|
|
Subsolidus re-equilibration of minerals that originally formed as phenocrysts in magmas may produce compositional changes at high, but sub-magmatic, temperatures through diffusion, exsolution, pseudomorphic replacement, and other processes. Hydrothermal fluids from several different sources may produce overlapping or overprinted mineral assemblages at intermediate temperatures. And, at the lowest temperatures ground waters, particularly in areas of torrential rainfall such as Ok Tedi, gain access to near surface rocks where they cause oxidation of magnetite and sulfides to form limonites, cause hydration of feldspars to form clays, and cause the conversion of iron-magnesium silicate minerals to mixtures of clay and limonite. Any of these processes may reset or obliterate geologic thermometers and barometers. Additionally, equilibrium must be assumed in many or all geothermometers and barometers, but textural features such as oscillatory zones in plagioclase feldspar and exsolution textures of ilmenite in magnetite, as in the least-altered igneous rocks of the Ok Tedi Intrusive Complex, suggest that equilibrium may not have been attained in these minerals even under the conditions of initial magmatic crystallization.
Burnham (1979, 1981) has discussed seven major constraints that control mineral stabilities in magmatic and hydrothermal systems and they are: H2O content, temperature or heat content, oxidation state (fO2), metal content, chlorine content, sulfur content, and the ratio of NaCl to KCl. Anderson (1996) has reviewed the status of thermobarometry in granitic batholiths and discussed several methods that have been used to learn about the intensive parameters that controlled the formation and emplacement of granitic rocks.
Several of the methods described by Anderson (1996) are applicable for estimates of the conditions under which the igneous and metasomatic rocks of the Ok Tedi Intrusive complex formed. These include the two-feldspar (Ghiorso, 1984; Fuhrman and Lindsley, 1988; Wen and Nekvasil, 1994) and ilmenite-magnetite thermometers (Anderson and others, 1993), the aluminum-in-hornblende barometer (Hammarstrom and Zen, 1986), and the sphene (titanite)-magnetite-quartz assemblage (Wones, 1989) that is an estimator of oxygen fugacity (fO2). In addition to the methods discussed by Anderson, temperature is estimated from the isotherms for clinopyroxenes given in Kretz (1982); the relative chemical activities of Na, K, and H+ in the zone of potassic alteration are estimated using activity-activity diagrams as described by Meyer and Hemley (1967), Helgeson (1974), and Montoya and Hemley (1974), and Bowers and others (1984); sulfur fugacity is assessed from the presence of pyrite and magnetite in the massive magnetite, magnetite-sulfide, and sulfide metasomatic bodies in, and peripheral to, the zone of potassic alteration; and the salinity of hydrothermal fluids is estimated from the presence of cubes of halite in, or the freezing point depression of, fluid inclusions.
Some of the available methods for estimating intensive and extensive parameters are more robust than others; they retain clues to the physical parameters under which they originated even though later conditions may change greatly. The hornblende-plagioclase feldspar and hornblende-pyroxene thermometers are examples of robust estimators. The two-feldspar and ilmenite-magnetite thermometers are examples of less robust estimators (Anderson, 1996). In order to combat the problems involved with single estimators, Anderson stresses that petrologists should use as many different thermometers and barometers as possible to access the temperatures, pressures, and other environmental conditions involved in the formation of igneous rocks.
The temperature of formation for crystals of pyroxene in the least-altered intrusive rocks can be estimated from isotherms on the clinopyroxene slope of the solvus surface derived by Kretz (1982) as shown in Figure 111. Microprobe-determined compositions of diopside crystals in samples from the Ok Tedi Intrusive Complex are also plotted on Figure 111. Formations of the pyroxene crystals in the least-altered samples are assigned temperatures of 700C, or less from the plotted positions of the compositions relative to the isotherms of Kretz. This temperature is also within the range of temperatures obtained from calculations of coexisting potassium and plagioclase feldspar as shown in Figure 23. The positions and shapes of solvus curves for 400, 500, and 800C that are drawn on ternary diagrams in Figures 23 and 24 were calculated by the program SOLVCALC (Wen and Nekvasil, 1994). Calculations using microprobe-determined compositions of potassium and plagioclase feldspars from the least altered rock samples and equations given by Elkins and Grove (1990) yielded temperatures of about 800 to 600C. In contrast, the weakly or moderately altered rocks yielded temperatures from about 550 to 400C (the lower limit of the method). Temperatures below 400C require a different thermometer. The temperatures obtained for individual samples are summarized in Table 27.
Three microprobe analyses of exsolution lamellae of ilmenite and host magnetite were made on polished thin-sections DDH 342-84.5 (Sydney Intrusion), DDH 331-218.5 (Sydney), and DDH 458-151.8 (Kalgoorlie). The compositions of the ilmenite and magnetite were entered into the program QUILF (Anderson, Lindsley, and Davidson, 1993) for calculation of temperature and oxygen fugacity. The temperatures obtained range from 380 to 505C and are compatible with subsolidus exsolution of ilmenite from magnetite. Values for oxygen fugacity obtained from the QUILF calculations range from 10-17 to 10-20 and plot near the magnetite-hematite buffer as shown in Figure 112.
Crystallization pressures of igneous rocks are commonly estimated from metamorphic mineral assemblages in the intruded wall rocks but the best way to approach the problem is to estimate the equilibration P directly from coexisting mineral assemblages of the granitic rocks themselves (Ague, 1997). Hammarstrom and Zen (1986) presented evidence that pressure could be evaluated from the total aluminum (Altot) content of hornblende. The theoretical basis of the "Al-in-hornblende" barometer is the Gibbs phase rule P = C-F+2 where P is the maximum number of independent phases, C is the number of independent components, and F is the number of independent degrees of freedom. Any amphibole-bearing igneous rock that can be described by the ten component system (SiO2-TiO2-Al2O3-Fe2O3-FeO-MgO-CaO-Na2O-K2O-H2O) and that contains the nine phase assemblage of plagioclase feldspar + hornblende + quartz + potassium feldspar + biotite + sphene + a FeTi oxide + melt + a fluid phase is trivarient. The three degrees of freedom that are possible according to the phase rule can be expressed as oxygen fugacity, temperature, and pressure. If oxygen fugacity is buffered by a second iron-titanium oxide mineral and the temperature of formation is near the solidus then pressure is the only unconstrained variable. Hammarstrom and Zen (1986) proposed that the content of Al2O3 is directly related to pressure in systems that comply with the above requirements (ten component system, nine phase mineral assemblage) and they proposed an equation by which pressure could be estimated. Since the publication of Hammarstrom and Zen's paper several authors have proposed modifications of the original equation. Two samples from the Ningi Intrusion contain hornblende and all of the other minerals of the required assemblage and have had their total alumina content measured by microprobe analyses. These can therefore be used for pressure estimation. The values of total alumina in multiple analyses of hornblende crystals from the two samples are listed in Table 28 along with the equations given by Hammarstrom and Zen (1986), Hollister and others (1987), Johnson and Rutherford (1989), and Schmidt (1990) and estimated pressures calculated from each of the calibrations. Pressure estimates were calculated from the maximum, average, and minimum alumina content in each sample. The minimum pressure at which hornblende is stable as a crystallizing phase is 500 bars (Burnham, 1979) to one kilobar (Ague, 1997). If values less than 500 bars are eliminated from Table 28 than the calculated the samples of the Ningi Intrusion range from 0.6 to 2.4 kilobars. The calculated pressures of DDH 356-166.5 are consistently about 0.6 kilobars less than those obtained from DDH 351-313.2 regardless of the calibration equation used. The important point is that all values obtains yield relatively low pressures and, therefore, a relatively shallow depth of emplacement (3 to 5 kilometers).
The minimum water content in the magma(s) that crystallized to form the Ok Tedi Intrusive Complex is constrained by the presence of biotite in most of the unaltered rocks and hornblende in a few of the samples. Burnham (1979) states that the water content of magma must be 2-4 weight percent and pressure must be greater than 500 bars for hornblende or biotite to stabilize as phenocryst phases. Naney (1983) determined that about 4 weight percent of water at 2 kilobars is necessary for hornblende whereas biotite may be stable in melts containing as little as 0.5 weight percent water. The presence of biotite as an early formed phenocryst phase and of hornblende in only a few samples, where it undoubtably formed late in the crystallization of the magmas at Ok Tedi, documents an increase in water content with increasing crystallinity.
Additionally, magmas that contain less that 2 weight percent water are incapable of releasing sufficient mechanical energy through second boiling to produce the extensive fracture systems that are required to provide the conduits through which hydrothermal fluids can flow, cause alteration of silicate and oxide minerals, and deposit metal sulfides and gold (Burnham, 1981, Anthony, 1983).
The presence of fluid inclusions in quartz crystals in the least-altered phanerites of the intrusions at Ok Tedi also demonstrate the presence of an aqueous phase late in the crystallization process.
Additional constraints on the water content and solidification temperatures of igneous melts can be obtained from experimental studies. Holtz and Johannes (1994) have summarized the liquidus and solidus phase relations for the hypothetical system Qz-Or-Ab (eutectic melts or haplogranites). Temperatures for the melting and solidification of igneous rocks such as those at Ok Tedi are undoubtably higher than those obtained for the hypothetical system because of the presence of additional phases such as anorthite in plagioclase feldspar and the iron and magnesium contents of ferromagnesian minerals. Nonetheless consideration of experimental data from study of these eutectic compositions does provide lower boundaries on the solidification temperature of igneous melts and on the amount of water that can be contained in magmas. Burnham (1979) has shown that the solubility of water is about the same in melts of a variety of compositions (andesite, albite, and Li pegmatite) for the range of pressures, 1 to 2 kilobars, likely to have prevailed at the crustal level in which the Ok Tedi Intrusive Complex was emplaced as shown in Figure 113. A curve delimiting the water-saturated solidus of eutectic or minimum granitic compositions (redrawn from Holtz and Johannes, 1994), liquidus curves for given amounts of H2O (solid lines), and H2O solubility curves (dashed lines) are illustrated in Figure 114. It can be seen from the position of the water saturated solidus curve that, regardless of water content, granitic magmas can be expected to have solidified by 630C which suggests that the temperatures less than this value that were calculated using two-feldspar thermometry represent subsolidus re-equilibration. The maximum solubility of water at pressures of 1-2 kilobars is 4 to 6 weight percent. In order to have produced a fluid phase it is necessary that the magmas at Ok Tedi must have been oversaturated with respect to water. It is likely that the magmas did not become oversaturated until some critical percentage of phenocrysts had crystallized. The average groundmass content of the igneous rocks with porphyritic texture at Ok Tedi is about 38 volume percent. The groundmass can be equated to liquid content in a crystal mush. If part or all of fluids responsible for hydrothermal alteration at Ok Tedi originated from the porphyritic magmas, it is possible that water did not become supersaturated with respect to the melt phase until greater than 50 volume percent of these magmas solidified as a result of crystallization.
The stability fields of iron oxide and sulfide minerals are commonly depicted on binary plots of the activity, or fugacity, of oxygen versus sulfur (Barton and Skinner, 1979) and these diagrams are used to describe and interpret phase relations in geologic systems. The stability of single minerals and coexisting two and three mineral assemblages are dependant on temperature as well as oxygen and sulfur fugacity. Two log fO2-fS2 diagrams for 700 and 450C are given in Figure 112. These temperatures were chosen because they provide reasonable upper and lower bounds on the transition from magmatic to hydrothermal conditions. The boundaries between pyrite and magnetite on these diagrams set limits on oxygen and sulfur fugacity at a particular temperature. The upper and lower values of oxygen fugacity defined by the two mineral assemblage magnetite-pyrite are projected from the log fO2-fS2 diagrams for 700 and 450C into a binary diagram of temperature versus log fO2 in Figure 112. The values of oxygen fugacity at 700C (10-12 to 10-14) and at 450C (10-21 to 10-26) are above the quartz + magnetite fayalite (QFM) buffer and as such are indicative of strongly oxidizing conditions. The values of fO2 implied by the presence of the pyrite-magnetite assemblage are consistent with the presence of quartz and sphene in the magmatic assemblage and the high oxygen fugacity suggested by the compositions of coexisting magnetite and exsolved ilmenite. The sulfide minerals, pyrite and chalcopyrite, are ubiquitous phases throughout the zone of potassic alteration whereas massive, hydrothermal, magnetite is rare in the center of porphyry mineralization. Most bodies of massive magnetite (± pyrite and chalcopyrite) at Ok Tedi are concentrated near the outer fringe of the potassic zone although a few pods of massive ore were exposed near the center of the potassic zone. The distribution of magnetite and the sulfide minerals in porphyry and massive ore bodies probably records fluctuations in the ratio of sulfur to oxygen as the hydrothermal fluids and gases evolved. Sulfur fugacity may have been higher than the pyrite-magnetite boundary in zones of normal porphyry ore, lower than this boundary in areas of massive magnetite, and at the boundary in areas of massive magnetite-pyrite±chalcopyrite. The fluctuations were likely related to both structural position within the hydrothermal system and time. Crosscutting relations and other textural features in all of the massive replacement ore bodies that I observed at Ok Tedi indicate that where both are present, magnetite preceded pyrite (±chalcopyrite). Clark and Aranciba (unpublished) argue that the strongly oxidized nature of the hydrothermal fluids in systems that contain early magnetite-rich veins is definable as a preponderance of SO2 over reduced sulfur species, including H2S; that under such conditions magnetite and gold can readily participate; and that copper and other chloride-complexed base metals remain in solution until later in the hydrothermal process. If the model of Clark and Aranciba can be extended to the massive replacement ore bodies at Ok Tedi, it provides an explanation for the textural relations of pyrite±chalcopyrite and magnetite and may imply that the massive bodies of magnetite formed prior to the deposition of sulfide minerals in the porphyry ore bodies. It also may indicate that the change from magnetite to pyrite and chalcopyrite deposition was due to cooling through the boundary between SO2 and H2S as shown at a temperature of about 500C for the cooling path shown in Figure 112. This tentative interpretation is consistent with thermodynamic data (Robie and others, 1978 and 1979) that demonstrate that H2S becomes increasingly stable relative to SO2 with diminishing temperature and at constant fO2.
Relevant to this discussion are related sulfur isotope studies that are currently being conducted on sulfide concentrates from the Ok Tedi deposit (Field and others, 1999, unpublished data). The 34S per mil values of 49 sulfide concentrates, excluding those of two anomalously depleted chalcocites of supergene origin, range from -3.4 to +2.2 per mil. This range is entirely compatible with those published for other major porphyry Cu-Mo deposits such as Bingham, Utah (Field, 1966), Butte, Montana (Zhang and others, 1999), and El Salvador, Chile (Field and Gustafson, 1976). Moreover, this narrow range about the zero per mil value of the meteorite standard is consistent with that of a deep-seated "magmatic" source of sulfur (Ohmoto and Rye, 1979). Most of the different sulfide minerals from Ok Tedi exhibit reasonable approximations to isotopic equilibrium, and using the fractionation equations of Ohmoto and Rye (1979) provide isotopic estimates of depositional temperatures. These include temperatures of 470C for one molybdenite-chalcopyrite mineral pair, 380 to 420C for two molybdenite-chalcocite pairs, 365 to 595 C for four pyrite-chalcopyrite pairs, and 360 to 545C for two chalcopyrite- bornite pairs. Thus, the sulfur isotope data suggest a general temperature range of about 450 to 550C for hydrothermal mineralization at Ok Tedi. The presence of fluid inclusions containing cubes of halite demonstrates that the fluids that evolved in the hydrothermal stage of magmatism were characterized by very high salinities, at least 35 weight percent NaCl equivalent (Roedder, 1971; Nash, 1976), through at least part of their evolution. Halite-bearing inclusions were observed in interstitial crystals of quartz, but not earlier formed minerals, in intrusive rocks with phaneritic texture suggesting that the saturation of the Ok Tedi magmas with respect to water and the separation of a hydrous phase did not occur until late in their sequence of crystallization. Homogenization temperatures from halite-bearing inclusions were reported as 650 to 450C (Wilkins, unpublished). Nedachi (1992) reported the presence of multiphase fluid inclusions in phlogopite-quartz-apatite veinlets of the Fubilan Intrusion that have filling temperatures of 380 to 470C. Separation of a fluid phase from the magmas at Ok Tedi is likely, based on the fluid inclusion data, to have occurred at temperatures of 650C or below.
The stabilities of most of the silicate gangue minerals of the potassic, phyllic, and argillic zones of porphyry deposits can be represented in the system Na2O-K2O-Al2O3-SiO2-HCl-H2O. Boundaries between the common hydrothermal phases are delineated on mineral stability diagrams in terms of the logarithm of the activities (or ratios of activities) of the dissolved species that control mineral-solution equilibria. The dissolved substances that play important roles in the hydrothermal fluids that produce porphyry deposits include H+, K+, and Na+. In addition, Ca2+, Mg2+, and iron (Fe2+ and Fe3+) are also important particularly in zones of propylitic alteration. However, calcium is not a major constituent in the samples of altered rock collected and analyzed as part of this study as has been shown earlier in Harker (Figure 60), Peacock (Figure 75B) and isocon (Figure 106-108) diagrams for potassically altered rocks. Accordingly, the stability of minerals containing CaO and their relations to other hydrothermal phases will not be included here. Mineral stability diagrams have been constructed from experimental and thermodynamic data by several authors including Hemley and others, (1961), Garrels and Christ, 1965, and Bowers and others (1984). Four mineral stability diagrams covering the likely range of hydrothermal temperatures (500-300C) experienced by the intrusive rocks of the Ok Tedi Intrusive Complex are redrawn from Bowers and others (1984) in Figure 115. The diagrams are for 1 kilobar pressure. The dominant aluminosilicate minerals in the potassic zone at Ok Tedi are potassium feldspar and subordinate albite. These two minerals can coexist throughout the range of temperatures shown in Figure 115, but only if the activities of K+ and Na+ are very much greater than the activity of H+. The stippled areas in Figure 115 show the likely compositions of the hydrothermal fluids that formed the potassic zone at Ok Tedi. In areas where potassic alteration was particularly strong, fluid compositions must have moved from the boundary between albite and potassium feldspar into the field of potassium feldspar because albite is not a major mineral phase in samples with greater than about 10 weight percent K2O.
The temperature estimates made in this chapter are combined with inferences drawn from petrographic observations to construct the paragenesis diagram of Figure 116. It must be reemphasized that the temperatures assigned are rough approximations. Only a few of the mineral phases have had temperature estimated by geologic thermometers and these values are likely to represent minimum or re-equilibration temperatures. Calculated temperature intervals are represented by solid line segments; dashed line segments represent mineral relations from microscopic examinations; and dotted lines are inferred from intuition. Crystallization of the earliest formed magmatic minerals, apatite and plagioclase feldspar, are inferred to have begun just below a liquidus temperature calculated from the composition of DDH 340-166.5 by the computer program PELE (Boudreau, 1999). Crystallization of quartz, the last magmatic mineral formed in most samples with phaneritic texture at Ok Tedi, is assigned a temperature of about 650C estimated from the water-saturated solidus at about 1.5 kilobars pressure (Figure 114). The crystallization of hornblende is assigned so as to straddle the location of the solidus temperature (650C) because the full buffering assemblage necessary for the successful application of the "Al-in-hornblende" barometer, that is present in some phanerites at Ok Tedi, is reported to only become stable near the solidus (Ague and Brandon, 1996). The high temperature ends of the lines representing the formation of hydrothermal feldspars are from the two-feldspar thermometer; the low temperature ends are unconstrained. The transition in stability from hydrothermal magnetite to pyrite is taken as the approximate point where the hypothetical cooling path of Figure 112 crosses the line separating SO2 from H2S. Water is assumed to have been present as a fluid phase throughout the process of formation of quartz, hornblende, albite, potassium feldspar, hydrothermal biotite, magnetite, and pyrite in the zone of potassic alteration. The lack of superimposed zones of quartz-sericite or advanced argillic alteration suggests that lower temperature magmatic or meteoric waters may not have had contact with the rocks of the potassic zone again until they were uplifted into the zone of weathering and supergene alteration. The formation of clays and limonites are inferred to have occurred in the zone of weathering and thus are assigned to the lowest temperature range on Figure 116.
