CONTRIBUTIONS TO THE GEOLOGY OF MINERAL DEPOSITS
_________________
EXPLORATION
FOR PORPHYRY METAL
DEPOSITS BASED ON RUTILE
DISTRIBUTION-A TEST IN SUMATERA
_______
By ERIC R. FORCE, -1- SURKIRNO DJASWADI, -2-
and THEO VAN LEEUWEN -3-
________
ABSTRACT
At the Tangse porphyry-copper prospect, rutile in thick soil reflects the distribution of the quartz-sericite and biotite-chlorite zones of hydrothermal alteration at depth. Detection of rutile in the samples is not simple, but studies of rutile distribution may nevertheless be a cheap exploration method for tropical porphyries.
INTRODUCTION
A program of investigation is being undertaken cooperatively by the Indonesian Directorate-General of Mines and the U.S. Geological Survey (USGS), sponsored by the Government of Indonesia and the U.S. Agency for International Development; it includes this study.
Recent work has documented the presence of rutile in the most severely altered parts of porphyry alteration systems (Lawrence and Savage, 1975; Force, 1976a; Williams and Cesbron, 1977; Force and others, 1980; Force, 1980a; Llewellyn and Sullivan, 1980; Czamanske and others, 1981). Because this rutile has a related origin and similar distribution to the copper mineralization, knowledge of rutile distribution should be useful in exploration for porphyry copper, as suggested by Lawrence and Savage and by Williams and Cesbron. As rutile is resistant to weathering, determination of rutile distribution in soil could be an important of an exploration method where porphyry deposits are concealed by thick soils leached of copper. We suspect, on the basis of knowledge of rutile occurrence summarized by Force (1976b, 1980b), that in the volcanotectonic arcs where many of these porphyries occur, the mere presence of
rutile is an indication of porphyry-related alteration in the broad sense. This hypothesis need further checking.
_____________
-1- U.S. Geological Survey.
-2- Directorate of Mineral Resources, Bandung, Indonesia.
-3- Rio Tinto Indonesia, Jakarta, Indonesia.
_______________
ACKNOWLEDGMENTS
This work would not have been possible without the help of Umar Olii of Rio Tinto, Indonesia, P. H. Silionga of the Directorate of Mineral Resources (Indonesia), W. W. Olive, Jr., of the U.S. Geological Survey, and their colleagues.
RUTILE DISTRIBUTION IN PORPHYRIES
Czamanske and others (1981) provide the most comprehensive picture so far of the distribution and origin of rutile in porphyries. Rutile is a secondary mineral that mimics the distribution of original magmatic titanium minerals such as sphene, biotite, and ilmenite. It reaches its greatest abundance and grain size in the biotite potassium feldspar alteration zone of porphyries in the Western United States. There it averages 0.3 percent or more as crystals, averaging about 0.03 by 0.06 mm, locally in mosaics where it forms pseudomorphs of primary, magmatic titanium minerals. In peripheral alteration zones, rutile abundance and grain size progressively diminish. In some porphyries, the distribution of rutile and of copper ore is about the same.
APPLICABILITY TO EXPLORATION IN THE TROPICS
The correspondence of rutile distribution to certain alteration zones, coupled with its resistance to weathering in soils, suggests that an exploration method for porphyry deposits could be based on rutile distribution. In the deeply weathered tropical terranes where this approach would be most useful, however, many porphyries are of different compositions and have different types of alteration than the Western U.S. quartz monzonitic porphyries most intensively studied by Czamanske and others. A few data suggest that these differences do not detract from the potential usefulness of rutile. Lawrence and Savage (1975) and Cox and others (1973) described quartz dioritic porphyries that contain rutile. We find that advanced argillic alteration assemblages with andalusite (common except in Western United States) contain rutile.
Two types of exploration with rutile distribution seem posible: proximal exploration in soils and local streams, and distal exploration in sediments of large streams.
THE TANGSE PORPHYRY
The Tangse porphyry-copper prospect in Aceh province, northern Sumatera (fig. 1), has been briefly described by Young and Johari (1978), Page and others (1979), and Taylor and van Leeuwen (1980). Subsequent geological mapping, geochemical sampling, ground magnetics, induced polarization, and diamond drilling (1,600 m) by Rio Tinto Indonesia have resulted in a more comprehensive documentation of the deposit.
The prospect area is near the confluence of two major rivers, Krung (river) Tangse and Kr. Bale (fig. 2). It forms a topographic depression, occupied by alluvial flats and low, flat-topped hills within the Barisan Range, a rugged mountain range that runs along the entire western edge of the island of Sumatera. Following closely the crest of the Barisan Range is a continuous system of axial valleys, including the Kr. Tangse valley, which marks the outcrop of the main fault line of the Sumateran fault system. This is essentially a right lateral fracture system, although gravity faulting is also important (Katili and Hehuwat, 1967; Page and others, 1979). Several other occurrences of porphyry copper are found along this fault zone farther to the southeast (Taylor and van Leeuwen, 1980).
The topographic morphology of the Tangse area is subdued because the rocks here are strongly fractured and altered. Primary copper mineralization is largely confined to an elongated multiphase stock consisting of various quartz diorite and dacite porphyries and having plan dimensions of 6 1/2 km by 2 km (northwest part shown in fig. 2). This stock was intruded into a large composite pluton of granitic to dioritic composition, which was emplased in a thick sequence of Mesozoic metavolcanic and metasedimentary rocks. The long axis of this intrusive complex is alined between two obliquely converging fault zones belonging to the Sumateran fault system. A major feature of the Tangse part of the fault system is the large mass of serpentinized ultramafic rocks. Numerous dikes (mostly postmineralization) cut the intrusive complex and adjacent wall rocks. Potassium-argon ages, determined on hornblende or biotite from five samples, indicate a middle Eocene age for the pluton and a middle to late Miocene age for the mineralized stock and late dikes.
Alteration at Tangse is multistage, and telescoping of alteration types has taken place. Fracture-controlled phyllic and advanced argillic alteration assemblages, the latter characterized by the presence of andalusite, have been superimposed on earlier biotite alteration, which has affected virtually the entire quartz diorite stock. The secondary biotite has also been selectively altered to chlorite throughout the stock, although the conversion is only locally complete. An extensive propylitic halo surrounds the strongly altered stock, but otherwise the areal distribution of alteration types does not conform to a zonal sequence even though the temporal relations are clear.
Primary sulfide minerals are pyrite, chalcopyrite, and molybdenite, which are present as disseminations among rock-forming minerals and in veinlets. Rocks showing only early-stage alteration seldom have total sulfide contents of more than 1-2 volume percent; rocks affected by late-stage alteration usually have total sulfide contents of more than 3 volume percent. Primary copper mineralization is widespread, although generally of low tenor, and is found in association with all alteration types, except propylitic alteration. The best mineralization is found in fault-controlled zones of chlorite-sericite-quartz alteration. Chalcopyrite is nowhere observed at the surface owing to strong weathering. Some chalcocite is commonly present directly below the oxidation zone over a relatively short interval. Zinc and lead form a well-defined geochemical halo to the zone of copper-molybdenum mineralization, but gold is absent.
Secondary rutile has already been detected under the microscope in several core and weathered outcrop samples before the present study began. It forms both single tiny crystals and massive to skeletal finely granular clusters. Some clusters appear to form pseudomorphs of former Fe/Ti oxide crystals, but more commonly the rutile is intimately associated with masses of chlorite with or without secondary biotite; this rutile is probably the byproduct of chloritization (and secondary biotitization?) of titaniferous mafic minerals, such as magmatic biotite. The common occurrence of zircon crystals within the rutile clusters supports this interpretation. Rutile is also commonly present in alteration assemblages that contain little or no chlorite (quartz-sericite; quartz-sericite-andalusite). In these associations, it is usally enclosed in sericite masses. Whether the rutile survived overprinting of biotite-chlorite alteration by later phyllic and advanced argillic alterations, or whether it is directly related to these late hydrothermal processes, has not been determined.
Rutile was not observed in unaltered quartz diorite or in postmineral dikes. The propylitic zone has not been studied in detail, but the available data from thin-section study suggests that rutile is absent in this zone also, even where it overlaps the zone of secondary biotite alteration. Sphene, however, is very common in the propylitic zone. Though our knowledge of rutile distribution in unweathered rock is sketchy at Tangse, it is in accord with results of Czamanske and others (1981) and of Williams and Cesbron (1977), who studied rutile from a large number of porphyry copper deposits. They observed that rutile may be present in the inner fringes of the propylitic zone, and is found throughout more intensely altered zones, but disappears outward in favor of the local titanium-bearing accessory in the host rocks.
SAMPLING METHODS
Soil samples were collected from about the upper meter of exploration trences. Stream sediments were from large active streams upstream and downstream from the deposit and small streams within the deposit. We found that most of the rutile was too fine to be concentrated in a pan. The best sample proved to be a -80 mesh screen fraction from which the clay-size material was decanted. Most of these fractions were prepared in the field. Bulk samples and +80 mesh pan concentrates were also collected for insurance.
LABORATORY METHODS
The rutile is too fine to be identified with confidence under a binocular microscope. In thin sections of weathered rock, we observed that goethite(?) and rutile crystallites were so similar that they were difficult to differentiate. Accordingly, we treated samples with acid to remove goethite, and examined them in grain mounts in oils under a petrographic microscope. Immersion in unheated but concentrated hydrochloric acid for 2 hours proved to be the least drastic treatment that worked. We identified rutile with condenser engaged under a high-power objective, using both plane and crosspolarized light. Some of the rutile was present as inclusions.
Presence or absence of rutile was determined in numbered, but otherwise unlabeled, samples by the first and second authors working independently. We examined pan concentrates also, and, though some rutile was identified, no information resulted beyond that obtained from -80 mesh fractions.
RECOMMENDED METHOD
A simple but effective method for rutile determination is (1) use an aliquot of a sample collected for soil geochemistry; (2) digest it in cold hydrochloric acid for 2 hours; (3) rinse, allowing the clay-sized material to escape; (4) remove the coarse fraction with an 80-mesh screen and dry the fines; (5) identify in grain mount with petrographic microscope, as explained above, and record rutile grain size. This should all be possible in a suitably equipped field office.
RESULTS OF PROXIMAL EXPLORATION
Soil samples.-The distribution of rutile in soil as Tangse correlates closely with the intensity of alteration of parent rock. Rutile is limited to soils over rock that has been altered to quartz-sericite (±andalusite) or biotite-chlorite assemblages. All soils over such rock contain rutile (fig. 2). In addition, the coarsest rutile is found in an axial belt of maximum alteration and sulfide concentration.
We were able to see postmineral dikes in trench bottoms and avoided taking soil samples over them. An exploration program based on rutile distribution without trench exposures, however, would have to allow for postmineral dikes that would yield samples without rutile in intensely altered and mineralized areas.
Stream sediments.-Three samples of sediments from short streams draining the deposit were analyzed (fig. 2). The fact that all contained rutile indicated that proximal alluvial sampling as well as soil sampling for rutile could be useful in delineating a porphyry body.
RESULTS OF DISTAL EXPLORATION
Nine sediment samples from the two largest streams were collected; six were downstream of the deposit (most are outside area shown on fig. 2). Rutile was not observed in any of these samples. Two problems are apparent: (1) Massive dilution with others debris has taken place, making rutile hard to find; (2) vigorous winnowing has removed most of the fine-grained rutile and transported it to lower energy depositional sites downstream. Thus, reconnaissance or distal exploration by means of rutile distribution may not be useful.
COMPARISON WITH OTHER EXPLORATION
TECHNIQUES
At Tangse, rutile exploration worked best in proximal samples-that is, in soil samples and in sediment samples from small streams draining the deposit. Thus, exploration based on rutile distribution is most appropriately compared with other proximal exploration techniques such as soil geochemistry and trenching.
Rutile distribution is consistent with results of soil geochemistry at Tangse (Young and Johari, 1978; Page and others, 1979) but can be determined easier and faster. Rutile distribution, like gold distribution, gives information even where other diagnostic elements have been leached from tropical soils. Where gold is absent over mineralized rock, as it was Tangse, prospecting with rutile may be the only effective surface technique.
Our observations of the rutile in soil collected at the top of soil profiles exposed in trenches corresponded well with our observations of rock alteration made on weathered samples at the bottom of the same trenches. Thus, to some extent, knowledge of rutile distribution can make extensive trenching unnecessary.
An integrated technique using soil geochemistry, trenching, and rutile distribution should provide more information at about the same cost as that for present exploration techniques.
Tangse is the wrong place to test the use of rutile in distal or reconnaissance exploration, as Kr. Tangse and Kr. Bale are powerful braided streams carrying immensely more material than is supplied by erosion of the subdued hills underlying the deposit. Our initial results were discouraging as were those of stream-sediment geochemistry for similar distal exploration. A better test could be done where a deposit is nearer a drainage divide.
CONCLUSION
The distribution of rutile in soil over the deeply weathered Tangse porphyry is the same as the distribution of intensely altered rock at depth. With the methods we have described here, the distribution of rutile in soil is not difficult to establish. Thus, rutile studies could be a valuable part of comprehensive exploration programs for porphyry deposits in the tropics. Alluvial prospecting for distant porphyries by means of this technique appears to be inefficient in our somewhat atypical example. More detailed work and tests over other deposits are certainly warranted.
REFERENCES CITED