Author:

Keywords:

Metallogenesis, Karagwe-Ankole belt, Tungsten belt, Rutongo area, Fluid inclusions, Tungsten, Tin

Abstract:

Although not well known by most people, tin (Sn) and tungsten (W) are two metals that are important in our society. Tin is for instance used for the production of corrosion-resistant steel, as a solder and for the production of bronze, while tungsten is used in cutting tools, heavy metal alloys, light bulbs and dental drills. Tin and tungsten deposits can be found worldwide. Tin and tungsten are often closely associated to one another and can even occur in the same deposit. The Mesoproterozoic Karagwe-Ankole belt (KAB) in Central Africa hosts numerous Sn and W deposits. These mineralisations principally occur in pegmatites and hydrothermal quartz veins that are presumably related to granite intrusions. This PhD study focuses on the Sn and W vein-type mineralisations in the central part of the KAB, in Rwanda. This study aims to determine the origin and evolution of the Sn and W vein-type deposits, focussing on the evolution of the mineralising fluids and their composition (geochemistry). The mineralising fluids are studied by means of a detailed petrographical, geochemical and microthermometrical study and by using specialised geochemical techniques (LA-ICP-MS, crush-leach, Raman analysis and stable isotope geochemistry). The absolute age of the ore deposits is determined by means of the 40Ar-39Ar dating technique. The metasedimentary rocks of the KAB in Central Africa have been affected at ~1375 Ma by a major intracratonic bimodal magmatic event (Tack et al., 2010). This event consists of widespread, voluminous S-type granitoid rocks with accompanying subordinate mafic intrusive rocks (Fernandez-Alonso et al. 1986; Tack et al., 2010). These granites have been called G1‑3 in Rwanda and are not associated with mineralisation (Dewaele et al., 2009). At ~1000 Ma, reactivation of structures in the underlying basement caused folding and thrusting of the rocks in the Western Domain of the Karagwe-Ankole belt, as the result of a far-field effect of the Southern Irumide Orogen (Theunissen, 1988, 1989; Fernandez-Alonso et al., 2012). At 986 ± 10 Ma, post-compressional relaxation gave rise to the emplacement of S-type granites, which have been called G4-granites in Rwanda (Tack et al., 2010; Fernandez-Alonso et al., 2012). These granites show enrichment and depletion trends typical for granites associated with hydrothermal mineralisation or rare element pegmatites (Dewaele et al., 2009). After emplacement of the G4-granites, pegmatites intruded at 969 ± 8 Ma (Brinckmann & Lehmann, 1983). This age overlaps with the age of the G4-granites. The pegmatites show a regional mineralogical zonation surrounding the granites (e.g. Varlamoff, 1954a, b). Recently, a study by Hulsbosch (2012) has shown that a geochemical relationship exists between the G4‑granites and these pegmatites. Based on REE-modelling, it is concluded that these pegmatites formed from a last, small melt fraction that remained after fractionation in a chemically zoned G4-granite (Hulsbosch, 2012). Some of these pegmatites are mineralised in columbite-tantalite and cassiterite. Columbite-tantalite mineralisation within the pegmatites has been dated between 975 ± 8 Ma and 966 ± 9 Ma, which overlaps with the timing of the pegmatites (Dewaele et al., 2011). Precipitation of minor cassiterite is related to this early stage of pegmatite crystallisation. The major proportion of cassiterite, however, precipitated during a later metasomatic-hydrothermal overprint of local zones of the pegmatites, associated with sericitisation and muscovitisation (Dewaele et al., 2011). Fluids and metals were introduced along fractures and faults, clearly after pegmatite crystallisation. In the Karagwe-Ankole belt, numerous quartz veins occur that are mineralised in tin and tungsten. In Rwanda, the Sn-mineralised veins are sometimes located in the vicinity of Nb-Ta-Sn mineralised pegmatites (e.g. Musha-Ntunga). In most places however, there is no clear relationship to pegmatites or granites in the field. Sn and W deposits are usually not spatially related to each other. In the central part of Rwanda, the location of the W and Sn vein-type deposits shows a typical structural and lithological control. The W vein-type deposits in the Tungsten-belt are located in the core and the flanks of complex anticlinal structures. The host rock is typically composed of alternating sequences of sandstones and black pyritiferous metapelitic rocks. Two types of W-mineralised quartz veins have been observed: thick bedding-parallel and crosscutting quartz veins that are at high angle to the bedding. The bedding-parallel quartz veins are typically hosted by dark shales, where they intruded cleavage planes sub-parallel to S0. These veins have been interpreted to have been formed during or in a late stage of a compressional deformation event, which is thought to be related to the intrusion of the G4-granites and post-dating the far-field effect of the Southern Irumide orogeny during amalgamation of Rodinia around 1000 Ma. The W-mineralised crosscutting veins are interpreted to have also been formed in a late stage of a deformation event, possibly overlapping in time with the formation of the W mineralised bedding parallel veins. The crosscutting quartz veins formed by opening of pre-existing planar structures (cleavage planes, axial planes, joints ...), some of which were connected to form important pathways for the mineralising fluid. Both vein types are associated with small alteration zones, comprising silicification, tourmalinisation, muscovitisation and precipitation of biotite. Dating of muscovite crystals at the border of the veins indicates that the closure temperature (~350-400°C) was reached between 992.4 ± 1.5 Ma and 984.6 ± 2.4 Ma. These ages are within error identical to the ages obtained for the G4‑granites (i.e. ~986 Ma), but are older than the ages obtained for the Nb‑Ta‑Sn mineralised pegmatites (i.e. between ~975 and ~966 Ma). The W-bearing minerals formed during two different phases. The first phase is characterised by scheelite and massive wolframite. These minerals formed somewhat late in the evolution of the massive quartz veins, sometimes in fractures that crosscut the veins. The ore minerals precipitated from an aqueous-gaseous fluid with a low to moderate salinity (0.6-13.8 eq. mass % NaCl), at temperatures and pressures between 310° and 510°C and 60 and 270 MPa, respectively. The gas phase was dominantly composed of CO2 (53-92 vol. %), with lesser amounts of N2 (6-38 vol. %) and minor amounts of CH4 (1-9 vol. %). Na is the dominant cation in solution, with lesser amounts of Mg (Mg/Na: 0.01-0.41), Ca (Ca/Na: 0.10-0.32), K (K/Na: 0.08-0.11), Ba (Ba/Na: 0.04-0.09), Fe (Fe/Na: 0.02-0.07) and Mn (Mn/Na: 0.01-0.07). Cl is the dominant anion. Tungsten was likely transported as NaHWO4, NaWO4-, HWO4- and WO42-. The dark metapelitic rocks, which typically host the W vein-type deposits, were a controlling factor in mineral precipitation. The metapelitic rocks could have acted as a source of iron and calcium necessary for the formation of the W-bearing ore minerals. Mineral precipitation could have resulted from a combination of falling temperature, addition of volatiles and an increase in pH through reaction with metapelitic wall rocks. In a second phase, scheelite became unstable and was replaced by fibrous ferberite to form pseudomorphs. This could simply have been caused by a decrease in temperature, which makes ferberite more stable than scheelite. Other possible mechanisms are a decrease of the Ca/Fe ratio of the fluid in contact with the ore minerals, a change in pressure and in the chloride concentration. Pyrite crystals inside the quartz veins and in the surrounding rocks could have been altered to supply the necessary iron for this replacement reaction. Afterwards, the quartz veins have been crosscut by a sulphide phase. Scheelite inside the pseudomorphs that had not been replaced by ferberite was altered to secondary tungstates in a late stage. The delta18O values of the W-mineralised quartz veins are relatively homogeneous (14.4 to 16.0‰ V‑SMOW), but the deltaD values vary significantly (-33 to -57‰ V-SMOW). The delta18O‑deltaD signature of the ambient fluids indicates a dominant metamorphic signature. The delta18O data of the massive wolframite samples analysed are 0.3 and 1.8‰ V-SMOW. Their deltaD composition ranges between -93 and -101‰ V‑SMOW. The delta18O values of the fibrous wolframite samples cover a small range, except for one sample (-3.3 to 3.9‰ V-SMOW), and there is a large spread in deltaD values (-87 to ‑133‰ V-SMOW). The delta18O-deltaD signature of the ambient fluids indicates that the W-mineralising fluid interacted with metamorphic rocks that contained organic material and NH4+-rich micas or feldspars. Based on the lithological setting (black shales), the stable isotope data, the similar Pb isotopic composition of sulphides in the veins and in the metapelitic rocks, the gaseous composition of the W-mineralising fluid (N2 and CH4),and their elemental composition, it is concluded that the ore forming fluid from which scheelite and massive wolframite precipitated, have been largely influenced by interaction with metamorphic rocks. However, based on the close spatial link, the structural setting of the ore veins and an overlap in ages for the W-mineralised veins and the G4-granites, a magmatic origin for the W‑mineralising fluid can not be excluded. In this case, the magmatic fluid equilibrated with metamorphic rocks. This magmatic fluid could have been a primary magmatic fluid, which interacted with the metasedimentary rocks or a metamorphic fluid that equilibrated isotopically both with metasedimentary and magmatic rocks. In the second model, the emplacement of G4-granites triggered hydrothermal circulation around the contact zone between the intrusion and the country rocks. The Sn vein-type deposits in the Rutongo area are also located in secondary anticlines. In contrast to the W deposits, the Sn deposits are typically hosted by a number of massive sandstone and/or quartzite units that are separated by units of alternating metapelitic rocks and sandstones. The Sn‑mineralised quartz veins occur in fields or swarms of several hundreds of sub-parallel crosscutting veins that are oriented perpendicular to the fold axis. The veins are associated with intense alteration, comprising silicification, tourmalinisation, muscovitisation and precipitation of biotite. The Sn‑mineralised quartz veins and the associated alteration zone crosscut the foliation planes in the host rock. Structural analysis indicates that Sn-mineralised veins formed due to extension, with the maximum compressive stress oriented at high angle to the bedding. The structural setting indicates a relatively late formation of the Sn-mineralised veins in the general evolution of the Rutongo anticline, after (G4-)granite emplacement. Cassiterite is the only ore mineral in the Sn vein-type deposits in the Central part of Rwanda. It formed in a late stage of the evolution of the mineralised veins. It is intimately associated with muscovite crystals in fractures in the quartz veins. Cassiterite was deposited from an aqueous-gaseous fluid with a low to moderate salinity (6.2-15.4 eq. mass % NaCl) and a minimal temperature of formation between 225° and 349°C. The gas phase was composed of CO2 (50-78 vol. %), N2 (11-40 vol. %) and smaller amounts of CH4 (10-15 vol. %). Na is the dominant cation in solution, with lesser amounts of K (K/Na: 0.14), Li (Li/Na: 0.08) and Cs (Cs/Na: 0.07). Cl is the dominant anion. Tin was probably transported in solution to the site of deposition as Sn2+-chloride complexes. Precipitation of cassiterite could have resulted from a pH increase of the mineralising fluid, which was caused by muscovitisation of the host rock and possibly transformation of CO2 to CH4 in the mineralising solution. Afterwards, the quartz veins have been crosscut by a sulphide phase. The sulphide-mineralising fluid was a moderate saline (13.0-18.3 eq. mass % NaCl) H2O-(X)-CaCl2-NaCl fluid, with a minimal temperature of formation between 249° and 265°C. The delta18O values of the Sn-mineralised quartz veins are similar, i.e. 13.6 and 14.1‰ V-SMOW. The deltaD values show a larger spread between -47 and -67‰ V-SMOW. The delta18O-deltaD signature of the ambient fluids indicates a metamorphic and possibly magmatic signature. Both delta18O and deltaD values of the cassiterite samples are relatively homogeneous (7.5 and 7.9‰ V-SMOW delta18O, and -90 and -105‰ V‑SMOW deltaD). The delta18O-deltaD signature of the ambient fluids indicates that the Sn-mineralising fluid interacted with metamorphic rocks that contained organic material and NH4+-rich micas or feldspars. Based on the stable isotopic composition of the Sn-mineralised veins, the similar Pb isotope ratios in the sulphides of the Sn mineralised veins and the country rocks, and the gaseous composition of the Sn-mineralising fluid (N2, CH4), it is concluded that the Sn-mineralising fluid strongly interacted with metamorphic rocks that contained NH4+-rich micas or feldspars and organic material or graphite. Fluids that predate cassiterite precipitation in the quartz veins also show a metamorphic signature. It is clear that both the Sn and W vein-type deposits in Rwanda have many things in common, such as the structural setting (secondary anticlines), the relative timing (during or in a late stage of a deformation event, after intrusion of the G4-granites), host rock alteration (muscovitisation, tourmalinisation and the precipitation of biotite), the mineral paragenesis (quartz veins that have been crosscut by a late sulphide phase), the ore-mineralising fluid (an aqueous-gaseous fluid with a low to moderate salinity dominantly composed of Na and Cl, a minimal temperature of formation between ~230° and ~350°C, a gas phase dominantly composed of CO2 with lesser amounts of N2 and CH4,), and the stable isotope composition of the ore minerals (points to interaction of the mineralising fluids with metamorphic rocks that contain NH4+-micas and feldspars and mature organic material). Although both the Sn and W vein-type deposits show this large amount of similarities, they do not occur together in the same location. The main difference between both deposit types is their lithological control, i.e.: the W vein-type deposits are typically hosted by units of alternating metapelitic rocks and sandstones, while the Sn vein-type deposits are hosted by thick sandstone/quartzite units. This leads to the conclusion that the host rock of the ore deposits played a controlling factor in ore deposition through water-rock interactions. Based on the similar setting, mineralogy and fluid composition it is stated that both types of ore deposits have been caused by the same mineralising event. Given that there is an overlap in ages between the G4‑granites and the W-mineralised quartz veins, and the typical association of Sn- and W-deposits with granite intrusions worldwide, the conclusion is drawn that both the Sn- and W-mineralising fluids originate from the geochemically specialised G4-granites in Rwanda. Strong water-rock interactions have subsequently overprinted this original signal and gave rise to the complex fluid chemistry as it can be seen today.