Ketilidian uranium mineral occurrences in South Greenland

A notable contrast between the western and eastern parts of the lower Eriksfjord Formation is seen in the relative amounts of basic and evolved lavas. In the west, the vo1canics in the lower part of the Eriksfjord Formation are almost exclusively basaltic (Stewart, 1964), whereas the eastern remnants around, and within, the Motzfeldt Centre have an overwhelming preponderance of differentiated lavas, mostly trachytes, while basalts apparently constitute less than 10 per cent of the vo1canics. The only other sequence with differentiated lavas is the Ilimaussaq Vo1canic Member from the upper westernmost part of the Eriksfjord Formation around the Ilimaussaq intrusion. The trachytic and trachyandesitic lavas make up about one third of the vo1canics in this member (Larsen, 1977). There is an apparent correlation between the occurrence of differentiated lavas and later syenitic intrusive centres. Even though the lavas and syenites in one place are not directly related petrogenetically they may owe their origin to the same structures in the deep crust or at the crust-mantle boundary, which facilitated the development of differentiated magrnas. Such structures could be cupolas on larger magma pools.

A notable contrast between the western and eastern parts of the lower Eriksfjord Formation is seen in the relative amounts of basic and evolved lavas. In the west, the vo1canics in the lower part of the Eriksfjord Formation are almost exclusively basaltic (Stewart, 1964), whereas the eastern remnants around, and within, the Motzfeldt Centre have an overwhelming preponderance of differentiated lavas, mostly trachytes, while basalts apparently constitute less than 10 per cent of the vo1canics. The only other sequence with differentiated lavas is the Ilimaussaq Vo1canic Member from the upper westernmost part of the Eriksfjord Formation around the Ilimaussaq intrusion. The trachytic and trachyandesitic lavas make up about one third of the vo1canics in this member (Larsen, 1977). There is an apparent correlation between the occurrence of differentiated lavas and later syenitic intrusive centres. Even though the lavas and syenites in one place are not directly related petrogenetically they may owe their origin to the same structures in the deep crust or at the crust-mantle boundary, which facilitated the development of differentiated magrnas. Such structures could be cupolas on larger magma pools. igdlorssuil Igdlorssuit is located at the northcrly Jimil ol' tlle fjord system abouI 60 km north af Kap Farvel ((i0023'N; 46°06'\V). Thc uranium minern! showing is Oll a smal! alp an the castem side, 500 m vertically ubovc (hc fjord. Detailed plane table rnapping (l: 1(00), and dcwilcd radiomctric measurernents with a lead col1imated scintillomctcr calibraled for uranium delineatecl this occurrcllcc. Regional marring (I: IO (00), however. showed that il was onll' OIlC af many similar uranium ucnlrrences in the afea albeit the largest and richest.
Tllc afca contains a series af Tafts ur pendants ol' rnetasedilllcntary and metavolcanic supracrustal rocks dirring gently to the north-east in rapakivi granite. Tlle uranium minerals are eithcr disseminated <dang the layering in the supracrustal rarts, ar more usually C011CCI1trated in small fraetures which are thcmselvcs stratahound.
Isoclinal folds and sedimentary features indicating overturned beds have been obsen'ed so it ean be safcly assumcd that the present layering does not represent the original bedding sequencc, <"lnd that repetition af the beds, due to structural disturhances and the intrusion af the rapakivi granite, ean be expected.
The supracrustal units have been divided an [he basis af their texture and mineralogical composition illlO rive units. Their distribution is hest illustrated an an uncontrolled section laken from an obliquc phOIO because af the steepness of (he area wllich averages 50 0 below the alp, and 3D' ahove il ( fig. 21).
The most easily rnappable unit is the mClaconglamerate. cabbles af felsie compu:)ilion in a c,)!c-silicate matrix. Thc cobb1cs have been slighrly deformcd hy Icctonlsm. Thc unit outcrops over a horizootai distance uf 250 m and has a maxi-Illum widtl1 af 50 rn with its southcrn limit l'oTmcd by the crest of a Tcc\ining antiform whh il horizollral NNW-SSE axis. It is pinched OUI by the rapakivi granite to thc nortl\. It cOlltains only one vcry rich (1.0-15% U) uraniferous fraeture which strikcs approximatcly NNE-SSW and ean bc traeet! over some 50 m ( fig. 21). "fhe fraeture and {he mineral rich zone is narrow (5-20 cm). and the uranium minerals ean only bc found at cithcr end of the fraelure and nol cOlltinllollsly along il.
The conglomerate is apparently overlain by a metavolcanic tuff of intermediate composition. This is a medium-grained, massive brittle rock which is green when fresh and weathers white, and in some places the weathered surfaces have a lithic or crystal tuff texture. It is the most common supracrustal rock type in the area and a common host rock for the uranium minerals. It is often broken up into large angular blocks within the rapakivi granite, which suggests that it was a relatively brittie rock type at the time of the intrusion of the granite.
The metavolcanic unit is interlayered with a mafic rich volcanic unit and a meta-arkose. The meta-arkose is a fine-grained, grey granular rock with sedimentary layering. It has been differentiated from the metavolcanic unit purely on the basis that it contains sedimentary structures, and it may well contain waterlain tuffs as well as purely sedimentary material. It varies from 5-20 m in thickness and can be traced along strike for considerable distances. One horizon, just below the conglomerate is partieularly uraniferous, hosting numerous uranium occurrences, ranging between 0.01 and 0.3% U over a horizontal distance of some 500 m. These are mostly in minor cross-fractures but also occur along the layering which may possibly correspond to the bedding.
The pyroxenite is composed mostly of medium to fine-grained dark green pyroxene, which in places forms up to 90 per cent of the rock, with coarse biotite and other mafie minerals in a grey plagioclase matrix. A lithic tuff texture can occasionally be seen on weathered surfaces in less structurallY disturbed areas. As the lensed outcrop pattern forms boudins up to 50 m wide and 200 m long it is suggested that this rock unit was less competent than the other units during metamorphism and folding, and the pyroxenite was forrned by metamorphic recrystallisation of either basic volcanic units or basic intrusive sills.
The sulphide-rich metasedimentary unit is, however, intimately associated with the mafic rich volcanic unit and usually outcrops immediately above or below it. It is not usually more than 3 m thick and rarely outcrops along strike for more than 200 m. It is a grey rock, when fresh, of medium to fine grain size with a granular texture and disserninated sulphide minerals in a quartz-feldspar matrix, but it is usually friable, rusty and deeply weathered. The sulphides often include appreciable amounts of molybdenite. Uranium minerals are not usually present.
The supracrustal rocks are intruded by two granite units. The older of the two outcrops only in a relatively small raft in the younger rapakivi granite in the mapped area. It is medium-grained biotite granite with a rather high radioactivity of up to 280 ur*. This is interpreted as being a remnant of an earlier migmatitic granite (Bridgwater et al., 1966), and its high radioactive background reflects the uraniferous nature of the parent supracrustal rock.
Rapakivi granite is the major rock unit in the area and is uniform in both composition and texture. It is very coarse grained with large ovoids of orthoclase feldspar, blue quartz and accessory biotite, and occasionally hornblende. It is a 'biotite pyterIite' in the Finnish classification system for rapakivi granite (Vorma, 1976) because the feldspar ovoids do not have any well developed plagioclase rims. In contrast to the older granite it has aremarkably constant and relatively low radioactivity (25-28 ur) even when adjacent to uraniferous radioactive supracrustal units. This is interpreted as reflecting its largely intrusive allochthonous nature. On a broad scale it appears to have intruded preferentially along the planar struc-*1 ur is a unit of gamma-radiation which is proportional to the radiation from 1 part per million uranium by weight. rections, but they ean also hc observed along the sides uf the main glacier-fjord valley trcndĩ ng in a NNW-SSE direetian, which is also an important fraeture direetion. They are COIllposed af finc-grained to aphanitic. dark green to black rock which is occasionally vesicular.
Thc:y have been classified as Gardar dykes by previous workers (Sutton & Wattcrson,lY6:>;) but (his has never been establishcd with certainty. Stmcture. Prior 10 the intrusian af the rapakivi granite, the supracrustal units have becn iso-clinaHy folded and fractured. The isoc1inal folding was presumably a synmetamorphic feature when (he rocks were re1ativeiy hot and plastic. In contrast to other areas af the Migmatite Complex (Dawes. 1970;Escher, 1966), only one period of isociinal folding has becn rccognised so far. This could well provc to be an oversimplification if the mapping is extended to areas less disflipted by the rapakivi granite, Tile axis af tlle folding trends more Ol' less NNW-SSE with a horizontai plunge, and Ihc axial plane dips gently (1G-300) to thc WNW. Fig. 22 illustrates somcthing of the sharpness of this folding. At locality B ( fig. 21) the maric vo!canic unit is terminated in a sharp antiform which is overlain by intermediate mctavolcanic rocks which at this point are highly uraniferous (2.5% U). This suggests thal the folding was one of the episodes that helped to mobilise and cOl1centrate tlle uranium minerals.
The pre-rapakivi fraeturing is characterised by narrow (0.5-1.0 cm), irregular fraetures without a prcfcrrcd orientation, which are often curved in ptygmatic-like folds, but alsa tend to be en ecllelon and/ol' to horsetail into smal! breccia zones ( fig. 23). Their form suggests relatively brittie deformation compared to the isaclinal folding. They are usually filled with marie minerals, magnetite, and cOlTllllonly but not always, host uranium minerals. They have not bccn obscrvcd l:utting the alder granite and are always terminated by the rapakivi granite. They are, thcrefore, alder than the rapakivi granitc and possibly ulder than the ulder granitc_ Their mode af formation is uncertain alrhough they could be related to either late stage teetanie evcnls connceted with the regional metamorphism ar to tcctonic adjustments prior to the inrrusian af {he rapakivi granite. Post-magmatic faulting is only of minor significance within the field area where E-W faults have a minor dextral sense of movement of a few metres, and two inc!ined faults dipping moderately (40-50) to the NE have a smal! (2-3 m) reverse sense of movement.
Uranium mineral occurrences. Over 35 uranium mineral occurrences have been found scattered over the hillside (fig. 21). They tend to be restricted to certain members of the metaarkose and intermediate metavolcanic units. The uranium bearing mineral is uraninite which is disseminated as fine grains through the strata or concentrated as medium sized grains along the pre-rapakivi fractures and associated breccias. This mineralisation, therefore, can be classified as stratabound but local!y control!ed by folds and veins where it has been concentrated by structural and metamorphic events ( fig. 24). Both the lithological setting and structural control of this mineralisation are similar to that found at the Kitts and Michelin deposits in Labrador (Gower et al., 1982).
The largest, highest grade uranium-mineral zone remains the area that was found by B. Wallin in 1982  and mapped with plane table this season ( fig.  21, A-A). Radiometric readings with the calibrated scintillometer were taken over a grid (115 ± 10 m) at 1 m intervals every 5 m in order to measure the extent and value of the surface grades. The results outlined a main zone 50 m long by 1-3 m wide with over 0.1 % U, and a central zone about 15 m by 2-3 m with grades of over 1.2% U and a maximum point source of over 2.5% U. To the north this unit is broken up into a series of rafts in the rapakivi granite with some radioactive anomalies. To the south the same unit is similarly broken up and dips below the accessibie outcrop level. Because of the massive, structureless nature of the host rock it has not been possibIe to relate the uranium minerals in this showing to any particular feature within the rock, but it seems likely that its brittIe nature contributed to its fracturing, which would give more ready access to the uranium-rich fluids and form traps for the uranium minerals.

Conc!usions on Igdtorssuit uranium showing.
The results of the mapping and sampling have established that this type of mineral occurrence can reach economic grades, and its surface expression suggests a size which could approach economic proportions. The wide distribution of occurrences suggests that uranium was present over larger area than just Igdlorssuit. The stratabound nature of the mineralisation suggests that the uranium was aIready present in the supracrustal units but was mobilised by the regional metamorphism and concentrated in zones of lower pressure such as on the crests of folds, as at locality B, or in the fractures.
This type of mineral occurrence probably accounts for the many uranium anomalies identified in the Migmatite Complex, both by the airborne gamma-spectrometer and geochemical sampling, and should constitute the type of target to be sought in future exploration in the area.

Uranium mineral occurrences on the nunatak north of Nordre Sermilik
In 1982 B. Wallin found uranium-rich boulders (Armour-Brown etat., 1984) and defined a gamma-spectrometer anomaly, fol!owing a detailed helicopter gamma-spectrometer survey  on the nunatak which lies to the north of Nordre Sermilik ( fig.  20). The disseminated character of the uranite in biotite gneiss suggested that it was formed earlier than the pitchblende associated with the Gardar veins which have been found previ-ously in the Granite Zone and eould, therefore, eonstitute both another type of mineral oeeurrenee in the area, and be a souree of the uranium in the veins.
Mapping at 1: 10 000 seale showed that the geology of the nunatak is dominated by weakly foliated and massive members of the Julianehåb Granite whieh eontains sub-horizontal rafts of mostly feldspathie gneiss with some biotite and magnetite-rieh gneiss. In general, these rafts are distributed as sub-horizontal sheets and are similar to the supraerustal rafts in the rapakivi granite at Igdlorssuit. They also have a eonsiderably higher radioaetive baekground than the surrounding granite (70-700 ur as eompared to 10-20 ur). The gamma-speetrometer anomaly oeeurs where these uraniferous gneissie units are thiekest and are exposed on a cliff on the south-east faeing slopes of the highest summit whieh eoineides with the flat-lying erest of a broad antiform of the rafts. This would be enough to aeeount for the 25 ppm eD anomaly measured from the helieopter.
Many other uranium-enriehed gneissic rafts were found on the nunatak, partieularly in the northem part. The riehest of these proved to have eoneentrations of uranium up to 1.3% and thorium values of 1131 ppm. These values oeeur disserninated in a raft of feldspathic gneiss surrrounded by weakly foliated, eoarse biotite granite. The raft measures 30 X 7 m and varies from 1.5-2.0 m in thiekness and has a very irregular outerop pattern as it lay on a dip slope. Any important down-dip extension of this raft ean be precluded beeause of its thinness and general disposition. The whole raft is uraniferous but the rieher part, whieh eontains over 1% D, measures 5 x 2 x 1 m. It is not possible, on the basis offield evidenee, to deduee the parent material of the feldspathie gneiss. Its feldspathie eomposition suggests that it is a member of the acid volcanic roeks whieh have been mapped in the area (Allaart, 1983). The very high D/Th ratio also suggests that the uranium was present in the parent material of the gneiss, sinee partitioning of these two elements normally takes plaee at temperatures well below those indieated by the prevailing gneissic eondition of the bedroek (Langmuir, 1978) and thorium usually dominates over uranium under sueh eonditions.
Canciusians an the uranium accurrences an the nunatak. The most important eonclusions that ean be drawn from the field results deseribed above are that uranium minerals were present in the pre-Gardar Ketilidian gneisses, and that it oeeurs as uraninite. This type of oeeurrenee is, therefore, a possibIe souree for the uranium in the Gardar veins. This eould have been dissolved, transported and redeposited either by meteorie W'ater during the subaerial Gardar erosion or by the hydrothermal aetivity re1ated to the Gardar igneous events or a eombination of the two. They constitute another type of uranium mineral oeeurrenee whieh ean be expeeted and sought after in any future mineral exploration in the Granite Zone.