 |
| Titel |
Diamond formation through isochemical cooling of CHO fluids vs redox buffering: examples from Marange peridotitic and Zimmi eclogitic diamonds |
| VerfasserIn |
Karen V. Smit, Thomas Stachel, Richard A. Stern, Steven B. Shirey, Andrew Steele |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250145264
|
| Publikation (Nr.) |
 EGU/EGU2017-9187.pdf |
|
|
|
|
|
| Zusammenfassung |
| Traditional models for diamond formation within the lithospheric mantle invoke either
carbonate reduction or methane oxidation. Both these mechanisms require some oxygen
exchange with the surrounding wall-rock at the site of diamond precipitation. However,
peridotite does not have sufficient buffering capacity to allow for diamond formation via
these traditional models and instead peridotitic diamonds may form through isochemical
cooling of H2O-rich CHO fluids [1].
Marange mixed-habit diamonds from eastern Zimbabwe provide the first natural
confirmation of this new diamond growth model [2]. Although Marange diamonds do not
contain any silicate or sulphide inclusions, they contain Ni-N-vacancy complexes detected
through photoluminescence (PL) spectroscopy that suggest the source fluids equilibrated in
the Ni-rich depleted peridotitic lithosphere. Cuboid sectors also contain abundant
micro-inclusions of CH4, the first direct observation of reduced CH4-rich fluids that are
thought to percolate through the lithospheric mantle [2].
In fluid inclusion-free diamonds, core-to-rim trends in δ13C and N content are used to
infer the speciation of the diamond-forming fluid. Core to rim trends of increasing
δ13C with decreasing N content are interpreted as diamond growth from oxidized
CO2- or carbonate-bearing fluids. Diamond growth from reduced species should
show the opposite trends - decreasing δ13C from core to rim with decreasing N
content. Within the CH4-bearing growth sectors of Marange diamonds, however,
such a ’reduced’ trend is not observed. Rather, δ13C increases from core to rim
within a homogeneously grown zone [2]. These contradictory observations can
be explained through either mixing between CH4- and CO2-rich end-members of
hydrous fluids [2] or through closed system precipitation from an already mixed
CH4-CO2 H2O-maximum fluid with XCO2 (CO2/[CO2+CH4]) between 0.3 and 0.7
[3].
These results demonstrate that Marange diamonds precipitated from cooling
CH4-CO2-bearing hydrous fluids rather than through redox buffering. As this growth
mechanism applies to both the fluid-rich cuboid and gem-like octahedral sectors of Marange
diamonds, a non-redox model for diamond formation from mixed CH4-CO2 fluids is
indicated for a wider range of gem-quality peridotitic diamonds. Indeed, at the redox
conditions of global diamond-bearing lithospheric mantle (FMQ -2 to -4; [4]), CHO fluids are
strongly water-dominated and contain both CH4 and CO2 as dominant carbon species
[5].
By contrast diamond formation in eclogitic assemblages, through either redox buffering
or cooling of carbon-bearing fluids, is not as well constrained. Zimmi diamonds from the
West African craton have eclogitic sulphide inclusions (with low Ni and high Re/Os) and
formed at 650 Ma, overlapping with the timing of subduction [6]. In one Zimmi diamond, a
core to rim trend of decreasing δ13C (-23.4 to -24.5 ) and N content is indicative of formation
from reduced C2H6/CH4-rich fluids, likely derived from oceanic crust recycled during
Neoproterozoic subduction. Unlike mixed CH4-CO2 fluids near the water maximum,
isochemical cooling or ascent of such reduced CHO fluids is not effficient at diamond
precipitation. Furthermore, measurable carbon isotopic variations in diamond are not
predicted in this model and therefore cannot be reconciled with the ∼1 ‰ internal variation
seen. Consequently, this Zimmi eclogitic diamond likely formed through redox
buffering of reduced subduction-related fluids, infiltrating into sulphide-bearing
eclogite.
References 1. Luth and Stachel, 2014. CMP, 168, 1083 2. Smit et al., 2016. Lithos, 265,
68-81 3. Stachel et al., in review 4. Stagno et al., 2013. Nature, 493, 84–88 5. Zhang and
Duan, 2009. GCA 73, 2089–2102 6. Smit et al., 2016. Precamb Res, 286, 152-166 |
|
|
|
|
|
|