|
|
Plagioclase and potassium feldspar are the most abundant minerals in all of the intrusive rocks at Ok Tedi. This is as true for hydrothermally altered intrusions as it is for unaltered rock. The feldspars comprise up to 80 volume percent, or more, of the phanerites and probably are similar in abundance in the porphyries. Pyroxene is the next most abundant mineral, comprising as much as nine volume percent, in most samples of unaltered rock but it is not present in rocks affected by potassic alteration. Hornblende is present in rocks from the Kalgoorlie, Ningi, and Sydney intrusions. Where present it never exceeds five volume percent in feldspar porphyry or rocks with phaneritic texture. It is a major constituent of hornblende porphyry dikes that cut the Fubilan and Kalgoorlie intrusions, but these are volumetrically insignificant. Quartz is a minor constituent in most thin-sections of unaltered rock. Its forms anhedral interstitial crystals in phanerites and is sparse as a phenocryst phase in porphyries. Veinlets with quartz as the major infilling gangue mineral cut across many samples and exposures of phanerite and porphyry. These veinlets are most abundant in and surrounding the quartz core. Magmatic biotite is common in many unaltered rocks with phaneritic and porphyritic texture. Relict magmatic biotite and finely crystalline aggregates of hydrothermal biotite are present in rocks that have undergone potassic alteration. Magnetite and sphene are common accessory minerals in unaltered intrusive rock as well as in propylitically altered rock, whereas various associations of chalcopyrite, pyrite, chalcocite, marcasite, pyrrhotite, bornite, molybdenite, and rutile are typically present in altered equivalents. Hydrothermal magnetite and sphene were noted in a few samples that have undergone potassic alteration, but, in general, these minerals do not seem to be a part of the process of potassic alteration. They may have been deposited in an earlier or later stage of hydrothermal alteration than the alkali feldspars and biotite. Apatite and zircon are present in both altered and unaltered host rocks.
Mineral assemblages indicative of propylitic, potassic, and argillic alteration are
common at Ok Tedi and in both suites of thin-sections examined as part of this study. Propylitic
alteration is characterized by the presence of epidote, calcite, actinolite, and chlorite. One or
more of these minerals are present in many samples from the Kalgoorlie, Ningi, and Sydney
Intrusions but are rare, or entirely lacking, in the Fubilan Intrusion. Potassic alteration is typified
by potassium feldspar and/or hydrothermal biotite. Virtually all rocks of the Fubilan Intrusion
have been subjected to potassium feldspar alteration and most also contain hydrothermal biotite
either as pseudomorphs after ferromagnesian minerals, disseminations, or in veinlets. In
addition, potassic alteration is present locally within the Kalgoorlie and Ningi intrusions.
Argillic alteration is also found at many locations. It is typically manifest as smectite in the
propylitically altered rock and as halloysite in areas of potassic alteration. Samples with strong
argillic alteration were not used in detailed petrographic examinations.
The thin sections chosen for detailed microscopy and microprobe analyses included samples with diverse textures and mineral assemblages. In terms of feldspar assemblages, however, most samples fall into one of two major types: (1) igneous rocks that are unaltered or weakly altered, and (2) igneous rocks that have undergone moderate to strong potassium metasomatism. Rocks that are characterized by strong propylitic or endoskarn alteration, although certainly present at Ok Tedi, are relatively rare and were not selected for detailed study.
Six hundred seventy-three microprobe analyses were done on the feldspars in 21 polished thin-sections. Representative analyses are given in Table 3. The entire sets of analyses are given in Appendix 4. The chemical compositions obtained for the plagioclase feldspars in these rocks clearly indicate the mineralogical differences between rocks with andesine and those with albite. The least altered rocks contain plagioclase feldspars with microprobe determined compositions that range from anorthite to albite but with most analyses falling in the range from calcic-andesine to calcic-oligoclase (Figure 23). The plagioclase feldspars in rocks that have undergone potassic alteration range from sodic-oligoclase to near end-member albite in composition (Figure 24). In contrast to the strong chemical differences in plagioclase feldspars in rocks of magmatic and metasomatic origin, the potassium feldspar in magmatic assemblages is not significantly different in composition from that in rocks that have undergone potassium metasomatism. Potassium feldspar ranges in composition from Or90 to Or50 when in association with either andesine or albite (Figures 23 and 24).
Temperature curves based on calculations using the SOLVCALC program of Wen and Nekvasil (1994) are shown on these Figures 23 and 24. The curves shown on each diagram were calculated using analytical data (Ab-Or-An contents) of coexisting plagioclase and potassium feldspars. A pressure of one kilobar was assumed but the positions of the curves were not found to be strongly dependant on pressure. The curves chosen probably represent reasonable ranges of temperatures for primary magmatic crystallization (andesine-potassium feldspar) and hydrothermal alteration (albite-potassium feldspar). The compositions of feldspars of andesine-potassium feldspar association fall between the curves for 500 and 800 C and represent the temperatures of magma solidification. The feldspars in the plot of albite-potassium feldspar association were formed at temperatures below 400 C. Curves for temperatures below 350 C were not calculated because the method used is not considered to be accurate at these temperatures.
Feldspar displays a spectacular diversity in texture and internal structure in the unaltered igneous rocks of the Ok Tedi Intrusive Complex and their altered equivalents. Specific features that merit elaboration include: twins, growth zones, synnuesis, resorption, crystal shapes, replacement textures, and inclusions. The crystals of feldspar in phanerites and the phenocrysts in porphyries range in longest dimension from about one to three mm on average and attain lengths of approximately 7 mm in a few of the examined thin-sections
Albite twinning is characteristic of plagioclase feldspars and is present in all but the most potassically altered rocks. Most crystals with albite twins display many narrow and sharply defined lamellae when viewed in cross-polarized light (Figures 25A and B); others show only a few lamellae that are diffuse in detail (Figure 25C) and yet others are defined by two sets of lamellae that are oriented nearly perpendicular to one another (Figure 25D). All feldspars that have lamellae are called "polysynthetically twinned feldspar" in point counter analyses and are listed as "PST feldspar in Appendices 2 and 3."
Plagioclase feldspars with oscillatory zones are equal, or greater, in abundance than those with albite twins in many thin sections. Many zoned crystals display concentric patterns of zones (Figures 25E-H) that go to extinction sequentially from rim to core as the microscope stage is rotated between crossed polarizing filters. Others with oscillatory zones have more complex extinction patterns. All crystals of plagioclase with oscillatory zoning are described in petrographic examinations as "CZ (concentrically zoned) feldspar" and are listed as such in Appendix 3 (this category was not counted separately in Suite 1 analyses).
Crystals of plagioclase feldspar with twins are not significantly different in composition from those with oscillatory zones. The average anorthite content of both twinned and zoned crystals from unaltered or weakly propylitically altered phanerites and porphyries is about An40 whereas that of the crystals in porphyries that have undergone strong potassic alteration is about An2.5 (not enough analyses were available from potassically altered phanerites for comparison).
Many crystals of plagioclase feldspar exhibit complex growth patterns. Glomerocrysts consisting of synnuesis twins (Vogt, 1921; Vance, 1969) and other forms of crystal coalescence are common. An example, shown in Figures 26A and B, consists of two intergrown phenocrysts of plagioclase feldspar. One phenocryst has two internal resorption surfaces indicating that it had at least three periods of growth, and that it was resorbed at least twice (A and B in Figure 26B).
Potassium feldspar is a common mineral component in thin-sections of porphyritic and phaneritic texture. It is generally subhedral to euhedral in shape in the least altered rocks and in those that have undergone propylitic alteration. A typical example of euhedral potassium feldspar from a sample of phanerite is shown in Figure 26C. Pseudomorphs of potassium feldspar after plagioclase feldspar are characteristic of the rocks that have undergone potassic alteration and these commonly reproduce the subhedral or euhedral shapes of their magmatic plagioclase precursors. In contrast, a few samples of potassically altered phanerites from the Kalgoorlie and Ningi intrusions were found to consist mainly of anhedral crystals of potassium feldspar (Figures 21C and D, and 21G-H).
Poikilitic textures consisting of chadacrysts of plagioclase feldspar and pyroxene enclosed by oikocrysts of potassium feldspar are common in phaneritic rocks. A good example is shown as Figure 26D. Large (> 5mm) crystals of euhedral potassium feldspar with inclusions of plagioclase feldspar that are aligned along internal growth zones are illustrated in Figures 26E and F. This texture suggests that potassium feldspar was actively forming up to the completion of magmatic crystallization or may represent metasomatic replacement.
Crystals of plagioclase feldspar with mantles of potassium feldspar are common in the least altered samples of phaneritic rock (Figures 26G and H). The mantles of potassium feldspar are magmatic in origin and result from the late stage growth of potassium feldspar into the void spaces between earlier formed plagioclase feldspars. In contrast, other examples of plagioclase feldspar with rinds of potassium feldspar are present only in strongly altered rocks. These rinds are not formed by overgrowth but rather by hydrothermal replacement which has progressed from the rims inward. A variety of crystals consisting partly of plagioclase feldspar and partly of potassium feldspar are shown in Figure 27A-D. These crystals represent arrested stages in the pseudomorphic replacement of plagioclase by potassium feldspar. From observations of such textures, it is possible to infer that many crystals of potassium feldspar in potassically altered rock were originally of plagioclase before hydrothermal alteration ensued.
All varieties of potassium feldspar are counted as "non-polysynthetically twinned feldspar" or "gray feldspar" in point counter analyses and are listed as "NPST feldspar" in Appendices 2 and 3. The average Or contents of potassium feldspar crystals in altered and unaltered porphyries and phanerites range from Or73 to Or80.
Where replacement of feldspar is nearly complete, tiny relicts of the original plagioclase
host may be completely enveloped by secondary potassium feldspar. This texture is herein
referred to as "sieve texture." Feldspars that show this feature superficially appear to be
perthitic. However, the plagioclase feldspar relicts are in optical continuity and thus are more
likely to be unreplaced remnants from hydrothermal alteration than products of exsolution.
Examples of this texture are illustrated in Figures 27E-G.
Pyroxene is present in thin-sections of unaltered and in propylitically altered intrusive rocks having both porphyritic and phaneritic texture. Crystals of pyroxene are subhedral to euhedral in shape in both textural varieties of the intrusions. Elongate prisms (Figure 28A-D) and octagonal cross-sections are common (Figure 28E-F). The crystals are pale green in color and weakly pleochroic in plane-polarized light. Most display even extinction when viewed between crossed polarizers. Pyroxene accounts for 4 to 10 volume percent of the least-altered rocks. Crystals of pyroxene are commonly juxtaposed against or contain inclusions of magnetite (Figure 28 A-H), apatite, or sphene. They are typically equal in size to the feldspars with which they coexist. The longest dimensions of individual crystals range from less than 1 mm to up to about 5 mm. The relative abundance, crystallinity, and mineralogical relationships of pyroxene to feldspar and magnetite are shown in Figure 28G-H.
A map showing the spatial distribution of pyroxene is given as Figure 29. The map shows that pyroxene is a common mineral phase in the Kalgoorlie, Ningi, and Sydney intrusions, whereas it is rare, or entirely absent, in the Mt. Fubilan intrusion.
One hundred and three microprobe analyses were made of pyroxenes in eight polished thin-sections. Representative analyses are given in Table 4. The entire sets of analyses are given in Appendix 5. Structural formulae for the analyses were calculated using the computer program FORMULA (Ercit, unpublished). The formulae were calculated on the basis of four cations and with all iron treated as FeO. The cation percentages of Mg, Ca, and Fe were plotted on triangular diagrams for classification. The classification scheme used (Figure 30A) is that of the International Mineralogical Association, Committee on New Minerals and Mineral Names, Subcommittee on Pyroxenes (Morimoto and others, 1988). The nomenclature of the Commission is simplified from older classifications and contains fewer names. The Subcommission's recommendations on the use of the term diopside and the dropping of salite as a mineral name are of particular interest to this study.
All of the analyses of pyroxene (Figure 30B and C) from the analyzed samples fall within the field of diopside as defined by the Commission and include examples that would have been categorized as diopside and as salite under older classifications. The terminology used in this dissertation follows the recommendations of the Subcommittee. The term diopside includes all pyroxenes that would have previously been named salite and diopside. However the term salite should be kept in mind for comparison to pyroxenes in publications that predate the Subcommittee's report or for those that don't follow these recommendations. Two analyses from one thin-section (DDH 459-102.9; Figure 30C) are characterized by higher CaO contents than the others. This thin-section contains garnet and it is possible that the analyses are of metasomatic pyroxene. Katchan (1982) has noted that skarn clinopyroxenes can be distinguished from igneous pyroxenes by slightly higher CaO and lower Na2O and that they plot above the tie line between diopside and hedenbergite compositions.
Pe-Piper (1984) reported that "recent reviews of orogenic rocks have suggested that the clinopyroxenes in potash-rich rocks may differ from those in lower-K calc-alkali rocks." Whereas augite is characteristic of most calc-alkali rocks, the compositions of clinopyroxenes in potash-rich calc-alkali andesites may extend into the fields of diopside and salite (Ewart, 1982). This is consistent with an interpretation of the Ok Tedi Intrusive Complex as a part of a high-potassium magma suite.
The average value of Fe/(Fe+Mg) in the analyzed pyroxenes is 0.28. In contrast to hornblende and biotite, the analyzed pyroxenes contain low amounts of fluorine. Typically fluorine abundances are below the detection limits in the analyses of pyroxene whereas hornblende and biotite were each found to contain greater than one percent fluorine.
Pyroxene is not present in thin-sections of samples collected from most mineralized locations within the Mount Fubilan, Kalgoorlie, or Ningi Intrusions. However pseudomorphs of mica or smectite after pyroxene in many samples from these areas attest to the fact that pyroxene was present before alteration occurred. With increasing potassic or argillic alteration the pseudomorphs after pyroxene have been destroyed and neither pyroxene nor its pseudomorphs are observed in rocks dominated by intense alterations of these types.
Although pyroxene is not found in most samples of porphyry that display strong potassic
alteration, it was found in a few thin-sections of anhedral metasomatic alkali feldspar phanerite
from the Kalgoorlie area. The samples which contain this association may be the periskarns
observed by Duncan (1972 ). The association is unusual, occurring in only a few specimens,
and confusing in light of the general antithesis of CaO-rich minerals in the potassic assemblage
at Ok Tedi. Further study of rocks of this type is necessary but I was unable to find any samples
during my visit in 1994 and the thin-sections of the first suite of samples (which includes
examples of the association) were not suitable for microprobe analyses.
Two distinct types of amphibole are present in the intrusive rocks at Ok Tedi. These are hornblende and actinolite. The hornblende is magmatic in origin, whereas the actinolite formed by replacement of pyroxene in rocks that have undergone propylitic alteration.
Hornblende is strongly subordinate to pyroxene in both suites of samples examined in this study. It was found only in thin-sections of samples from the Kalgoorlie and Ningi Intrusions, and in two thin-sections from one drill hole in the Sydney Intrusion (DDH-331). Hornblende is scarce or absent in rocks that have undergone potassic alteration.
Crystals of hornblende are anhedral to subhedral in shape, and pleochroic from pale green to medium green in color. A few show the two directions of cleavage that are characteristic of amphiboles (Figures 31B, C, and D), but others display only a singe direction or none at all (Figures 31A and B). Some hornblende may have formed by replacement of pyroxene. A possible example of this is depicted in Figures 31E-F where hornblende clearly has enveloped and may have partly replaced magmatic pyroxene. Actinolite is characterized by fibrous habit and pale green color. Many clusters, or fibre bundles, of actinolite formed by pseudomorphic replacement of magmatic pyroxene, hornblende, and biotite (Figures 31G-H) are present in propylitically altered rocks.
Eighty-seven analyses were made of hornblende and actinolite in five polished thin-sections. Representative analyses are given in Table 5. The entire sets of analyses are given in Appendix 6. Their mineral formulae were calculated using the computer program AMPHIBOL (Richard and Clarke, 1990). This program produces five sets of output based on various assumptions that can be made about the iron and water content of the samples. The mineral names assigned here were taken from the standard calculation method which assumes that all iron is present as Fe2+ and on a basis of 23 anions.
All of the analyzed amphiboles from the Ok Tedi Complex are calcic according to the classification scheme of Leake (1978) shown in Figure 32. The hornblendes are edenite in composition (Figure 32A) whereas the actinolites are actinolite or actinolite hornblende (Figure 32B).
Several microprobe analyses of pyroxene were carried out using an amphibole analytical file. These are shown plotted on an amphibole classification diagram in Figure 32C. It can be seen from this figure that the analyzed actinolites are closely similar in chemistry to pyroxene. This similarity supports an interpretation that much of the actinolite in rocks from the Ok Tedi complex is derived from magmatic pyroxene by the process of metasomatic replacement.
The average value of the ratio Fe/(Fe+Mg) in this group of amphiboles is 0.26. The average fluorine content of 53 edenites is 1.75 weight percent, of 17 actinolitic hornblendes it is 1.10 percent, and of 13 actinolites it is 1.03 percent. The fluorine content of the magmatic amphiboles is higher than that of the alteration amphiboles.
A map showing the spatial distribution of hornblende is given in Figure 29. Hornblende
was found in one drill core sample each of feldspar porphyry and phanerite from the Sydney
Intrusion north of Folomian (DDH-331) and in several specimens from the Kalgoorlie and Ningi
intrusion. In addition, hornblende was present in dikes of hornblende porphyry in the Mt.
Fubilan and Kalgoorlie intrusions before replacement by hydrothermal biotite.
Micas of the biotite group that have formed by both magmatic and hydrothermal processes are present in the rocks of the Ok Tedi Intrusive Complex. Ayres and Bamford (1976) recognized three distinct types of biotite that they believed could be distinguished petrographically: (1) dark brown laths and euhedral crystals with textural relationships that suggest they are probably early formed and magmatic in origin; (2) brown to red-brown laths or aggregates that replace pyroxene or hornblende at many locations; and (3) colorless to light-brown laths which are more phlogopitic in composition and which form veins and replacements. These authors believed that the last two types are of hydrothermal origin.
Magmatic biotite is characterized by euhedral or subhedral shape and even extinction between crossed polarizing filters (Figures 33A-C). It is typically pleochroic from dark brown to pale yellowish brown. Many examples of magmatic biotite display the polygonal shape typical of micas and, in cross-section, parallel planes of exfoliation or cleavage lamellae. Magmatic biotite is commonly associated, or intergrown, with pyroxene, magnetite, apatite, or sphene (Figure 33D).
A crystal of magmatic biotite showing evidence of resorption is shown in Figures 33E and F. The resultant crystal is skeletal in form and is surrounded by a ring of magnetite granules. Resorption surfaces in feldspar were also noted earlier in this chapter.
Hydrothermal biotite displays a wider range of pleochroic colors and crystal shapes than does biotite of magmatic origin. The maximum pleochroic colors include: orange, red, and yellowish brown. The minimum pleochroic color is pale yellowish brown. Most crystals of hydrothermal biotite are ragged in shape and form shreds, flakes, and patchwork aggregates; typical examples are shown in Figures 33 G-H and Figures 34A-B. Many crystals, and crystal clusters, are secondary after earlier mafic minerals (pyroxene, hornblende, or magmatic biotite). Others form along veinlets or at apparently random sites.
Hydrothermal biotite is associated with rutile or quartz in many thin-sections. Intergrowths of biotite and quartz are common and together they form pseudomorphic replacements of earlier ferromagnesian minerals as shown in Figures 34C and D).
Two pseudomorphs of hydrothermal biotite that have replaced hornblende in hornblende porphyry are shown in Figures 34E-G. An example of hydrothermal biotite that has replaced preexisting mafic pyroxene (or hornblende) in a sample of texturally destroyed rock is shown in Figure 34H.
Veinlets and veins of hydrothermal biotite are locally abundant in areas of potassic alteration. Complex mesh-like networks of hydrothermal biotite were also noted in many thin-sections.
Microprobe analyses were made on 58 biotites present as phenocrysts and other crystal forms of mica in six polished thin-sections. Representative analyses are given in Table 6. The entire sets of analyses are given in Appendix 7. These analyses are from both magmatic and hydrothermal biotites.
The compositions of all micas analyzed as part of this dissertation are shown in Figure 35 a compound classification diagram based on the end-member biotite minerals: annite, phlogopite, eastonite, and siderophyllite (Speer, 1984; and Guidotti, 1984). The upper part of the figure is a plot of the ratio Fe/(Fe+Mg) versus the number of aluminum cations assigned to the octahedral site and the lower plot shows Fe/(Fe+Mg) versus the number of aluminum cations assigned to the tetrahedral site. The analyzed compositions cluster near the vertical dashed line at Fe/(Fe+Mg) = 0.33. This value has been used to distinguish biotite (>0.33) from phlogopite by Mason (October 1993). Magmatic biotite (Figure 35A, B) is characterized by higher Fe/(Fe+Mg) ratios (average=0.33) than hydrothermal biotite (Figure 35C, D average=0.26). Many of the analyses of magmatic biotite have Fe/(Fe+Mg) ratios lower than 0.33; therefore, this ratio cannot be considered an unequivocal method for distinguishing between igneous and metasomatic biotites. Optical properties are the best criteria for discrimination in this case. The average value of the ratio Fe/(Fe+Mg) for the hydrothermal biotites is 0.26. This value is closely similar to the value of the same ratio for pyroxenes (0.28) and amphiboles (0.26) from the Ok Tedi Intrusive Complex and may indicate that the hydrothermal biotite that was analyzed was formed by pseudomorphic replacement of magmatic pyroxene or hornblende. The values of tetrahedral Al calculated from many analyses of magmatic and hydrothermal biotites plot below the tie-line between phlogopite and annite. These micas can, therefore, probably considered aluminum-poor.
Nockolds (1947) reported that the amount of Al2O3 with respect to MgO and FeO* in biotite mica is dependant on the nature of the other "mafic" minerals associated with it. The Al2O3 content is greatest in biotites associated with topaz or muscovite (Nockolds extended the use of the term mafic to include these minerals), intermediate in biotites that are not associated with other "mafic" minerals, and lowest in those associated with hornblende or pyroxene. He noted that the ratio of MgO to FeO* is not related to the paragenesis but to the degree of differentiation or contamination of the magma from which the biotite crystallized. The amounts of Al2O3, MgO, and FeO* (recalculated to 100 percent) in biotites from the Ok Tedi Intrusive Complex are plotted in Figure 36. This diagram has fields representing the paragenesis of the biotites (from Nockolds, 1947). Almost all of the analyses of biotite from the intrusive rocks at Ok Tedi fall in field of micas that are expected to coexist with pyroxene or hornblende. This is true for both magmatic and hydrothermal biotites. Hydrothermal biotite can be distinguished from magmatic biotite in Figure 36 by having a slightly lower ratio of FeO* to MgO. Both types have similar Al2O3 content but that of hydrothermal biotite is slightly less than that of magmatic biotite. The fact that pyroxene coexists with magmatic biotite in the least-altered intrusive rocks of the Ok Tedi Intrusive Complex is consistent with the Al2O3-poor nature of the biotites plotted in Figures 35A and B. The amount of Al2O3 in hydrothermal biotite also is consistant with its presence in zones of potassic alteration and with the very low amount of sericite (muscovite) in this environment.
The average fluorine content of the magmatic biotite is 1.76 weight percent whereas that of the hydrothermal biotite is 2.81 percent. This suggests that the fluorine content of the hydrothermal fluids may have been higher than that of the magma. TiO2 averages 3.50 weight percent in magmatic biotite and 2.16 percent in hydrothermal biotite. The average MgO content of magmatic biotite is 15.9 percent whereas that of hydrothermal biotite is 18.5 percent.
A map showing the spatial distribution of magmatic and hydrothermal biotite is given in
Figure 37. All, but one, of the thin-sections from the Sydney Intrusion have only magmatic
biotite. The remaining sample is from the Darien-East Cheam area and may be from rock
associated with skarn alteration. A few thin-sections of samples from the Kalgoorlie and Ningi
intrusions contain only magmatic biotite, but most contain either hydrothermal biotite alone or
combinations of magmatic and hydrothermal biotite. Thin-sections of all but a few samples
from the Mt. Fubilan Intrusion contain hydrothermal biotite exclusively, and this intrusion is the
principal host to the Ok Tedi ore deposit.
