< Back to previous page
Project
Water-rock interaction and the origin of the metals in the Central African Copperblet.
The Central African Copperbelt is an exceptionally rich metallogenetic province in the
Democratic Republic of Congo (DRC) and Zambia. Decades of research yielded a
vast amount of knowledge concerning the formation of the various ore deposits.
However, the source of the ore metals is still not well understood. Many
authors refer in a rather vaguemanner to the igneous or metamorphic basement
underlying the deposits as the metal source. This research therefore focuses on
the origin of the ore metals in the Copperbelt deposits.
A promising way of relating an ore deposit with its metal source, is to compare the Sr and
Nd isotopic composition of potential source rocks, and the gangue minerals that
are associated with the ore minerals. This concept has been systematically
applied to stratiform Cu-Co deposits in the DRC (i.e. Kamoto, Kambove West, and
Luiswishi) and Zambia (i.e. Nkana and Konkola). The role of Co in these deposits
is crucial with respect to the metal sources and is discussed in detail. Next
to the stratiform deposits, two vein-type Cu-Zn-Pb mineralisations in the DRC
(i.e. Dikulushi and Kipushi) were also used as case studies
The study has been designed to solve three research questions. The first question is
whether the metal sources for the case study deposits can be constrainedusing
the isotopic signature approach. The second question asks whatthe differences
are between the stratiform Cu-Co deposits in the DRCand Zambia, and how they
relate to their sources. The last question addresses the possible application
of ore sulphide trace element compositions in the search of the ore metal sources.
The analytical equipment used to determine the major and minor element composition
ofbasement samples was an Inductively Coupled Plasma Optical Emission
Spectrometer (ICP-OES). The Sr and Nd isotopic signature of these samples, and
of gangue carbonates from the different case studies, was characterised with a
Multi-Collector Inductively Coupled Plasma Mass Spectrometer (MC-ICP-MS). The
ore sulphide element composition was measured with an Electron Probe
Microanalyser (EPMA).
For the stratiform Cu-Co deposits in the Katanga Copperbelt (i.e. the Congolese part of
the Central African Copperbelt), gangue carbonates of the diagenetic
mineralisation phase display variable Sr and Nd isotopic compositions. The low
number of samples available and their variable isotopic signatures make it
difficult to relate them to a specific basement unit. In addition, the basement
directly below the Katanga Copperbelt isunknown. However, the isotopic
composition of the diagenetic gangue carbonates generally corresponds with the
signature of felsic basement rocks. Nevertheless, the high Co content of these
deposits is most likely derived from mafic basement rocks. Syn-orogenic
mineralisationin Katanga resulted from remobilisation of diagenetic sulphides.
This caused homogeneous radiogenic isotope ratios and an ore mineral assemblage
similar to the diagenetic phase. A small new metal contribution from felsic
basement rocks might be indicated by the isotopic composition of some samples
from the Kamoto deposit. This contribution is likely still detectable because
of the relatively low metamorphic gradethat was reached at Kamoto.
The pre- to syn-kinematic mineralisation phase at the Zambian stratiform Cu-Co deposits
contains gangue carbonates with isotopic compositions that correspond with
felsic rocks of the Domes Region. However, the isotopic signatures of the Nkana
and Konkola deposits differ slightly, indicating that the ore metals were
leached from Domes Region basement rocks occurring close to each deposit. The
gangue minerals of the successive syn-kinematic mineralisation phase have
isotopic compositions resembling the earlier mineralisation phase at each
deposit. Remobilisation of the precursor ore minerals most likely occurred
during the syn-kinematic mineralisation. Because of the appearance of Co and Ni
sulphides in the late-kinematicphase at Nkana, a new source probably provided
these metals to the ore forming fluid. Nevertheless, the isotopic signatures
from gangue carbonates associated with the late-kinematic phase do not reflect
this new source and are similar to the gangue signatures of the earlier phases.
The gangue carbonates of the Cu-Co deposits in the DRC have clearly different isotopic
signatures than in Zambia. In the DRC, 87Sr/86Sr ratios
of the gangue minerals are lower and the eNd values arehigher. Two other major
differences between the Katanga and Zambian Copperbelts are the higher Co
concentrations in Katangan deposits, and the carbonate-dominated host rocks in
the DRC versus the mainly siliciclastic host rocks in Zambia. These
dissimilarities are in line with each other and strongly suggest that the
source rocks for the Katanga deposits included an important mafic component.
Unfortunately, the scarse information about the basement in Katanga does not
allow to indicate a specific source rock. If the isotopic compositions of
gangue carbonates of all stratiform deposits in the DRC are considered
together, they appear to lie on a vertical trend in the Sr-eNd isotopic space.
The endmembers of this trend may represent on one side felsic rocks similar to
the felsic rocks of the Domes Region, and on the otherside unknown mafic rocks
situated in the Katanga subsurface. In analogy to the local sources for the
Nkana and Konkola deposits, local basement rocks (including mafic rocks) are
the most plausible metal source for the Katanga deposits.
The other important difference concerns the host rock type. The siliciclastics in Zambia
could have influenced the isotopic composition of the gangue carbonates of the
Nkanaand Konkola deposits. This is especially so for the Rb-Sr isotopic
system, which is prone to remobilisation and alteration. However, it is argued
that it would not change the interpretation of the Domes Regionfelsic rocks as
metal source for these deposits. Firstly, previous research has shown that the
sedimentary Katanga Supergroup rocks, whichhost the ore bodies, did not
contain the required metal budget to bea viable metal source for the known
deposits. Secondly, the metasedimentary rocks were derived from the underlying
Domes Region basement (i.e. the interpreted metal source), which would result
in similar isotopic compositions for the host rocks as for this basement.
Hence, even if the siliciclastic rocks influenced the isotopic composition of
the mineralising fluid and gangue carbonates of the Zambian deposits, the main
metal source was the local basement.
The difference inCo concentrations and related isotopic signatures between the
Katanga and Zambian stratiform deposits is clear. However, a similar but more
subtle effect is recognised for the Nkana and Konkola deposits in theZambian
Copperbelt. The Nkana deposit contains more Co and its gangue minerals have
higher eNd values and lower 87Sr/86Sr ratios than the
Konkola deposit. Mixing calculations indicate that the metal source for Nkana
must have contained at least 5% more mafic rocks than the metal source for the
Konkola deposit. Co is absent in the stratiform Cu deposits at the
north-eastern side of the Kafue Anticline. Hence, the mafic rocks that could
have provided Co to the Zambian deposits probably occur at the south-western side
of the Kafue Anticline (i.e. close to case study deposits).
For the vein-type Cu-Ag and Cu-Zn deposits of Dikulushi and Kipushi (DRC),
respectively, very high eNd signatures were found. This concurs with the
presence of mafic rocks at each mineralisation site, i.e. gabbros within the
Axial Breccia of the Kipushi anticline, and basalts in a breccia observed in
the Dikulushi mine. Nevertheless, the required mafic rocks could equally reside
in the deep subsurface. The gangue minerals of the syn-orogenic mineralisation
at Dikulushi lie on a trend between the isotopic compositions of the Gombela
Subgroup host rocks and a required mafic component. Mixing calculations demonstrate
that a mixture of both metal sources could easily have yielded the isotopic
compositions of the gangue minerals. In contrast with the very 'mafic' isotopic
signature of the Kipushi gangue carbonates, previous research on fluid
inclusions has suggested a felsic source rock. The possible influence of felsic
basement rocks on the isotopic composition of the gangue carbonates has been
assessed. If all measured variation in the eNd values is explained in terms of
more or less influence of felsic rocks, about 10% more felsic rock contribution
is needed to explain the lowest eNd signature compared to the highest.
Nevertheless, mafic rocks must have been the dominant rock type determining the
isotopic signatures at the Kipushi deposit.
Microprobe analysis of the ore sulphides has shown that the intracrystalline trace element
variability is about as largeas the intercrystalline variability for the case
study deposits. This hampers the applicability of this technique to
characterise sulphidesfrom different mineralisation phases. In addition, few
systematic differences were found between the case studies, which could have
been expected for the typical 'mafic' elements Co, Cr or Ni. However,
chalcopyrites from the late-kinematic mineralisation phase at Nkana have
slightly higher concentrations of Co than chalcopyrites from the two earlier
mineralisation phases at Nkana. This is in agreement with the input of 'mafic'
metals during this mineralisation phase. Moreover, the Ge concentration was
markably low for sulphides from the Nkana deposit, whereas the highest Ge and
Ag levels were found for samples from the Dikulushi deposit.
More in general, the results of this research show that the Sr and Nd isotopic
signatures of gangue carbonates are a valuable tool for constraining the metal
source(s) ofsulphide ore deposits. The right basement unit and source rock
type can be determined by comparing the gangue mineral signatures with the
isotopic composition of basement rocks. For the Central African Copperbelt, the
local basement beneath or surrounding the different ore deposits has shown to
be the most likely metal source to these deposits. With regard to the role of
Co, the Co-rich deposits did also display characteristic isotopic signatures,
and strongly support the existence of (still unknown) mafic bodies in the
basement that provided the Co to the mineralising fluids. For subsequent
mineralisation phases, usually related to the Lufilian orogenesis, the
importance of remobilisation of ore metals is emphasised. The next step forward
in this research domain will be the detailed charactarisation of the local
basement beneath the ore deposits, in order to relate the geochemical signature
of each ore deposit with specific rock units.
Democratic Republic of Congo (DRC) and Zambia. Decades of research yielded a
vast amount of knowledge concerning the formation of the various ore deposits.
However, the source of the ore metals is still not well understood. Many
authors refer in a rather vaguemanner to the igneous or metamorphic basement
underlying the deposits as the metal source. This research therefore focuses on
the origin of the ore metals in the Copperbelt deposits.
A promising way of relating an ore deposit with its metal source, is to compare the Sr and
Nd isotopic composition of potential source rocks, and the gangue minerals that
are associated with the ore minerals. This concept has been systematically
applied to stratiform Cu-Co deposits in the DRC (i.e. Kamoto, Kambove West, and
Luiswishi) and Zambia (i.e. Nkana and Konkola). The role of Co in these deposits
is crucial with respect to the metal sources and is discussed in detail. Next
to the stratiform deposits, two vein-type Cu-Zn-Pb mineralisations in the DRC
(i.e. Dikulushi and Kipushi) were also used as case studies
The study has been designed to solve three research questions. The first question is
whether the metal sources for the case study deposits can be constrainedusing
the isotopic signature approach. The second question asks whatthe differences
are between the stratiform Cu-Co deposits in the DRCand Zambia, and how they
relate to their sources. The last question addresses the possible application
of ore sulphide trace element compositions in the search of the ore metal sources.
The analytical equipment used to determine the major and minor element composition
ofbasement samples was an Inductively Coupled Plasma Optical Emission
Spectrometer (ICP-OES). The Sr and Nd isotopic signature of these samples, and
of gangue carbonates from the different case studies, was characterised with a
Multi-Collector Inductively Coupled Plasma Mass Spectrometer (MC-ICP-MS). The
ore sulphide element composition was measured with an Electron Probe
Microanalyser (EPMA).
For the stratiform Cu-Co deposits in the Katanga Copperbelt (i.e. the Congolese part of
the Central African Copperbelt), gangue carbonates of the diagenetic
mineralisation phase display variable Sr and Nd isotopic compositions. The low
number of samples available and their variable isotopic signatures make it
difficult to relate them to a specific basement unit. In addition, the basement
directly below the Katanga Copperbelt isunknown. However, the isotopic
composition of the diagenetic gangue carbonates generally corresponds with the
signature of felsic basement rocks. Nevertheless, the high Co content of these
deposits is most likely derived from mafic basement rocks. Syn-orogenic
mineralisationin Katanga resulted from remobilisation of diagenetic sulphides.
This caused homogeneous radiogenic isotope ratios and an ore mineral assemblage
similar to the diagenetic phase. A small new metal contribution from felsic
basement rocks might be indicated by the isotopic composition of some samples
from the Kamoto deposit. This contribution is likely still detectable because
of the relatively low metamorphic gradethat was reached at Kamoto.
The pre- to syn-kinematic mineralisation phase at the Zambian stratiform Cu-Co deposits
contains gangue carbonates with isotopic compositions that correspond with
felsic rocks of the Domes Region. However, the isotopic signatures of the Nkana
and Konkola deposits differ slightly, indicating that the ore metals were
leached from Domes Region basement rocks occurring close to each deposit. The
gangue minerals of the successive syn-kinematic mineralisation phase have
isotopic compositions resembling the earlier mineralisation phase at each
deposit. Remobilisation of the precursor ore minerals most likely occurred
during the syn-kinematic mineralisation. Because of the appearance of Co and Ni
sulphides in the late-kinematicphase at Nkana, a new source probably provided
these metals to the ore forming fluid. Nevertheless, the isotopic signatures
from gangue carbonates associated with the late-kinematic phase do not reflect
this new source and are similar to the gangue signatures of the earlier phases.
The gangue carbonates of the Cu-Co deposits in the DRC have clearly different isotopic
signatures than in Zambia. In the DRC, 87Sr/86Sr ratios
of the gangue minerals are lower and the eNd values arehigher. Two other major
differences between the Katanga and Zambian Copperbelts are the higher Co
concentrations in Katangan deposits, and the carbonate-dominated host rocks in
the DRC versus the mainly siliciclastic host rocks in Zambia. These
dissimilarities are in line with each other and strongly suggest that the
source rocks for the Katanga deposits included an important mafic component.
Unfortunately, the scarse information about the basement in Katanga does not
allow to indicate a specific source rock. If the isotopic compositions of
gangue carbonates of all stratiform deposits in the DRC are considered
together, they appear to lie on a vertical trend in the Sr-eNd isotopic space.
The endmembers of this trend may represent on one side felsic rocks similar to
the felsic rocks of the Domes Region, and on the otherside unknown mafic rocks
situated in the Katanga subsurface. In analogy to the local sources for the
Nkana and Konkola deposits, local basement rocks (including mafic rocks) are
the most plausible metal source for the Katanga deposits.
The other important difference concerns the host rock type. The siliciclastics in Zambia
could have influenced the isotopic composition of the gangue carbonates of the
Nkanaand Konkola deposits. This is especially so for the Rb-Sr isotopic
system, which is prone to remobilisation and alteration. However, it is argued
that it would not change the interpretation of the Domes Regionfelsic rocks as
metal source for these deposits. Firstly, previous research has shown that the
sedimentary Katanga Supergroup rocks, whichhost the ore bodies, did not
contain the required metal budget to bea viable metal source for the known
deposits. Secondly, the metasedimentary rocks were derived from the underlying
Domes Region basement (i.e. the interpreted metal source), which would result
in similar isotopic compositions for the host rocks as for this basement.
Hence, even if the siliciclastic rocks influenced the isotopic composition of
the mineralising fluid and gangue carbonates of the Zambian deposits, the main
metal source was the local basement.
The difference inCo concentrations and related isotopic signatures between the
Katanga and Zambian stratiform deposits is clear. However, a similar but more
subtle effect is recognised for the Nkana and Konkola deposits in theZambian
Copperbelt. The Nkana deposit contains more Co and its gangue minerals have
higher eNd values and lower 87Sr/86Sr ratios than the
Konkola deposit. Mixing calculations indicate that the metal source for Nkana
must have contained at least 5% more mafic rocks than the metal source for the
Konkola deposit. Co is absent in the stratiform Cu deposits at the
north-eastern side of the Kafue Anticline. Hence, the mafic rocks that could
have provided Co to the Zambian deposits probably occur at the south-western side
of the Kafue Anticline (i.e. close to case study deposits).
For the vein-type Cu-Ag and Cu-Zn deposits of Dikulushi and Kipushi (DRC),
respectively, very high eNd signatures were found. This concurs with the
presence of mafic rocks at each mineralisation site, i.e. gabbros within the
Axial Breccia of the Kipushi anticline, and basalts in a breccia observed in
the Dikulushi mine. Nevertheless, the required mafic rocks could equally reside
in the deep subsurface. The gangue minerals of the syn-orogenic mineralisation
at Dikulushi lie on a trend between the isotopic compositions of the Gombela
Subgroup host rocks and a required mafic component. Mixing calculations demonstrate
that a mixture of both metal sources could easily have yielded the isotopic
compositions of the gangue minerals. In contrast with the very 'mafic' isotopic
signature of the Kipushi gangue carbonates, previous research on fluid
inclusions has suggested a felsic source rock. The possible influence of felsic
basement rocks on the isotopic composition of the gangue carbonates has been
assessed. If all measured variation in the eNd values is explained in terms of
more or less influence of felsic rocks, about 10% more felsic rock contribution
is needed to explain the lowest eNd signature compared to the highest.
Nevertheless, mafic rocks must have been the dominant rock type determining the
isotopic signatures at the Kipushi deposit.
Microprobe analysis of the ore sulphides has shown that the intracrystalline trace element
variability is about as largeas the intercrystalline variability for the case
study deposits. This hampers the applicability of this technique to
characterise sulphidesfrom different mineralisation phases. In addition, few
systematic differences were found between the case studies, which could have
been expected for the typical 'mafic' elements Co, Cr or Ni. However,
chalcopyrites from the late-kinematic mineralisation phase at Nkana have
slightly higher concentrations of Co than chalcopyrites from the two earlier
mineralisation phases at Nkana. This is in agreement with the input of 'mafic'
metals during this mineralisation phase. Moreover, the Ge concentration was
markably low for sulphides from the Nkana deposit, whereas the highest Ge and
Ag levels were found for samples from the Dikulushi deposit.
More in general, the results of this research show that the Sr and Nd isotopic
signatures of gangue carbonates are a valuable tool for constraining the metal
source(s) ofsulphide ore deposits. The right basement unit and source rock
type can be determined by comparing the gangue mineral signatures with the
isotopic composition of basement rocks. For the Central African Copperbelt, the
local basement beneath or surrounding the different ore deposits has shown to
be the most likely metal source to these deposits. With regard to the role of
Co, the Co-rich deposits did also display characteristic isotopic signatures,
and strongly support the existence of (still unknown) mafic bodies in the
basement that provided the Co to the mineralising fluids. For subsequent
mineralisation phases, usually related to the Lufilian orogenesis, the
importance of remobilisation of ore metals is emphasised. The next step forward
in this research domain will be the detailed charactarisation of the local
basement beneath the ore deposits, in order to relate the geochemical signature
of each ore deposit with specific rock units.
Date:1 Oct 2010 → 31 Dec 2014
Keywords:Geochemistry, Nd and Sr isotopes, Metallogenesis, Central African Copperbelt
Disciplines:Geology, Other chemical sciences, Geochemistry, Other biological sciences
Project type:PhD project