|
|
Quartz accounts for an average of 4.5 volume percent in rocks with phaneritic texture. Most quartz is anhedral in shape and is found in the interstices between crystals of feldspar (Figures 38A and B) or crystals of feldspar and pyroxene (Figures 38C-F). The anhedral shape of the quartz and its presence in interstitial sites suggests that it crystallized during a late stage of the solidification of the magmas that formed the phanerites.
Quartz is present as an anhedral phase that is sparsely distributed throughout the groundmass of unaltered porphyries at Ok Tedi. It was noted as a possible phenocryst phase in only a few of the thin-sections of porphyry samples examined as part of this study. However most of these samples consisted of highly altered rock in which quartz may have been part of a veinlet, vug filling, or miarolitic cavity rather than a discrete magmatic crystal. Euhedral crystals of quartz such as the "quartz-eyes" that are found in many other porphyry copper systems are entirely lacking at Ok Tedi.
Many crystals of quartz contain fluid inclusions. Most of the inclusions include multiple phases consisting of a liquid, a gas bubble, and one or more crystal phases. The largest and most common solid phase is invariably a cube of halite (Figures 38G and H). The presence of saline fluid inclusions in unaltered phanerites is significant in that it suggests that a fluid phase was present in the late stages of magma consolidation. Such a saline fluid may have been responsible for the hydrothermal alteration that transformed the mineralogical composition of the intrusive rocks and that deposited the hydrothermal copper and gold of the Ok Tedi deposit. A study by R.W.T. Wilkins (unpublished) of CSIRO Laboratories of Australia has shown that both high-density multiphase liquid inclusions and low-density gas-rich inclusions are common at Ok Tedi; that homogenization temperatures of the inclusions range from 400 to 650 C; and that the multiphase inclusions contain up to 64 weight percent NaCl + KCl.
Veinlets of quartz are present in both phanerites and porphyries throughout the Ok Tedi Intrusive Complex, but are most common and abundant in and near the quartz core. Within veinlets, the quartz forms anhedral granular mosaics. Fluid inclusions are also common in the vein quartz. Chalcopyrite, pyrite, and (possibly) bornite are common associates of quartz in the veinlets. Limonite as an oxidation product of earlier formed sulfides is also a common associate of quartz in the zone of oxidation.
Quartz also occurs as intergrowths with hydrothermal biotite in sites previously occupied
by magmatic pyroxene (possibly also hornblende or biotite.) Fluid inclusions are common in the
quartz in these many of these pseudomorphic intergrowths. The quartz in these sites also
commonly contains fluid inclusions. Many of these inclusions are elongate and tube-like in
shape.
Groundmass was a counted phase in a few of the phanerites listed in Appendices 2 and 3.
Where used in such descriptions of phanerites the term denotes areas of very finely crystalline
texture.
Most crystals of sphene in the intrusive rocks from Ok Tedi are euhedral in shape. They display the wedge, diamond (Figure 39A), lozenge (Figure 39B-D), and dovetail shapes that are characteristic of this mineral. Sphene is closely associated with the mafic minerals (pyroxene, hornblende, and magmatic biotite), apatite, and magnetite. This association is typified by inclusions of sphene in the mafic minerals, ragged growths of sphene on magnetite, and the close physical proximity of sphene to the mafic minerals in samples believed to represent magmatic rock. Inclusions of apatite and magnetite are present in some samples (Figures 39C and D). The common association of sphene with the mafic minerals, apatite, and magnetite as well as as inclusions of these minerals in one another suggests that they formed contemporaneously.
A map showing the distribution of sphene in thin-sections from the Ok Tedi Intrusive Complex is given as Figure 40. The average modal content of sphene in samples that contain either pyroxene or hornblende (and therefore are representative of magmatic assemblages) is 1.04 percent whereas the average in samples which do not contain these minerals (i.e., those with potassic alteration) is 0.13. This suggests that sphene was unstable, and therefore destroyed, in the hydrothermal environment.
Fifteen microprobe analyses were made on crystals of magmatic sphene in three thin sections. Representative analyses are listed in Table 7. The entire sets of analyses are given in Appendix 8. The weight percentages of CaO, TiO2, and SiO2 of stoichometric sphene, as calculated from the ideal formula, should be 28.6, 40.8, and 30.65, respectively. The average of these oxides in the fifteen analyses performed are 27.3, 35.9, and 29.9 weight percent, respectively. The measured values of CaO and SiO2 are reasonably close to their stoichiometric values but the amount of TiO2 is low. The analyses contain, besides the expected oxides, an average of 1.2 weight percent Al2O3, 1.9 percent FeO, and 0.53 weight percent fluorine. It is likely that some substitution of Fe for Ca(?) and possibly Al, for Ti has taken place, and that the deficiency of TiO2 can be related to such possible substitutions.
Apatite is present in altered and unaltered phanerites and porphyries from Ok Tedi. It has two distinctly different sizes and habits: 1.) as tiny rod-shaped crystals that are typically less than 0.01 mm in length, and 2.) as stout crystals with circular cross-sections, which are as large as one millimeter in diameter (Figures 41A-C). Apatite was noted in about 20 percent of the thin-sections with porphyry texture and in about 65 percent of those with phaneritic texture. It accounts for about 0.3 volume percent in both the phanerites and porphyries in which it was observed.
Zircon is present in about one-half of all thin-sections that were examined. It is present in phanerites and porphyries. Its abundance is about the same in unaltered intrusions and their counterparts that have been subjected to strong potassic alteration. Zircon accounts for less than 0.1 volume percent in the rocks in which it is found. Most crystals are rounded in shape (Figure 41D) and less than 0.1 mm in diameter. Zircon is found in association with, or is overgrown by apatite, magnetite, pyroxene, and magmatic biotite, which suggests an early formation in the paragenetic sequence.
The principal opaque oxide mineral of the least altered intrusive rocks is magnetite. It occurs as cubes and rounded cubes (Figures 41E-G). Exsolution lamellae of ilmenite were observed in the magnetite crystals of a few thin sections. Two size populations are present in many samples: 1.) magnetite granules (larger than 0.1 mm), and 2.) magnetite dust (less than 0.01 mm). Although granules are less abundant in terms of total number of crystals, they are probably much greater in aggregate volume as compared to the dust variety. Magnetite granules form rings around ragged crystals of magmatic biotite in some samples (e.g., Figures 33E and F). These may have originated as oxidation or resorption products of the biotite and, therefore, the result of a magmatic process. Magnetite is present as inclusions in pyroxene, hornblende, biotite, and sphene. Much of the magnetite in samples from Ok Tedi is enveloped by thin fringes of ragged anhedral biotite which suggests that it may have served as a nucleation site for the biotite.
Microprobe analyses were made of magnetite in six thin-sections. Representative
analyses of magnetite and ilmenite are listed in Table 8. The entire set of analyses which
includes magnetite, ilmenite, and mixtures of magnetite and ilmenite is given in Appendix 9.
The results of the analyses are portrayed on ternary diagrams in Figures 42A and B. Many of the
analyses plot near Fe2O3 = 69 weight percent, FeO = 31 weight percent on these diagrams
(Figure 42A). Many other analyses plot between the composition of magnetite and ilmenite (or
ulvospinel). This indicates the presence of titanium at the sites that were hit by the electron
beam and may indicate the presence, in these locations, of thin exsolution lamellae of ilmenite
or concentrations of unexsolved titanium. Still other analyses plot near the composition of
ilmenite. These represent the composition of ilmenite lamellae.
Garnet and epidote are both common in samples from the Kalgoorlie, Ningi, and Sydney intrusion. Garnetite and epidotite that formed by wholesale replacement of intrusive rocks can be found in a few locations at Ok Tedi. These alteration minerals are likely the result of extreme propylitic alteration involving additions of calcium, in particular, and possibly of iron. These calc-silicate minerals and assemblages thereof can also be classified as skarns, particularly where the silicate minerals are in association with sulfide and/or oxide minerals. Katchan (1982) and Duncan (1972) suggest that most of the skarn at Ok Tedi is endoskarn and thus formed from intrusive host rocks. I excluded samples of massive garnetite, epidosite, and endoskarn from my detailed petrographic studies.
Epidote is found as partial pseudomorphic replacements of plagioclase feldspar, as an infilling mineral in veinlets (Figure 43A and B), and as disseminated anhedral to subhedral crystals that may have originated by the replacement of earlier-formed minerals or by the filling of void spaces (Figure 43C and D). Katchan (1982) reported that the Sydney Monzodiorite is the host for veinlets of epidote and that these increase in abundance toward bodies of skarn. He listed microprobe analyses of epidote and allanite and noted that epidote rims were found on allanite cores. I performed eight microprobe analyses on epidote in one thin-section as part of my study. The results of these analyses are given in Table 9. Not enough analyses were done to allow any discussion of systematic variation of the composition of epidote, but they conform the petrographic identification of the mineral. Lanthanum and cerium (characteristic of allanite) were not included in the analytical files and may account for the low total oxide percentages. Structural formulae were calculated on the basis of 13 anions (oxygen and fluorine) and with all iron as Fe2O3.
Garnet is similar to epidote in spatial distribution and mode of occurrence. It forms veinlets and disseminated crystals (Figures 43E and F) in rocks with propylitic alteration and culminates in abundance as massive garnetite in garnet skarn. It is found in association with the magmatic ferromagnesian minerals in some thin-sections (Figure 43E) and by itself in others (Figure 43F). Fifty-one microprobe analyses were performed on disseminated anhedral garnet in two thin-sections (DDH 341-141 and DDH 459-102.9). Representative analyses are given in Table 10; the complete sets are included as Appendix 10. Representative analyses from a set of 155 analyses performed by Lihua Zhang (using the microprobe laboratory at OSU) on one sample of garnet skarn from the Edinburgh area at Ok Tedi are also included in Table 10. All of these analyses are portrayed on Gr-Ad-Sp+Al ternary diagrams in Figure 44. The garnets are all grandite in composition consisting principally of the grossularite and andradite end members. The analyses from the garnet in DDH 342-141 cluster near the andradite corner of the diagram and range in Ad content from about 71 to 95. Most of the analyses of garnet in DDH 459 fall in the range of Ad60-79 but one analysis gave Ad35. The garnet in the sample of garnetite from Edinburgh shows a wide range of Ad content (about 27 to 81) and includes analyses similar to those of both other thin-sections.
The principal oxide mineral in samples that have undergone strong potassic alteration is rutile. Pseudomorphs of rutile after sphene are common in thin-sections from rocks which have undergone strong potassic alteration. Examples are shown in Figures 39E-H. Rutile is also present in the patches of hydrothermal biotite formed from mafic minerals.
Four microprobe analyses were performed on rutile in one thin-section (JDD-94-07). The analyses are given as Table 11. Three of these analyses fall on or near the Ti apex in Figure 45. The remaining analysis falls on the tie-line between Ti and Fe2+. These microprobe analyses confirm the identification of rutile but are insufficient in number to establish any patterns of chemical composition.
Veinlets of hydrothermal magnetite cut across intrusive rocks at several locations within
the Ok Tedi intrusive complex. Anhedral clusters of magnetite crystals were observed in one
specimen from an area of strong potassium feldspar metasomatism (JDD-94-07; Figure 46). The
anhedral shapes of the crystal clusters of hydrothermal magnetite are considerably different from
the characteristic cubic habit of the magmatic variety. Microprobe analyses of the hydrothermal
magnetite are given in Table 12 and are plotted on a ternary diagram (Ti-Fe2+-Fe3+/2) in Figure
47. This magnetite is indistinguishable in composition from that of magnetic origin.
The principal copper sulfides are chalcopyrite and bornite. They are most abundant and constitute mineable ore only in rocks affected by potassic alteration. Chalcopyrite and pyrite are also present in veinlets cutting across propylitically altered rock. Photomicrographs illustrating the typical appearance and crystal form of pyrite, chalcopyrite, and chalcocite are illustrated in Figure 48. Many of the crystals of sulfide minerals in these photomicrographs (and in porphyry ore in general) are strikingly similar in appearance, and in geometric relations to the silicate minerals, to magnetite in unaltered intrusive rock. The textural occurrence, distribution, and relationship to associated rock-forming silicates of the disseminated sulfides collectively suggests that most disseminated ore sulfides formed by hydrothermal replacement of magmatic magnetite. Pyrite could have been produced merely by the exchange of sulfur for oxygen in the magnetite crystal structure, whereas he formation of chalcopyrite would require the addition of copper as well as sulfur. Microprobe analyses were performed on pyrite, chalcopyrite, and chalcocite in polished thin-sections. Representative analyses are presented in Table 13; the complete set is given in Appendix 11. Most of these analyses are of nearly pure monomineralic sulfide material. The analyses are portrayed graphically in Figure 49 where symbols representing the analytical compositions of pyrite and chalcopyrite (pentagons) fall at, or immediately adjacent to, their expected compositions (open circles). Chalcocite analyses plot between chalcocite and covellite. Two analyses plot near the expected composition of bornite.
Zinc, silver, platinum, gold, and lead were included in the microprobe analytical files for
all sulfide mineral analyses. These elements were not, however, encountered in significant
quantities in any of the analyses. Most, in fact, have low values that fall within the expected
counting error for the analyses. Although microscopic evidence suggests that some gold is
present as inclusions in the sulfide minerals, these results do not contradict the observed
microscopic evidence. They do, however, suggest that these elements are not evenly distributed
throughout matrices of the sulfide minerals.
The second type of clay is white or gray in color, very fine-grained, and found in the sites of corroded crystals of feldspar. This type of clay is present in rocks that have undergone strong propylitic and potassic alteration. It is most abundant in areas of supergene alteration that are largely coincident with those of intense potassic alteration. White clay is present in local fault structures including one low angle fault that is located between the Vancouver and Edinburgh areas and that is over two meters in thickness. This type of clay was provisionally identified as kaolinite in petrographic studies. It is designated as Qz/Wm/Clay in Appendices 2 and 3 because of the possibility that, in many instances, it was an admixture with very fine-grained drusy quartz or white mica.
Four samples containing these clays were submitted for x-ray diffraction (XRD) analyses.
This analytical work confirmed the identity of smectite in samples with yellow waxy clays.
Prior analytical work in the exploration and feasibility stages of development (Goode and
Gilbert, 1976; Katchan, 1982; Ayres and Bamford, 1976, revised 1987) also found yellowish
clays to be smectites (montmorillonites). The white clays in the samples analyzed were found to
be mixtures of illite and halloysite. The identification of smectite in thin-sections is now well-established and can be easily applied in field identification. The identification of the white
clays, however, remains more tentative and further analytical work using samples from related
locations and alteration types is necessary for positive identification and mineralogical
interpretations.
Limonite has also formed by the in situ oxidation (rusting) of disseminated crystals of
pyrite and chalcopyrite within areas of porphyry-style mineralization in intrusive rock and in the
siliciclastic Ieru Formation. Pseudomorphs of limonite after pyrite were common in the leached
capping and in the upper part of the zone of supergene enrichment. Limonite is commonly
associated with malachite and azurite in porphyry, phanerite, and siltstone hosts. These
oxidation minerals are also accompanied by cuprite and native copper at some locations.
The samples portrayed in the accompanying figures are subdivided into two categories: 1) unaltered (or weakly altered, Figure 50), and 2) strongly altered (Figure 51). This distinction is based on the presence or absence of pyroxene (±hornblende). Samples with one or both of these minerals are considered to be unaltered. The figures are further subdivided according to which intrusion the samples are assigned. There are no samples of unaltered rock from the Fubilan Intrusion. Only one sample of altered (texturally-destroyed) intrusive rock is included in the samples from the Sydney Intrusion. This sample was not subjected to petrographic analysis, and it is not portrayed in Figure 51 because it has been changed by alteration to a very fine-grained texture, in which the distinction of quartz and the feldspars is difficult or impossible using a conventional petrographic microscope.
The recommendation of the I.U.G.S. Subcommittee on Igneous Rocks is to name rocks with phaneritic texture using plutonic rock names, and to name those with aphanitic texture using volcanic rock names. The Subcommittee does not make any recommendations for rocks with porphyritic texture. The convention used in this dissertation is to assign plutonic terminology to porphyritic rocks with phaneritic groundmass, and to assign volcanic terminology to those having aphanitic groundmass. This convention is in agreement with older classifications including that of Shand (1943). The QAP data from rocks with porphyry and phaneritic texture are plotted together on the same ternary diagrams in Figures 50 and 51. The textural varieties are distinguished by symbols: porphyries by unfilled circles, phanerites by filled circles. Samples of hornblende porphyry are depicted as unfilled diamonds.
Three samples of phaneritic rock from the Sydney Intrusion fall in field 9* and are, therefore, quartz monzodiorites (Figure 50). The thin-sections from these samples are among the least-altered encountered in examinations of both suites of petrographic samples. Two samples fall into field 10*. One has phaneritic texture and is classified as quartz diorite; the other has porphyritic texture and is classified as andesite. One sample (DDH 342-141) which contains a minor amount of garnet and is therefore considered to have undergone weak propylitic or endoskarn alteration, is classified as quartz monzonite.
All of the samples from Mount Fubilan are strongly altered and virtually all have porphyritic texture. Most plot in fields of 9 and 9* (andesite) as given in Figure 51. However, two samples fall in field 8 (latite) and one plots near the A-Q join in field 6* (quartz alkali-feldspar trachyte). A thin-section of the latter quartz alkali-feldspar trachyte sample (DDH 302-69) shows the most intense potassium feldspar alteration and the highest K2O content of any analyzed sample. The samples classified as latite and quartz alkali-feldspar trachyte most likely crystallized as andesites and were subsequently modified by potassium metasomatism. The names applied to these rocks, and to the others portrayed in Figure 51 imply metasomatic, and not igneous compositions. The example of hornblende porphyry (sample JDD-94-12) was collected from a dike within the Fubilan Intrusion; it plots in field 9 (andesite).
Four of the unaltered rocks from the Kalgoorlie Intrusion (Figure 50) plot in field 10 (andesite or diorite), whereas the fifth (DDH 389-271), a phanerite, plots in field 9 (monzodiorite). Five samples with strong potassic alteration are shown in Figure 51. Two of these are hornblende porphyry (samples JDD-5 and 11). They plot in field 9* and are, therefore, classified as andesites. One sample of feldspar porphyry plots near the boundary between fields 8 and 9. Two samples of phanerite are shown in Figure 51. One of these plots near the boundary between fields 9 and 10 and is a monzodiorite based on modal criteria. The other falls in field 8* (monzonite).
Samples of unaltered phanerite from the Ningi Intrusion plot in fields 9 (quartz monzodiorite) and 9* ( monzodiorite). One of the unaltered porphyries occupies field 9 (andesite), the other plots near the boundary between fields 9 and 10 (andesite). Samples of altered rock from this intrusion spread across the QAP diagram from field 10 to near the boundary between fields 7* and 3.
In summary, most of the unaltered samples of phanerites (7 of 10) fall in fields 9 and 9* whereas those with porphyry texture (4 of 5) fall closer to the P apex within fields 10 or 10*. This reflects the fact that plagioclase feldspar was an early phase in the crystallization sequence and that most of the rocks with porphyry texture were quenched before potassium feldspar was on the solidus, or shortly after it began to form. Presumably, this implies that the groundmass of unaltered samples with porphyry texture has high ratios of potassium to plagioclase feldspar, while most of the strongly altered samples that have porphyry texture fall in field 9 and 9*. Subsequently it will be demonstrated that these samples are metasomatic (quartz) trachytes and (quartz) alkali trachytes when classified using chemical criteria. They should, therefore, fall in fields 7, 7*, 6, and 6*. This discrepancy arises because most of the feldspar in these samples that was counted as plagioclase has An content of less than 5 and is, therefore, technically alkali feldspar and should be part of the A component. In addition, the potassium feldspar which dominates the groundmass in stained samples, and by implication all of the groundmass in potassically-altered rock, was not counted or added to the A component. The displacement of the strongly altered samples of porphyry from near the P apex to fields 9 and 9* results from partial replacement of many of the crystals of plagioclase feldspar by pseudomorphs or rinds of potassium feldspar. Modal analyses of those samples in which potassium feldspar replacement of plagioclase has been exceptionally well developed plot in fields 6*, 8, and 8*.
The mineralogical effects of metasomatic alteration on the feldpars in samples of strongly altered phanerites are more complex. Only one of these samples (JDD-94-14) show much displacement toward the A apex. The remaining samples from altered phanerites may have undergone albitization of the plagioclase feldspar with but minor pseudomorphic replacement of plagioclase by potassium feldspar. However, many samples with phaneritic texture that were studied as part of the first suite of petrographic analyses display strong alkali-feldspar replacement textures demonstrating that such alteration is common in rocks with phaneritic texture.
