Content uploaded by Axel Gerdes
Author content
All content in this area was uploaded by Axel Gerdes on Feb 17, 2016
Content may be subject to copyright.
Analytical Fingerprint of Columbite-Tantalite (Coltan)
Mineralisation in Pegmatites – Focus on Africa
F Melcher
1
, T Graupner
2
, F Henjes-Kunst
3
, T Oberthür
4
, M Sitnikova
5
, E Gäbler
6
,
A Gerdes
7
, H Brätz
8
, D Davis
9
and S Dewaele
10
ABSTRACT
Following the United Nations initiative to fingerprint the origin of
conflict materials, the German Federal Ministry for Economic
Cooperation and Development has decided to fund a pilot study on coltan
ores. Since 2006, our working group has been investigating columbite-
tantalite (coltan) mineralisation, especially in Africa, also within the
wider framework of establishing certified trade chains.
More than 300 samples were obtained from the world’s major coltan
producing areas. Special attention is, however, directed to samples and
concentrates from Ta-Nb-Sn provinces in Africa: Democratic Republic of
the Congo, Rwanda, Mozambique, Ethiopia and Namibia. Using state of
the art analytical tools, we investigate mineralogical and chemical
parameters obtained from columbite-tantalite ores and concentrates in
order to distinguish between ore provinces, likely even down to the deposit
scale. Methods employed include fully automated electron microscopy
(mineral liberation analysis), electron microprobe analysis (major and
minor elements), laser ablation plasma-source mass spectrometry (trace
elements and U-Pb dating), X-ray fluorescence spectroscopy (bulk major
and trace elements), X-ray diffraction analysis (mineralogy and structure)
and thermal-ionisation mass-spectrometry (U-Pb dating).
Major and trace element concentration patterns, mineral assemblages
in the ore concentrates, and zoning characteristics in the different
pegmatites from Africa distinctly differ from each other. Furthermore, the
following age populations are evident:
• Archaean (>2.6 Ga);
• Palaeoproterozoic (1.9 - 2.1 Ga);
•
early Neoproterozoic (‘Kibaran’; 0.98 - 0.93 Ga); and
•
late Neoproterozoic to early Palaeozoic (ca 0.5 Ga).
Currently, we focus on the resolution of the fingerprinting system from
region via ore province down to deposit scale. Our preliminary analytical
results indicate that a certification scheme including fingerprinting of
sources of coltan ores is feasible.
INTRODUCTION
The chemical properties of tantalum are increasingly used in
various technological developments. Tantalum capacitors are
invaluable in the production of mobile phones, digital cameras,
computers and cars. Tantalum is almost exclusively mined from
rare-element pegmatites and a few specialised granites, with
minor production coming from niobium-rich carbonatites and
residues of tin smelting. Annually, approximately 1300 tonnes of
Ta metal is mined in Western Australia, Brazil, Canada, China
and numerous African countries. Future mines will eventually
open up in Egypt, Saudi Arabia, Russia, Finland and Greenland.
Today, Talison (Australia) produces in excess of 50 per cent of
the world production from its Wodgina mine. However,
increasing pressure on the world market has initiated and
renewed extensive exploration in many African countries, in the
hope that ore may be produced at a lower price. In most African
countries, tantalum is mined by artisanal miners from eluvial and
alluvial deposits: miners produce ‘coltan’, which is the Central
African trade name for concentrates containing minerals of the
columbite-tantalite solid solution series (abbreviated CGM,
columbite-group minerals). Such concentrates contain from ten
to 40 per cent Ta
2
O
5
, in addition to Nb, Sn, W, Ti, U, Th, REE,
Zr and other metals.
Coltan has been identified as one of several raw materials that
were used to finance the civil wars in central Africa. The term
‘blood coltan’ was coined in the Congolese civil wars as the sale
of this mineral powered the fighting, especially in the eastern
provinces of the Democratic Republic of the Congo (DRC). The
various armies in this war-torn region, both official and amateur,
moved in to take over the trade. A sharp price increase for
tantalum on the market at the beginning of the century caused by
speculation (from US$60 to US$480/kg Ta
2
O
5
) made this trade
highly profitable. After the ‘coltan boom’ in 2000, large
quantities of coltan were smuggled from the DRC into the
neighbouring countries to be sold illegally on the black market.
The United Nations took the initiative and an expert group
proposed that measures should be taken to certify tantalum-
bearing mineral products along their trade chain. Analytical
schemes should be worked out that allow distinguishing mineral
matter produced within regions affected by the civil war from
other sources. Results of a pilot study funded by the German
Federal Ministry for Economic Cooperation and Development
(BMZ) are presented here.
TANTALUM PRODUCTION IN AFRICA
Tantalum mineralisation has been reported from many countries
in Africa (Fetherston, 2004). Central African countries such as
the DRC, Uganda, Burundi and Rwanda have been significant
suppliers of tantalum concentrates for at least 40 years. Other
Ninth International Congress for Applied Mineralogy Brisbane, QLD, 8 - 10 September 2008 615
1. German Federal Institute for Geosciences and Natural Resources
(BGR), Stilleweg 2, Hannover 30655, Germany.
Email: F.Melcher@bgr.de
2. German Federal Institute for Geosciences and Natural Resources
(BGR), Stilleweg 2, Hannover 30655, Germany.
Email: Torsten.Graupner@bgr.de
3. German Federal Institute for Geosciences and Natural Resources
(BGR), Stilleweg 2, Hannover 30655, Germany.
Email: Friedhelm.Henjes-Kunst@bgr.de
4. German Federal Institute for Geosciences and Natural Resources
(BGR), Stilleweg 2, Hannover 30655, Germany.
Email: Thomas.Oberthuer@bgr.de
5. German Federal Institute for Geosciences and Natural Resources
(BGR), Stilleweg 2, Hannover 30655, Germany.
Email: MariaAlexandrovna.Sitnikova@bgr.de
6. German Federal Institute for Geosciences and Natural Resources
(BGR), Stilleweg 2, Hannover 30655, Germany.
Email: hans-eike.gäbler@bgr.de
7. Institute of Geosciences, Petrology and Geochemistry,
Altenhöferallee, Frankfurt am Main 60438, Germany.
Email: gerdes@em.uni-frankfurt.de
8. GeoZentrum Nordbayern, Universität Erlangen-Nürnberg,
Schlossgarten 5, Erlangen 95054, Germany.
Email: braetz@geol.uni-erlangen.de
9. Jack Satterly Geochronology Laboratory, Department of Geology,
Earth Sciences Centre, University of Toronto, 22 Russell Street,
Toronto ON M5S 3B1, Canada. Email: dond@geology.utoronto.ca
10. Department of Geology and Mineralogy, Royal Museum for Central
Africa, 13 Leuvensesteenweg, Tervuren 3080, Belgium.
Email: stijn.dewaele@africamuseum.be
HOME
tantalum producing countries include Mozambique, Nigeria,
Ethiopia, Namibia and Zimbabwe. Recently, the largest single
producers are the Kenticha mine, Ethiopia and Marropino,
Mozambique, which are both mined in a semi-industrial way. In
all other areas, tantalum mining continues as small-scale,
artisanal mining.
According to the USGS Mineral Commodity Summaries,
African countries have a share of ~20 per cent of the world
tantalum production. That production developed from low
quantities (<100 tonnes of Ta metal) – before a dramatic price
increase in 2000 (the ‘coltan boom’) – to 350 tonnes in 2000, and
has stayed above 250 tonnes per annum since then. The major
share of the production was reported from the DRC and Rwanda
in 2000 - 2001, but since then has shifted to Rwanda (2001 -
2002), Zimbabwe (2002 - 2003), and Mozambique (2003 -
2004). These ‘official’ numbers suggest significant transfer of
Congolese material into neighbouring countries (Figure 1). In
contrast, the Ethiopian production (Kenticha mine) steadily
increased its output since 1995. Mozambique did not contribute
significantly to the world market before 2007.
Almost all tantalum deposits in Africa appear to be related to
granitic rare-metal pegmatites or their regolith derivatives, such
as deeply weathered deposits, or to eluvial and alluvial placer
deposits (Varlamoff, 1972; Fetherston, 2004). At least five
periods of tantalum mineralisation are identified in Africa
(Figure 2):
1. Late Archaean to early Palaeoproterozoic pegmatites host
tantalum mineralisation on the Zimbabwe and northern
DRC-Central African Republic (DRC-CAR) cratons.
2. The Palaeoproterozoic granite-greenstone belt terranes of
the Eburnean Province in West Africa carry small rare-
metal pegmatite deposits, eg in the Ivory Coast (Allou et al,
2005) and Ghana (Kokobin near Oda).
3. A major period of rare-element granite emplacement is
manifested in the Kibaran Belt of central Africa.
Mineralisation is slightly younger than 1.0 Ga (billion
years). Identical ages are found in the Tantalite Valley
pegmatite field, Namaqualand Province of southern
Namibia and South Africa.
4. The Neoproterozoic to early Palaeozoic ‘Panafrican’
mobile belts, especially along the eastern coast of Africa,
are locally highly endowed with rare-metal granites (Egypt)
and pegmatites (Ethiopia, Mozambique, Madagascar).
Intrusion ages range from 0.45 to 0.6 Ga. Pegmatites of
similar ages are known from the Central Zone of the
Damara orogen in central Namibia, and from the
Panafrican basement in Nigeria.
5. The youngest tantalum mineralisation in Africa appears to
be related to the Jurassic granitic ring complexes of the Jos
Plateau in central Nigeria.
Areas of the pilot study
Democratic Republic of the Congo, Rwanda, Burundi
and Uganda
Tantalum-rich placer deposits have been intermittently mined in
the Ituri Province (DRC), within the ‘northern DRC-CAR
craton’. However, the Kibaran Belt (DRC, Rwanda, Burundi and
Uganda) is the major producer of Ta-Nb, Sn, W, REE (monazite)
and Au. The world’s largest reserves of tantalum are probably
contained in the Kivu Province of the eastern DRC. Mining took
place from about 1910 into the Mobutu era. The largest single
pegmatite body in the Kibaran Belt (Manono-Kitotolo; Katanga
Province, DRC) hosted total reserves of ~100 million tonnes of
eluvial/alluvial and primary ore (Bassot and Morio, 1989). With
the outbreak of the Congolese war in 1998, mining, distribution
and sale of coltan in the Kivu Province came under the control of
the Rwandan-backed rebel army, which was not withdrawn
before mid-2002 (Fetherston, 2004). Recently, mining licenses
have been granted to internationally operating mining
companies. However, the artisanal miners are out of control in
most areas, and industrial production of coltan is not likely to
restart in the very near future. In Rwanda, the state-owned Régie
d´Exploitation et de Developpement des Mines (REDEMI) still
controls many concessions, but has also granted licenses to
mining companies that produce cassiterite and coltan
concentrates in cooperation with the local artisanal miners.
Historical productions of cassiterite and coltan of Rwanda from
1958 to 2005 are ~60 000 and 5000 tonnes, respectively (BRGM,
1987; USGS Mineral Commodity Summaries 1990 - 2005).
616 Brisbane, QLD, 8 - 10 September 2008 Ninth International Congress for Applied Mineralogy
F MELCHER et al
FIG 1 - Mine production of tantalum metal from African countries.
Sources: USGS Mineral Commodity Summaries, BGR mineral
database.
FIG 2 - Location of important ore provinces with Nb-Ta mining
activities in Africa. See text for discussion of the ages. The position
of the Archaean to Palaeoproterozoic craton areas in Africa is
added for illustration (modified from Schlüter, 2006).
The NNE-SSW-oriented Kibaran Belt extends from Uganda
via Rwanda and finally into the Katanga region of the DRC (Pohl,
1994; Dewaele et al, in press). Palaeo- and Mesoproterozoic
clastic sediments are ubiquitous, and have been intruded by three
generations of granites. The oldest granites have been dated at
1.38 Ga, whereas the youngest are slightly younger than 1.0 Ga.
The Ta-Nb, Sn and W ores are exclusively connected with the
youngest post-orogenic ‘tin granites’ (Late-Kibaran), also called
G4 granites. Mineralisation is developed in quartz veins, greisens
and small pegmatitic bodies that are sometimes zoned, in some
distance to the G4 granites (Varlamoff, 1972). The rare-metal
pegmatites are of the lithium-caesium-tantalum type (LCT;
erný, 1991) and intrude metasediments, basic intrusive rocks
(metadiorites) or rarely older granites. They generally carry Sn,
and many contain Ta-Nb, in addition to locally abundant Li, Be
and phosphate mineralisation. Cassiterite is also abundant in
veins that are locally spatially related to pegmatites. However,
cassiterite-bearing veins are usually devoid of tantalum
mineralisation. The quartz-wolframite veins are usually free of
Sn and Ta.
Ethiopia
All production of Ta comes from the Kenticha pegmatite in the
Oromia Regional State, which is presently mined by the
Ethiopian Mineral Development Share Company (EMDSC). It
produces 70 tonnes Ta metal per annum (120 tonnes of
concentrate at 60 per cent Ta
2
O
5
). The probable reserve of
primary ore is 17 000 tonnes Ta
2
O
5
at a grade of 0.017 per cent
Ta
2
O
5
, whereas the reserve of Ta
2
O
5
in the weathered zone was
calculated to 2400 tonnes at 0.015 per cent Ta
2
O
5
.
The pegmatites in the Oromia Regional State intruded into a
Precambrian terrane that is dominated by granitoids and
ophiolites emplaced between 0.9 and 0.7 Ga. Collisional granitic
magmatism occurred between 0.7 and 0.55 Ma. The pegmatites
were classified as barren, beryl-columbite, complex spodumene
and albite-spodumene types (Desta, Garbarino and Valera, 1995;
Tadesse and Desta, 1996). In the zoned Kenticha LCT rare-metal
granite pegmatite, the tantalum mineralisation is hosted by
post-tectonic alaskite and associated granitic pegmatite within
fractured and partly sheared serpentinite along a regional-scale
thrust fault separating the low-grade Kenticha greenstone belt
from medium- to high-grade gneisses.
Mozambique
From the 1950s until the civil war, various rare-element
pegmatite deposits in the Alto Ligonha region have been mined
for gemstones, industrial minerals and rare metals (REE, Be,
Nb-Ta, Li, U-Th). After the civil war, mining and exploration
licences have been granted to several mining companies,
including NOVENTA. About 81 tonnes of Ta were produced
from the Marropino deposit in 2006 (Mining Journal
Supplement, 2007).
The pegmatites are part of the Zambesia Province in NE
Mozambique. The Marropino pegmatite, ca 0.48 Ga old, is a
deeply weathered, kaolinitised, zoned LCT rare-metal pegmatite
that intruded mafic gneiss and schist of the Proterozoic Morrua
Formation. The main orebody extends for 1 km in ENE-WSE
direction and is up to 80 m thick. The pegmatites at Morrua,
ca 50 km north of Marropino, are less altered. Drilling has
indicated six pegmatite bodies each >3 m thick, and extending up
to 1 km along strike (Cronwright, 2005). Tantalite (1900 tonnes
concentrate) has been mined at Morrua from 1957 to 1979,
besides spodumene, beryl and gold. At Mutala, 90 km north of
Marropino, zoned pegmatites form up to 50 m thick bodies
intruded into mica- and amphibole-schist of the Morrua
Formation. Ancient mines have been reactivated by local
garimpeiros to produce tantalite (Munhamola, Moneia).
Namibia
Lithium-beryllium and tin-columbite-tantalite occurrences are
associated with rare-metal pegmatites in the Central Zone of the
Damara orogenic belt in central Namibia. The tin pegmatites of
Uis are historically the most important Ta producers of Namibia.
Ta
2
O
5
resources are estimated at 7.2 million tonnes (at 0.05 per
cent) and 2.0 million tonnes (at 0.024 per cent; Fetherston,
2004). At present, coltan is only produced by local artisanal
miners. Tin-bearing pegmatites are concentrated in four broad
belts. The Cape Cross-Uis belt is a narrow NE-trending belt of
up to 8 km in width and 100 km in length, and contains both
zoned and unzoned cassiterite-bearing pegmatites. Most
pegmatites are unzoned and up to 50 m thick. CGM, ixiolite,
tapiolite and wodginite have been identified as Ta minerals.
Mining of the Tantalite Valley pegmatites of southernmost
Namibia took place up to the 1970s and early during this century,
but is closed at the moment. Reserves are estimated at
0.74 million tonnes at 0.043 per cent Ta
2
O
5
(Fetherston, 2004).
The pegmatites are part of the Mesoproterozoic Tantalite Valley
Basic complex, which is composed of olivine gabbro and
gneisses. Up to 1 km long and >10 m thick rare metal pegmatites
of the LCT type intruded along a shear zone 0.93 to 0.88 Ga ago
(Diehl, 1992). Pegmatites commonly show symmetrical
zonation. The major minerals are quartz, K-spar, albite and white
mica, accompanied by spodumene, lepidolite, amblygonite,
beryllium and bismuth minerals, CGM and phosphates.
THE ANALYTICAL FINGERPRINT
The focus of this study is to develop a methodological approach
that is capable of identifying the origins of tantalum ore
concentrates. There are a number of factors which have to be
taken into consideration.
1. The analytical time and effort have to be kept at a
reasonable level. The costs for the certification should not
raise the price for coltan unreasonably (eg application of a
two-step procedure).
2. The quality and composition of the coltan ore concentrates
available on the market may vary considerably depending
on the technical equipment used for ore processing and the
experience of the miners.
3. The mineralogical and chemical composition of Ta-Nb ores
is extremely complex, based on the wide range of minerals
of the columbite-tantalite solid solution series (CGM) and
the ability of CGM to incorporate a large number of
additional elements. Furthermore, coltan ores may also
contain other tantalum-bearing mineral phases, like
tapiolite (FeTa
2
O
6
), wodginite [(Mn,Sn,Fe,Ti,Li)Ta
2
O
8
],
ixiolite [(Nb,Ta,Sn,Fe,Mn,Ti)
4
O
8
], bismutotantalite [Bi(Nb,
Ta)O
4
], stibiotantalite [Sb(Nb,Ta)O
4
], minerals of the
pyrochlore group such as microlite [(Ca,Na)
2
Ta
2
O
6
(O,OH,F)], and further minerals of the complex
fergusonite, aeschynite and euxenite mineral groups.
Although confusing at first, these large variations in Ta-Nb
minerals and ores also offer chances for a scheme of
fingerprinting.
In this study we demonstrate the usefulness of a combined
mineralogical-geochemical approach to distinguish the origin of
coltan concentrates from five African countries; namely the
DRC, Rwanda, Mozambique, Ethiopia and (southern) Namibia.
This approach is based on an extensive database acquired and
compiled for samples obtained from Africa’s major coltan
producing areas. Most data have been acquired from
concentrates sold by artisanal miners, or recovered on site from
mineral concentration plants. Most samples from the DRC,
however, are from the Mineralogical Collection of the Museum
for Central Africa in Tervuren.
Ninth International Congress for Applied Mineralogy Brisbane, QLD, 8 - 10 September 2008 617
ANALYTICAL FINGERPRINT OF COLUMBITE-TANTALITE (COLTAN) MINERALISATION IN PEGMATITES
Methodological approach
Coltan concentrates are studied in a step-by-step mode (Figure 3)
subdivided into three different ‘paths’, which evolve from:
1. bulk methods to,
2. single grain, and
3. in situ methods.
In first approximation, this sequence also coincides with
increases in time and costs involved, but also with an enhanced
knowledge on the grain-scale. Which path is followed depends
on the information needed, questions asked, and also on the
analytical equipment available, if the fingerprint will be applied
by other laboratories in the future.
The first path (1; Figure 3) comprises bulk analysis of sample
powders. Major and trace element concentrations are obtained by
wavelength-dispersive-X-ray fluorescence analysis (XRF) on
bulk samples (PANAlytical Axios and Philips PW2400). XRF on
fused glass discs provides fast and cheap average concentration
data of major and most relevant trace elements, except Li, Be, B
and some of the REE. Due to the heterogeneous nature
of the concentrates (sampling approaches; preconcentration
techniques), comparison of the data is not easily possible.
However, the method provides important data on the quality of a
concentrate. The mineralogical composition of bulk samples is
determined by X-ray diffraction analysis (XRD; Philipps PW
3710).
Analyses performed following the second path (2; Figure 3)
include major and trace element analysis, including U-Pb dating,
of single grains or fragments of single grains, requiring a
minimum size of the grains used, and careful pre-examination by
scanning electron microscopy (SEM). For major and trace
element analysis by magnetic sector ICP-MS (Element 1) and
ICP-OES one or several hand-picked grains (5 to 100 mg of
sample material) are ground and dissolved in a mixture of
hydrofluoric acid 48 per cent (20 - 200 μl) and nitric acid 65 per
cent (200 μl). After complete dissolution deionised water is
added to bring the volume to 20 ml. Aliquots from this solution
are diluted by 0.15 M nitric acid and analysed by ICP-OES (Nb,
Ta, Mn, Fe, Sn) and magnetic sector ICP-MS (32 trace elements
including the REE). The dilution factors depend on the sample
weight and the applied instrument and vary between 2.5 and 50.
Analyses of CGM grains (or fragments of grains) that have been
carefully selected from concentrates provide a reasonably fast
and cheap method to chemically characterise single grains.
Weighted sample portions are smaller, and detection limits are
lower compared to XRF. However, contribution from mineral
inclusions and the effects of zoning are neglected. The method
provides quantitative major and trace element data of single
grains. However, only a limited number of grains (about five to
ten grains) from a concentrate can be analysed in a reasonable
time interval. Nevertheless, the results agree well with in situ
methods such as electron probe microanalysis and laser ablation
ICP-MS, with the exception of some more mobile trace elements
(eg Rb, LREE) that may be present in secondary phases, which
are avoided during in situ analyses.
Following path 3 (Figure 3), polished sections are prepared
and investigated by quantitative mineralogical analysis using the
mineral liberation analysis software (MLA; JK Tech Pty Ltd,
Australia) on a Quanta 600 FEG scanning electron microscope
(FEI company), equipped with an EDAX 32 module. The MLA
software combines backscattered electron (BSE) images with
EDX spectra. For the MLA a series of BSE images including an
X-ray spectrum for each mineral particle is collected. The offline
processing routine compares the measured mineral spectra with
known mineral standards to determine the mineral identity for
each grain. The MLA is a fast and accurate method for
quantitative determination of all particles in a sample, and is
particularly well suited for mineral concentrates.
CGM and other Ta-Nb-bearing mineral phases are analysed for
major and trace elements by electron microprobe (CAMECA
SX100), with detection limits (LOD) of 200 ppm for trace
elements. The advantages of wavelength-dispersive electron
microprobe analysis (EPMA) of Ta-bearing phases are the high
spatial resolution (ca 1 μm), the non-destructive nature of the
method, simple analytical procedures including standardisation
(against natural CGM and pure metallic standards), and the
possibility of automatisation. The major disadvantages are long
counting times for trace elements at reasonable LODs. In order
to collect a representative number of analyses from a tantalum
concentrate, ~100 - 150 grains are analysed; taking abundant
zoning into account, between 200 and 500 analyses are carried
out, consuming altogether ~50 to 125 hours. The results are
populations, or fields of analytical data in binary diagrams
which, in principle, represent fractionation and post-magmatic
evolution trends of CGM (Figure 4; eg erný and Ercit, 1985;
erný, 1989). Often, plots of the major element ratios XMn
(100 × Mn/(Mn+Fe)) and XTa (100 × Ta/(Ta+Nb)) enable
discrimination of different ore provinces even down to a deposit
scale. Many trace elements also follow fractionation trends with
XMn and/or XTa. Others, however, do not show coherent
behaviour and may be used to discriminate the origin of the
sample. EPMA is the only method available to date to
quantitatively resolve complex zoning patterns of CGM (eg
Lahti, 1987). Many zones are less than a few micrometres wide
(Figure 5) and cannot be measured by laser ablation inductively
coupled plasma mass spectrometry (LA-ICP-MS). Chemical
variation within zoned grains is substantial, and in some cases as
large as the overall variation of all CGM. Possibilities for
discrimination of tantalum pegmatite sources based on EPMA
thus include variations of major and trace element concentrations
(>200 ppm) in a population of grains, and in single crystals. Both
are considered viable fingerprints to their source.
For determination of low levels of trace elements the
LA-ICP-MS technique (Nd:YAG laser – 266 nm New Wave
Merchantek LUV 266x; Agilent 7500i quadrupole ICP-MS;
University of Würzburg) is applied. Thirty-seven trace elements
including the REE are determined. Argon is used as the carrier
618 Brisbane, QLD, 8 - 10 September 2008 Ninth International Congress for Applied Mineralogy
F MELCHER et al
FIG 3 - Methods used for characterisation of the mineralogical
parameters and geochemical compositions of Ta-Nb concentrates.
Abbreviations: EPMA – electron probe microanalysis; ICPMS –
inductively coupled plasma mass spectrometry; LA-ICP-MS –
laser ablation ICPMS; MLA – mineral liberation analysis; SEM –
scanning electron microscopy; TIMS – thermal ionisation mass
spectrometry; XRD – X-ray diffractometry; XRF – X-ray
fluorescence spectrometry. Analytical paths 1 to 3 are discussed
in the text.
gas. The spot size varies from 30 to 50 μm. The glass reference
materials NIST SRM 610 and 612 with the values of Pearce et al
(1997) are used for external calibration and calculation of trace
elements by the GLITTER Version 3.0 (Macquarie Research Ltd,
2000). Advantages of the LA-ICPMS method are the
significantly lower detection limits for trace elements (maximum
LOD values are mostly ≤1 ppm; higher maximum values for Mg,
Al, Si, Ca, Ti, As and Sn) compared to the analysis by EPMA.
Disadvantages include the destructive nature of the method and
its lower spatial resolution compared to EPMA. Possible
contamination of the analysis by micromineral inclusions ablated
at depth during single spot analysis is generally easily recognised
and can be eliminated in most cases.
Uranium-lead dating is carried out both on crystal fragments
using conventional thermal ion mass spectrometry (TIMS; BGR
and University of Toronto), and in situ using a Thermo-Scientific
Element II sector field ICP-MS coupled to a New Wave UP213
ultraviolet laser system with low-volume ablation cell (University
of Frankfurt). Spot size varies from 30 to 60 μm. Raw data are
corrected for background signal, common Pb, laser induced
elemental fractionation, instrumental mass discrimination, and
time-dependant elemental fractionation (Gerdes and Zeh, 2006).
Ninth International Congress for Applied Mineralogy Brisbane, QLD, 8 - 10 September 2008 619
ANALYTICAL FINGERPRINT OF COLUMBITE-TANTALITE (COLTAN) MINERALISATION IN PEGMATITES
FIG 5 - Backscatter electron images (CAMECA SX 100 electron microprobe) of coltan concentrates from (A) Gatumba, Rwanda;
(B) Ruhanga, Rwanda; (C) Nyarigamba, Rwanda; (D) Nyambisindu, Rwanda; (E) Yubili mine, Kivu Province, DRC; (F) Manono,
Katanga, DRC. Abbreviations: FeC, ferrocolumbite; FeT, ferrotantalite; MnC, manganocolumbite; MnT, manganotantalite; Mc, microlite;
Umc, Uranmicrolite; Wg, wodginite; Wolf, wolframite.
FIG 4 - General fractionation trends of CGM in the columbite
quadrilateral (after erný, 1989).
The analytical reproducibility (eg GJ-1 reference zircon) of the
206
Pb/
238
U and
207
Pb/
206
Pb is commonly about 0.8 and 0.5 per
cent, respectively. No matrix dependent U/Pb fraction has been
observed.
The chemical procedures to separate U and Pb for TIMS
analysis are adapted from Romer and Wright (1992) and Romer
and Smeds (1994). U and Pb are measured in multicollector
mode on a MM354 TIMS (University of Toronto) and a
ThermoFinnigan Triton (BGR). Isoplot (Ludwig, 2003) is used
for graphical presentation of U-Pb isotope data and age
calculation. U-Pb dating can yield highly precise and concordant
mineral ages provided CGM grains free of inclusions or
alteration phenomena are selected. However, this is nearly
impossible due to the opaque nature of the CGM. Thus, U-Pb
dating of CGM often yields complex and discordant ages. To
minimise the effects of disturbance of the U-Pb isotopic system
in CGM, strong HF leaching of the grains prior to dissolution has
been proposed (Romer and Wright, 1992). At the University of
Toronto, individual fragments (<100 μm; <1 mg) not leached in
HF are analysed separately (‘single-grain method’). This allows
careful inspection of the quality of the fragments to be dated.
However, variations in age within a concentrate cannot be
detected by this method. At the BGR, fragments from different
grains (fragment size 0.8 - 0.16 mm; mass 1 - 4 mg) are
investigated (‘multigrain method’). The fragments are leached in
several steps using warm diluted HF, HCl and HNO
3
prior to
dissolution. Both approaches of U-Pb TIMS dating are time
consuming because many steps of quality controlling
pre-analytical preparation, chemical dissolution, chemical U-Pb
separation and mass-spectrometric U-Pb isotopic measurements
are required. In addition, chemical processing of the grains and
subsequent handling of the U-Pb element fractions have to be
performed under clean-air conditions. On the other hand, TIMS
dating is the state-of-the-art method to obtain precise U-Pb dates
and useful for discrimination of sources with small (< 10 Ma)
differences in age as are present in the Kibaran province of
central Africa.
RESULTS AND DISCUSSION
Tantalum-enriched pegmatites are found in the intermediate to
outermost parts of zoned pegmatite aureoles surrounding
parental granites (erný, 1989). Granitic pegmatites are grouped
into five classes, namely the:
1. abyssal;
2. muscovite;
3. muscovite-rare element;
4. rare-element; and
5. miarolitic classes (erný et al, 2005; Ercit, 2005).
The LCT (Li, Cs, Ta) and NYF (Nb, Y, F) petrogenetic
families within the rare-element class are important hosts to
Ta-Nb mineralisation. Based on their mineral parageneses,
rare-element pegmatites may be further subdivided into five
pegmatite types: the rare earth, beryl, complex (with spodumene,
petalite, amblygonite, lepidolite and elbaite subtypes), albite-
spodumene and albite types (erný et al, 2005).
No matter which classification is adhered to, the chemical
composition of CGM, and other Ta-bearing phases
systematically varies according to the pegmatite type (Figure 4).
In beryl pegmatites, CGM are Fe-Nb dominated, mostly ranging
from ferrocolumbite to ferrotantalite, whereas in highly
fractionated pegmatites of the complex type, CGM are
commonly Mn- and Ta-rich and display fractionation trends from
manganocolumbite to manganotantalite. Large compositional
variations within distinct pegmatites are rather common.
Up to date, a vast amount of electron microprobe data on
CGM has accumulated in the literature. The XMn (100 ×
Mn/(Mn+Fe), atomic ratio) versus XTa (100 × Ta/(Ta+Nb),
atomic ratio) diagram identifies a pegmatite type or subtype (eg,
Breaks, Selway and Tindle, 2005), and discrimination of
different pegmatites might be achieved in cases as well. Most
data sets contain concentrations of Ti, Sn, rarely of W, Zr, U, Sc,
Y, Mg and other elements substituting into the CGM structure.
erný et al (2007) investigated the role of Zr and Hf in CGM and
wodginite. The trace element composition of CGM in the ppm
range has not been investigated previously.
The possibility to use major and trace element compositions to
fingerprint the origin of Ta concentrates from unknown locations
has not been taken into account. During the Congolese wars,
problems arose with coltan that was illegally mined from the
Kahuzi-Biega National Park in the DRC, which is a major habitat
of lowland gorillas. Poirier and Lastra (2002) postulated that
coltan concentrates derived from the National Park may be
distinguished from coltan concentrates from other parts of the
DRC and of Rwanda based on their mineralogical composition
(ie the presence or absence of cassiterite) and microprobe major
element data (XMn versus XTa diagram). However, this study
was based on seven CGM-bearing samples only and, therefore,
does not fulfil any statistical criteria. In the following, we will
demonstrate that major element chemistry of CGM and
mineralogical composition of concentrates in most cases does
not unequivocally discriminate between different mining areas
within a given region of interest. This will be demonstrated using
a sample set comprising 151 samples, including 86 concentrate
samples, from pegmatite deposits in the DRC, Rwanda, Ethiopia,
Mozambique and southern Namibia (Tantalite Valley Complex).
Data are grouped according to their country of origin: this is a
prerequisite for the fingerprint of minerals from conflict areas.
The fields defined by the major element compositions of the
CGM overlap to a large extent (Figure 6a). It is hardly possible
to prove or disprove the origin of a group of CGM grains,
especially when samples plot within an area of XMn ratios
between 40 and 100 and XTa ratios between ten and 70. The
presence of ferrotapiolite discriminates concentrates from the
DRC and Rwanda, because this phase has not been identified so
far in samples from the other countries. Fe-rich ferrocolumbite to
ferrotantalite (XMn <40) is also restricted to the DRC and
Rwanda.
Better discrimination of overlapping fields is achieved using
minor elements such as W, Sn, Ti, Zr, Hf and U that are routinely
measured by EPMA if concentrations exceed 200 ppm.
Although, in many cases, these elements will be present in lower
concentrations, they may be used for further discrimination
(Figure 6b). Tungsten concentrations are highest in CGM from
the DRC (up to four weight per cent WO
3
), whereas CGM from
other countries rarely exceed 0.6 weight per cent WO
3
.
Populations from Mozambique and Ethiopia will follow different
fractionation trends, especially in the XTa versus WO
3
diagram.
The manganotantalite from southern Namibia plots in a distinct
field to rather high WO
3
concentrations. The XMn and XTa
versus WO
3
diagrams will allow identification of concentrates
from the DRC only at high W concentrations. Low tungsten
concentrations, which are also present in some DRC samples
within the data set, will not be unequivocally discriminated using
this approach.
As a next step, a quantitative measurement of the
mineralogical composition is used to further discriminate
samples that plot at WO
3
concentrations <0.6 weight per cent.
Using the presence of minor phases such as tapiolite, cassiterite
or bismutotantalite discriminates concentrates from:
•
Mozambique (bismutotantalite present, cassiterite and
tapiolite absent);
•
Ethiopia and southern Namibia (bismutotantalite, cassiterite
and tapiolite largely absent); and
620 Brisbane, QLD, 8 - 10 September 2008 Ninth International Congress for Applied Mineralogy
F MELCHER et al
•
DRC and Rwanda (bismutotantalite absent, cassiterite and
tapiolite commonly present) (Table 1).
In case 2, age determination will unequivocally discriminate
Ethiopia (ca 530 Ma; Küster et al, 2007) from southern Namibia
(ca 980 Ma; Melcher et al, 2008). In case 3, age determination
will not be able to discriminate concentrates originating from
Kibaran-age (ca 930 - 1000 Ma) pegmatites of the DRC and
Rwanda (Figures 7c and 7d). However, CGM originating from
the northern Congo craton (Ituri Province, DRC) will be easily
depicted due to their significantly older ages (2000 - 2500 Ma;
Figure 7b).
Manganotantalites with moderate to low tungsten
concentrations from concentrates lacking cassiterite, tapiolite
and bismutotantalite may originate from southern Namibia,
the DRC, Rwanda or Mozambique. Mozambique will yield
Ninth International Congress for Applied Mineralogy Brisbane, QLD, 8 - 10 September 2008 621
ANALYTICAL FINGERPRINT OF COLUMBITE-TANTALITE (COLTAN) MINERALISATION IN PEGMATITES
A
B
FIG 6 - Electron microprobe analyses of CGM from African coltan
areas. Individual point analyses are grouped according to their
countries of origin. Numbers of analyses used for constructing the
fields are as follows: DRC (Kibaran occurrences only) 1383,
Rwanda 1962, Ethiopia 473, Mozambique 585, Namibia 538. Note
that both columbite-tantalite and ferrotapiolite data are plotted.
(A) XMn versus XTa diagram; (B) XMn and XTa versus WO
3
(wt per cent); note different scaling on y-axis. XMn = 100 ×
Mn/(Mn+Fe), XTa = 100 × Ta/(Ta+Nb), both in atomic per cent.
A
B
C
D
FIG 7 - U-Pb isochron diagrams for coltan concentrates from (A)
Marropino, Mozambique; (B) Mambasa, Ituri Province, DRC; (C)
Ruhanga, Rwanda; (D) Shabunda, Kivu Province, DRC;
LA-ICPMS data; each ellipse corresponds to one CGM grain.
a younger age (eg 450 - 500 Ma; Figure 7a), whereas the
remaining locations will give similar ages (930 - 1000 Ma). In
this case, trace element diagrams will help unravel their origin.
We find rare earth element concentrations and patterns to be very
useful to discriminate Kibaran samples from nearly all other
coltan pegmatite areas. Kibaran samples usually lack Eu
anomalies, or have small negative Eu anomalies (Eu*, calculated
as (0.5 × (Sm
N
+Gd
N
))/Eu
N
), 0.5 to 10) and low to moderate
MREE
N
/HREE
N
ratios (0.1 - 10). Manganotantalite from
southern Namibia is characterised by strong negative Eu
anomalies (Eu* > 20) and high MREE
N
/HREE
N
ratios (>10).
Using a trace element such as Bi, fields might be constructed that
overlap only to a small degree (Figure 8).
Having excluded southern Namibia as the origin of a
concentrate, we still have to resolve a possible origin from the
DRC and Rwanda. Analysis of many concentrates from different
areas within the Kibaran Province has yielded amazing
differences of major and trace element compositions. In most
cases, microprobe data of a large number of CGM in a given
concentrate will define a field at low, intermediate or high XMn
values, with zoning trends in XTa in individual crystals (Figure 9).
These reflect the general fractionation trends of beryl pegmatites
(low XMn), spodumene pegmatites and various complex
rare-element pegmatite groups (intermediate to high XMn)
(erný, 1989). Within such groups, further discrimination is
possible using trace element ratios (Figure 10). This last step will
finally enable to discriminate the origin of a sample even down
to the deposit scale.
A schematic flow chart illustrates the general options to
distinguish coltan concentrates from ore provinces of similar
geological age (Figure 11), arranged in a step-by-step mode from
modal mineralogy to major and trace element chemistry of
CGM.
CONCLUSIONS
Without doubt, there are regional and local variations in the
composition of coltan. These are due to differences in geological
age and mineralogical and chemical composition of host
pegmatites and their derivative heavy mineral concentrates.
Zoned CGM crystals perfectly mimic the chemical evolution of
pegmatitic melts and can thus be used as monitors of the
fractionation stage of the source rocks. This allows distinction of
locations even in districts and provinces of similar geological
ages, similar host rocks or similar parental melt compositions.
Each tantalum deposit has its unique characteristics. Therefore, a
fingerprint of samples of suspect or unknown origin should be
possible when a large and high-quality analytical database is
available.
However, it takes appreciable analytical efforts and time to
completely characterise a concentrate. In the future, methods will
be developed that allow fast screening based on modal
mineralogy and trace element and/or isotope geochemistry.
622 Brisbane, QLD, 8 - 10 September 2008 Ninth International Congress for Applied Mineralogy
F MELCHER et al
Critical
concentration
(%)
DR
Congo/
Kibaran
Mean DR Congo/
Eburnean
Mean Rwanda Mean Mozambique Mean Ethiopia Mean
Ferrocolumbite 5.0 xx 19.70 xx 4.82 xx 9.15 0.06 0.22
Ferrotantalite 5.0 x 5.33 xxx 26.80 xx 10.90 0.14 0.10
Manganocolumbite 5.0 xxx 18.30 xxx 20.00 xxx 15.30 x 0.97 xx 9.84
Manganotantalite 5.0 xxx 20.90 xxx 28.00 xxx 18.70 xxx 17.50 xxx 67.80
Ferrotapiolite 1.0 x 4.81 xxx 11.40 (x) 2.77 0.06 <0.01
Microlite 0.5 0.04 0.03 (x) 2.95 xxx 8.68
Uranmicrolite 0.2 0.02 0.01 (x) 0.87 xxx 0.61 x 1.22
Wodginite 0.5 (x) 1.64 <0.01 x 2.00 0.06 0.17
Bismutotantalite 0.2 <0.01 <0.01 <0.01 xx 0.55 <0.01
Cassiterite 1.0 xx 7.36 <0.01 xxx 16.80 <0.01 <0.01
Haematite/magnetite 1.0 x 6.09 <0.01 x 4.69 xx 8.46 x 6.19
Ilmenite 1.0 xx 4.53 xx 0.90 0.89 x 5.25 0.68
Monazite 0.2 x 0.27 <0.01 0.13 xx 6.56 0.03
Rutile 0.5 0.43 xx 5.46 0.44 0.30 0.05
Zircon 0.2 0.27 0.05 xx 1.56 xx 0.44 x 0.23
Number of samples 20 2 30 7 5
Legend: Fraction of samples with higher concentrations of a mineral than the defined critical concentration for the respective mineral phase: xxx – ~75 -
100 per cent; xx – about 50 - 75 per cent; x – about 25 - 50 per cent; (x) – individual samples with high concentrations of the mineral occur.
TABLE 1
Frequencies of minerals in coltan ore concentrates from pegmatite provinces in Africa as determined by MLA.
FIG 8 - Diagram of Bi concentration (ppm) versus Eu anomaly of
CGM from the central African Kibaran Province, the Tantalite
Valley Complex in southern Namibia and Panafrican pegmatites
from Mozambique.
Minerals are traded on an open global market. However, the
public is increasingly aware of ‘clean’ products, which are mined
in an environmentally sound and socially tolerable way in
countries which accept the rules of good governance. This is true
especially for minerals imported from conflict areas. The
worldwide implementation and acceptance of the ‘Kimberley
Process’ for diamonds proves that the international community is
no longer willing to accept materials from conflict areas or
materials produced under criminal circumstances. The analytical
fingerprint of ‘coltan’ may assist in the establishment of a control
instrument in an envisaged certification of the production and
trade chain of coltan.
ACKNOWLEDGEMENTS
Samples were kindly provided by a number of museums,
companies and individuals. Field work was supported by Jean
Ruzindana Munana (Redemi, Rwanda) and the Geological
Surveys of Namibia, Ethiopia and Mozambique. We also
appreciate cooperation with the project leaders, W Pohl
(Braunschweig) and B Lehmann (Clausthal), of the ‘Coltan
Environmental Management Pilot Project 2007’, funded by the
German Volkswagenstiftung. Thanks are due to many people in
the BGR for laboratory assistance, especially to J Lodziak,
P Rendschmidt, F Korte and M Bockrath.
Ninth International Congress for Applied Mineralogy Brisbane, QLD, 8 - 10 September 2008 623
ANALYTICAL FINGERPRINT OF COLUMBITE-TANTALITE (COLTAN) MINERALISATION IN PEGMATITES
FIG 10 - Zr+W (ppm) versus Ti/Sn for CGM from the central African
Kibaran Province, discriminating different geographical regions.
FIG 11 - Flow chart exemplifying the discrimination of tantalum concentrates (0.9 - 1.0 Ga old) from central and southern Africa. CGM =
columbite-group minerals; conc = concentrate.
FIG 9 - Variation of XMn and XTa in CGM and tapiolite in four coltan concentrates from Rwanda and the DRC.
REFERENCES
Allou, B A, Lu, H Z, Guha, J, Carignan, J, Naho, J, Pothin, K and Yobou,
R, 2005. Une corrélation génetique entre les roches granitiques, et les
dépôts éluvionnaires, colluvionnaires et alluvionnaires de colombo-
tantalite d’Issia, centre-ouest de la Côte d’Ivoire, Exploration and
Mining Geology, 14(4):61-77.
Bassot, J-P and Morio, M, 1989. Morphologie et mise en place de la
pegmatite kibarienne à Sn, Nb, Ta, Li de Manono (Zaïre), Chronique
de la Recherche Miniere, 496:41-56.
Breaks, F W, Selway, J B and Tindle, A G, 2005. Fertile peraluminous
granites and related rare-element pegmatites, Superior Province of
Ontario, in Rare-Element Geochemistry and Mineral Deposits (ed:
R L Linnen and I M Samson), pp 127-158, GAC Short Course Notes
17, St Catherines.
BRGM, 1987. Plan Minéral du Rwanda, 437 p (BRGM).
erný, P, 1989. Characteristics of pegmatite deposits of tantalum, in
Lanthanides, Tantalum and Niobium (ed: P Möller, P erný and
F Saupé), pp 195-239 (Springer-Verlag: Berlin).
erný, P, 1991. Rare-element granitic pegmatites. Part I: Anatomy and
internal evolution of pegmatite deposits, Geoscience Canada,
18:49-67.
erný, P, Blevin, M, Cuney, M and London, D, 2005. Granite-related ore
deposits, Economic Geology 100th Anniversary Volume, pp 337-370.
erný, P and Ercit, S, 1985. Some recent advances in the mineralogy and
geochemistry of Nb and Ta in rare-element granitic pegmatites,
Bulletin de Minéralogie, 108:499-532.
erný, P, Ercit, T S, Smeds, S A, Groat, L A and Chapman, R, 2007.
Zirconium and hafnium in minerals of the columbite and wodginite
groups from granitic pegmatites, Canadian Mineralogist, 45:185-202.
Cronwright, M S, 2005. A review of the rare-element pegmatites of the
Alto Ligonha Pegmatite Province, northern Mozambique and
exploration guidelines, MSc thesis (unpublished), Rhodes University,
Grahamstown.
Desta, Z, Garbarino, C and Valera R, 1995. Granite pegmatite system in
Kenticha (Adola, Sidamo, Ethiopia) rare metal pegmatite belt:
Petrochemistry, regional pegmatite zoning and classification, Sinet:
An Ethiopian Journal of Science, 18:119-157.
Dewaele, S, Tack, L, Fernandez-Alonso, Boyce, A, Muchez, P, Schneider,
J, Cooper, G and Wheeler, K, in press. Geology and mineralization
of the Gatumba area, Rwanda (Central Africa): Present state of
knowledge, Etudes Rwandaises.
Diehl, B J M, 1992. Niobium and tantalum, in The Mineral Resources of
Namibia, pp 3.5-1-3.5-15 (Ministry of Mines and Energy, Geological
Survey: Windhoek).
Ercit, T S, 2005. REE-enriched granitic pegmatites, in Rare-Element
Geochemistry and Mineral Deposits (ed: R L Linnen and I M
Samson), pp 175-199, GAC Short Course Notes 17, St Catherines.
Fetherston, J M, 2004. Tantalum in Western Australia, Mineral Resources
Bulletin, 22, 162 p (Western Australia Geological Survey).
Gerdes, A and Zeh, A, 2006. Combined U-Pb and Hf isotope
LA-(MC)-ICP-MS analyses of detrital zircons: Comparison with
SHRIMP and new constraints for the provenance and age of an
Armorican metasediment in Central Germany, Earth and Planetary
Sciences Letters, 249:47-61.
Küster, D, Romer, R L, Tolessa, D, Zerihun, D and Bheemalingeswara,
K, 2007. Geochemical evolution and age of the Kenticha tantalum
pegmatite, southern Ethiopia, in Granitic Pegmatites: The State of the
Art, International Symposium, Porto.
Lahti, S I, 1987. Zoning in columbite-tantalite crystals from the granitic
pegmatites of the Eräjärvi area, southern Finland, Geochimica et
Cosmochimica Acta, 51:509-517.
Ludwig, K R, 2003. User’s Manual for Isoplot 3.00: A Geochronological
Toolkit for Excel, Berkely Geochronological Center special
publication 4, 71 p.
Melcher, F, Sitnikova, M A, Graupner, T, Martin, N, Oberthür, T,
Henjes-Kunst, F, Gäbler, E, Gerdes, A, Brätz, H, Davis, D W and
Dewaele, S, 2008. Fingerprinting of conflict minerals: Columbite-
tantalite (‘coltan’) ores, SGA News, 23:1-14.
Mining Journal, 2007.
Tantalum,
12 p, Mining Journal Special
Publication.
Pearce, N J G, Perkins, W T, Westgate, J A, Gorton, M P, Jackson, S E,
Neal, C R and Chenery, S P, 1997. A compilation of new and
published major and trace element data for NIST SRM 610 and
NIST SRM 612 glass reference materials, Geostandard Newsletters,
21:115-144.
Pohl, W, 1994. Metallogeny of the northeastern Kibara belt, Central
Africa – Recent perspectives, Mineralium Deposita, 9:105-130.
Poirier, G and Lastra, F, 2002. Mineralogical study of columbite- tantalite
ores from Kahuzi-Biega National Park (Democratic Republic of
Congo) and Rwanda, CANMET MMSL report 02-022(IR).
Romer, R L and Smeds, S A, 1994. Implications of U-Pb ages of
columbite-tantalites from granitic pegmatites for the Paleoproterozoic
accretion of 1.90 - 1.85 Ga magmatic arcs to the Baltic Shield,
Precambrian Research, 67:141-158.
Romer, R L and Wright, J E, 1992. U-Pb dating of columbites: A
geochronologic tool to date magmatism and ore deposits,
Geochimica et Cosmochimica Acta, 56:2137-2142.
Schlüter, T, 2006. Geological Atlas of Africa: With notes on stratigraphy,
tectonics, economic geology, geohazards and geosites of each
country, 272 p (Springer: Berlin, Heidelberg).
Tadesse, S and Desta, Z, 1996. Composition, fractionation trend and
zoning accretion of the columbo-tantalite group of minerals in the
Kenticha rare metal field (Adola, Southern Ethiopia), Journal of
African Earth Sciences, 23:411-431.
Varlamoff, N, 1972. Central and West African rare-metal granitic
pegmatites, related aplites, quartz veins and mineral deposits,
Mineralium Deposita, 7:202-216.
624 Brisbane, QLD, 8 - 10 September 2008 Ninth International Congress for Applied Mineralogy
F MELCHER et al
