|
BUSHVELD PROJECT
The Stillwater complex in Montana and the Bushveld complex
in South Africa are among the few intrusions that contain
extensive stratigraphic sections in which the halogen-bearing
minerals are unusually rich in chlorine. This is in contrast
to the majority of layered igneous complexes that crystallized
from basaltic parent liquids (e.g., the Skaergaard intrusion
in Greenland, the Munni Munni and the Windimurra complexes
in Western Australia and the Great Dyke in Zimbabwe) in which
the halogen-bearing minerals are distinctly fluorian. Because
of the wide variation in halogen concentrations in these two
intrusions, they are ideal intrusions in which to study possible
causes for these variations and their potential for economic
metal transport.
This study has three major objectives:
1) Halogen variations in the Bushveld complex and related
rocks: Halogen geochemistry below the platiniferous Merensky
Reef of the Bushveld complex is still poorly known, as are
halogen contents of the associated marginal sills. We propose
to more fully sample these rocks to investigate the compositions
of halogen-bearing minerals using electron microprobe. The
associated sills are of particular interest as their compositions
may preserve evidence of the original halogen concentrations
of Bushveld magma(s). Modern magmas tend towards higher Cl/F
ratios with higher overall volatile concentrations, suggesting
that the parent magmas of the Stillwater and Bushveld complexes
were "wetter" than commonly assumed. If so, this
has important consequences for crystallization behavior and
the transport of ore elements.
2) Cl isotopic characteristics: Preliminary Cl isotopes
on a handful of Stillwater biotite suggest these fluids were
derived from a crustal source (i.e., relatively light Cl)
rather than degassed mantle as is characteristic of MORB (relatively
heavy Cl). We propose to analyze and compare stable Cl isotopic
characteristics with existing data for Sr, O and H isotopic
trends for these two intrusions (particularly the Bushveld
complex), their country rocks and associated contemporaneous
sills and dikes. Along with petrographic and field evidence,
this data should allow us to constrain models suggesting that
high Cl is due to a mantle fluxing event, contamination or
metamorphic fluid infiltration from the country rocks. Comparison
with Cl-poor intrusions and other particular magma types (esp.
boninites) will also test if Cl isotopes are distinct and
if they have similarities with magmas believed to have formed
by volatile fluxing of mantle source regions.
3) PGE-Halogen correlations: Evolution of Cl-rich fluids
from intercumulus liquids could be a potentially useful transport
mechanism for the platinum-group elements (PGE). It has been
suggested that the cumulates below the principle ore zones
in these intrusions were the source for the PGE in the deposits.
Hence, we want to determine to what extent of background PGE
(to be analyzed by ICP-MS) , S , base metals and Cl are correlated.
These data will be combined with petrographic and geochemical
studies, augmented with quantitative modeling of degassing,
PGE transport and mineral-liquid-fluid equilibria, to test
models of PGE transport and concentration.
Middle Banded Zone of the Stillwater Complex
Compaction of igneous cumulates is, potentially, a ubiquitous
process during accumulation of crystals in layered intrusions.
Local development of mineral lamination, granophyric , and
variations in grain morphology indicate that variable compaction
took place during the accumulation of portions of the Middle
Banded series of the Stillwater complex. This interpretation
is supported by preliminary microprobe data that show lower
Mg#'s for pyroxenes in layers interpreted (on the basis of
their texture and stratigraphic position) to have been enriched
in more evolved liquid compacted out of underlying cumulates.
The Bronzitite zone, Ultramafic series - Banded series contact,
and parts of Gabbronorite III provide ideal settings for assessing
the importance of various controls on compaction. A quantitative
approach to documenting mineral lamination and grain morphology
is proposed that, coupled with chemical data, will provide
a basis for assessing the extent of compaction that has taken
place in cumulates in general. This approach should be particularly
valuable for studying cumulates where exposure is limited
(e.g., oceanic drill core).
A quantitative understanding of the physics of compaction
is of fundamental importance in evaluating the role that compaction
has played in a particular setting. The proposed study would
include development of numerical models of compaction that
can be used to determine the importance of several parameters
including: variation in cumulate density with height, the
rate of heat loss and generation within the cumulus pile,
and the of accumulation of crystals. The theoretical model
will aid in understanding correlations between mineral lamination,
bulk major and trace-element composition, mineral compositions,
and liquid/fluid compositions as indicated by accessory phases
(e.g. apatite). Results of the field study can be used to
estimate properties of the cumulus matrix, in particular its
effective viscosity. Initial models will be designed to be
broadly applicable to all type of rocks formed by accumulation
of crystals, including not only layered continental intrusions,
but alsoophiolites and mid-ocean ridge settings. Specific
parameters applicable to the Stillwater may then be applied
to this model and the predicted results compared to field
observations. Qualitative results of preliminary modeling
of density variations in the cumulus pile have predicted distributions
of intercumulus liquid that can explain the incompatible trace
element distribution in the Great Dike and Munni Munni complexes.
Major results of this study will include: (1) A systematic
evaluation of the significance of compaction in the Stillwater
and its chemical, textural, and mineralogical consequences.
(2) The first large scale quantitative investigation of mineral
lamination and grain morphology in a major, layered, continental
intrusion. (3) Field based estimates of properties of the
cumulus pile, such as matrix viscosity, that govern compaction
rates. and (4) A numerical model of compaction that has been
field tested against a variety of different settings in the
Stillwater complex.
FINE-SCALE LAYERING PROJECT SUMMARY
We propose to use numerical models of crystallization processes
to further understand the origins of igneous layering. Most
previous work on in situ mechanisms for the formation of igneous
layering has involved nucleation mechanisms. We are interested
in how post-nucleation processes, specifically crystal aging,
may modify an original crystal distribution imposed either
by in situ nucleation or by crystal settling events. Previous
work by the Boudreau suggests that many features of fine-scale
layering in layered intrusions, such as the formation of "doublet"
layers and patterning developed in the plane of layering,
may be explained as the result of pattern development during
crystal aging, a mechanism originally developed by P. Ortoleva
and co-workers to describe similar phenomena in crystallizing
salt systems. We propose to extend this work by exploring
the effects of phase-phase variability in surface free energy
and include simple cotectic crystallization behavior coupled
with thermal and mass diffusion and advection in the numerical
modeling, and compare the results with examples from layered
intrusions.
|
