exploration concepts prior to the expensive stage of drilling.
The power of thermochronology to the explorationist, is the provision of the 4th dimension, vital to the assessment of the metallogenic evolution of a mineral district. The challenge for thermochronologists will be to develop new, low-cost techniques that enable the construction of high-density, temperature-integrated, regional paleothermometry data sets. As outlined in this and other chapters within this review volume, the inversion of thermochronology data to produce 4-D thermal evolution and landscape evolution models is well underway. The coupling of thermal inversion with geochemical and geophysical inversion techniques is still in its nascent stages, but by 2010 the explorationist will be using deterministic, fully coupled thermal-chemical transport models as part of their discovery toolkit.
With respect to porphyry deposit in particular, the following conclusions are offered:
Combining multiple chronometers and in particular apatite (U-Th)/He, zircon (U-Th)/He and zircon U/Pb, provides a thermal history for porphyry deposits over a temperature range of >700 °C. Information that can be obtained from thermal history analysis includes the timing and depth of emplacement of igneous units, the cooling rate during hypogene copper deposition and the exhumation rate of the porphyry deposit.
The disruption of the steady state geothermal gradient during the emplacement of igneous intrusions places limitations on the direct usage of (U-Th)/He age dating in the determination of emplacement depth and exhumation rates. Numerical modeling techniques provide an effective and complementary tool for quantifying cooling and emplacement parameters.
We postulate that strong thermal gradients present the ideal conditions for the generation of high-grade hypogene Cu ores, whereas more diffuse mineralization haloes would be expected for more slowly cooled igneous intrusions. Thermal history analysis using “triple dating” U/Pb-He techniques provides some support for a positive correlation between short duration, rapid cooling from 500 to 300 °C and emplacement depth, however the database is limited and additional district-scale studies are required.
Understanding the rate of exhumation for a mineral district or a metallogenic belt permits an assessment of the erosion potential of hypogene ores, as well as the formation potential of supergene ores from eroded hypogene deposits.
APPENDIX I: U/PB AND (U-TH)/HE ANALYTICAL PROCEDURES
Apatite and zircon grains for (U-Th)/He thermochronology were selected by hand picking in order to avoid U- and Th-rich mineral inclusions that may contribute excess helium. Images of selected grains were recorded digitally and grain measurements were taken for the calculation of an alpha correction factor (Ft; Farley et al. 1996). Helium is thermally extracted from single crystals, loaded into platinum micro-crucibles and heated using a 1064 nm Nd-YAG laser. 4He abundances were determined by isotope dilution using a pure 3He spike, calibrated daily against an independent 4He standard tank. The uncertainty in the sample 4He measurement is <1%. The U and Th content of degassed apatite is determined by isotope dilution using 235U and 230Th spikes. Apatite is digested in 7 M HNO3. Zircon is digested in Parr bombs using HF. Standard solutions containing the same spike amounts as samples were treated identically as were a series of unspiked reagent blanks. For single crystals digested in small volumes (0.3–0.5 ml), U and Th isotope ratios were measured to a precision of <3%. Overall the (U-Th)/He thermochronology method at CSIRO has a precision of 2.5% for apatite, based on multiple age determinations (n = 70) of Durango standard which produce an average age of 31.5 ± 1.6 (2σ) Ma.
Zircon U/Pb dating at Macquarie University utilizes a LA-ICP-MS facility combining a New Wave/Merchantek 213 nm UV laser ablation (LA) system and a HP 4500 ICP-MS, with analytical methods detailed by Jackson et al. (2004). A split sample of zircon grains from the (U-Th)/He study was mounted in epoxy discs and polished to expose the grains. The mounts were examined using back-scattered electron/cathodoluminescence microprobe imaging to record internal zonation features and external morphology prior to selecting grains for analysis. U/Pb geochronology results were based on the analysis of 207Pb/235U, 208Pb/232Th and 206Pb/238U on between 14 and 20 grains per sample. The analysis of the sample zircons was bracketed by multiple analyses of the gem quality GJ-1 zircon, and other in-house standards 91500 and Mud Tank zircon (Black and Gulson 1978; Wiedenbeck et al. 1995) were analyzed in every run as an independent control on reproducibility and instrument stability.
APPENDIX II: EXPLANATIONS AND CALCULATIONS OF MODELED PARAMETERS
1. Sample position, eroded thickness of the porphyry, and initial sample depth
In order to run the algorithm, the position of the dated sample within the porphyry unit must be known or assumed. In most cases, it was assumed that the sample was taken from an exposure or outcrop, which was not the original “top” of the porphyry. If there had been significant erosion since exposure or a portion of the porphyry had been removed by mining, the sample position was somewhat deeper in the body. As shown in Figure 7⇓, the distance from the “top” of the porphyry to the pre-mine topography surface is defined as the “Eroded Thickness” of the porphyry since exposure, a variable we attempt to estimate.
The small circle within the porphyry as shown in Figure 7⇓ represents the position of the samples, and the “Sampling Depth” is defined as the vertical distance between the pre-mine surface and the position of samples. If the samples were taken from the outcrop, the Sampling Depth equals zero; if they were either from the mine pits or from drill holes, the Sampling Depth is some value larger than zero. The position of the samples at the time of emplacement (called Initial Sample Depth for simplicity) is defined as the Emplacement depth + Eroded Thickness + Sample Depth.
In the case of the Sar Cheshmeh Porphyry (SCP), the Sample Depth is set to be 100 m because that about 100 m of porphyry has been previously removed by mining. In this case, the position of the sample from SCP at the emplacement time (Initial Sample Depth) = Emplacement Depth + Eroded Thickness of the porphyry + 100 m.
Iterations of the model indicated that pluton emplacement depth and dimension are the key factors controlling the cooling histories of igneous bodies. It can also be proved that different positions within an igneous body have different cooling histories, although symmetry points may have same cooling histories if the igneous body is of regular shape. So the Initial Sample Depth (sample position) is very important.
2. Determination of emplacement depth
Determination of emplacement depth and calculation of erosion rates are based on the following two assumptions:
Weathering and erosion processes have been occurring at the surface since the intrusion of the porphyry. This infers erosion is occurring during the intrusion and cooling of the body as well as during and after exposure;
The average erosion rate remains constant for a given rock type. In our model, eroded country rocks above the intrusion are assumed to be a single (mixed) rock. We have assumed that the erosion rate for the intrusion is slower than that for the country rock which is assumed to contain sediments and/or sedimentary rock.
For each run of the model, we assign initial values for the Emplacement Depth and Eroded Thickness of stock and then calculate a cooling curve for the sample. If this cooling curve passes through all age data points determined during dating of the sample, this run of the model is considered to be successful, and the Emplacement Depth and Eroded Thickness are potentially valid values. The word “potentially” is used here because there are a number of other pairs of Emplacement Depth and Eroded Thickness values that can also satisfy the model and result in successful cases. This introduces some uncertainty in the final result. Fortunately, the Emplacement Depth can be limited to a certain range. Beyond this range, no matter what value of Eroded Thickness is taken, the modeling will fail. Similarly, the Eroded Thickness can also be limited to a narrow range.
We can further reduce the uncertainties by considering the erosion rates for both country rock and igneous stock separately and by comparing the modeling results with those from other deposits in the region.
3. Calculation of exhumation rates
The apatite (U-Th)/He age was used to calculate a depth (called He Depth for simplicity) which corresponds to the position of the sample at the time of closure for apatite (U-Th)/He. For example, by assuming that the apatite (U-Th)/He age is 1 Ma with a closure temperature = 90 °C, surface temperature = 10 °C, thermal gradient = 50 °C/km, then the depth of the sample at 1 Ma:
So, we can say, under the above conditions, that the sample was at the depth of 1.6 km below the surface at 1 Ma.
The erosion rate before exposure (rate 1) is defined as:
The erosion rate (rate 2) after exposure of the intrusion is defined as:
where Eroded Thickness of Intrusion = Initial Sample Depth (rate 1) × (Emplacement Time − Exposure Age).
The Exposure Age will be generated automatically by the model when the top of the intrusion is exposed and begins to erode.
The calculation of He Depth is based on the assumption that the igneous stock is already “cooled” (as defined in Table 4⇓ and the text) before it passes through the closure temperature of apatite (U-Th)/He. Most deposits in this work fall into this category. If it is not “cooled” (like Grasberg), the calculation is more complicated but follows the same general assumptions as outlined above.
4. Example: determination of emplacement depth and exhumation rate for the Batu Hijau Porphyry
|Thermochronology||Closure temp||Age (± 2σ Ma)||Note|
|Apatite (U-Th)/He||90 °C||2.23 ± 0.09||He Depth = 1600 m|
|K-Ar||400 °C||3.73 ± 0.08|
|U/Pb||750 °C||3.74 ± 0.14||Emplacement Time = 3.74 Ma|
Batu Hijau was modeled as a single composite stock with a cylindrical shape where the width = 500 m and height = 1700 m (residual size). The U/Pb and (U-Th)/He samples of late-mineralization Young Tonalite were taken from drill core at about 350m below the pre-mine topography surface. Thus, the Sampling Depth = 350 m. Theoretical modeling studies show that igneous bodies of 500 m width and about 2000 m height would be “cooled” within <1.0 m.y. if the Emplacement Depth was <10 km. So, the Batu Hijau tonalite porphyry body would have “cooled” before its temperature passed through the apatite (U-Th)/He closure temperature (90 °C).
We can limit the valid range of the emplacement depths for Batu Hijau before running the model. For example, if Emplacement Depth = 2.0 km and Eroded Thickness of porphyry = 0 m, the erosion rate before exposure is about 0.89 mm/yr using Equation (1).
At this rate, Batu Hijau would still not be exposed because the eroded thickness of country rock is 1.86 km (0.497 km/m.y. × 3.74 m.y.) which is less that the emplacement depth (2.0 km). Therefore, the emplacement depth must be deeper than 2.0 km.
If we assume the Eroded Thickness of the porphyry is 0~500 m, the minimum Emplacement Depth for Batu Hijau is 1.5~2.0 km;
Due to the absence of zircon (U-Th)/He age data, the maximum emplacement depth for Batu Hijau could be up to 5.0 km. However, if we consider reasonable values for the erosion rates of the porphyry and country rock, the maximum emplacement depth can be limited <5 km. For example, assuming Emplacement Depth = 3.5 km, if Eroded Thickness of porphyry = 100 m, the model will generate an erosion rate before exposure of about 1.56 mm/yr and an erosion rate after exposure of about 0.05 mm/yr. If we increase the Eroded Thickness of the porphyry to 1000 m, the erosion rates before and after exposure would be 2.15 mm/yr and 0.46 mm/yr. However, because the erosion rate before exposure is too fast and the difference between the two rates is too large, this is an unlikely scenario and the maximum emplacement depth is 3.5 km.
If the emplacement depth can be determined, we can also limit the Eroded Thickness of the porphyry to a reasonable range by considering the erosion rate of the overlying country rock and porphyry rock units.
Then, based on the ranges of Emplacement Depth and Eroded Thickness obtained, we can generate a cooling curve that matches all the real age data and produces reasonable erosion rates. An Emplacement Depth of 2.4 km and Eroded Thickness of about 300 m are geologically reasonable and produce a cooling curve that successfully passes through all age data points (Fig. 5e⇓). The calculated average erosion rate for the porphyry is 0.24 mm/yr, consistent with other porphyry units in the region—the calculated erosion rates for the Grasberg porphyry and for Ciemas porphyry are 0.23 mm/yr and 0.11 mm/yr, respectively.
5. Limitations and future improvements
Due to the limited information on sample position and the plethora of uncertainties, the above algorithms are not ideal. However, a feasible solution is yielded under the current conditions and improvements are constantly being made. For example, if two or more samples (with as large an age difference as possible) are obtained from same porphyry deposit for dating, we can generate a unique solution to the porphyry emplacement depth. A range of emplacement depths that satisfy the cooling curve constraints of the real age data for each sample can be identified. At a certain point, the “possible” emplacement depth ranges for both samples will intersect and we can solve for the true emplacement depth for the porphyry.
The management of NICICO is gratefully acknowledged for providing access to Sar Cheshmeh, Miduk and Abdar. Thanks to Jeff Davis and Ahmed Ali for mineral separations, Ratih Woodhouse and Marcus Gregson for hand-picking/quality control, Lesley Dotter for zircon dissolutions, Peter Pollard for provision of samples and images from the Grasberg deposit, and to Travis Naughton and Angelo Vartesi for drafting. We are grateful to Rio Tinto Mining and Exploration Ltd, in particular Ross Andrew, Neil McLaurin and John Bartram, for supporting this work. Field activities in Indonesia were supported by AusAID and the Indonesian Department of Energy and Mineral Resources. Jo-Ann Wortho and Barry Kohn provided helpful comments and suggested additional references. The thoughtful reviews of Anthony Harris and Ken Hickey greatly improved the chapter.