Rit (Vísindafélag Íslendinga) - 01.06.1984, Blaðsíða 164

160
PÁLL IMSLAND
the more likely are they to be found in the tetrahedral complex state in the
liquid. The ionic potentials ofGr3+ and Al3+ are 4.77 and 5.88 respectively.
Al3+ is thus far more likely than Cr3+, to occur as a part of a tetrahedral
complex. Ni2+, Cr3+, V3+, Sc3+ (according to Ringwood (op.cit.)), behave
as free cations in a liquid of low anion to Si+A1 ratio, while in a liquid of
high anion to Si+Al ratio they may at least partly form tetrahedral
complexes, while Al3+ does so on a large scale. Such complexes are not as
readily accepted by the crystal structure as free cations during crystalliza-
tion. Under these conditions, elements like Cr may, contrary to their more
usual behaviour, become concentrated in the residual liquid. Analyses of F
and S content of few whole rock samples (Table 11) and the hydrous
minerals (see later) might indicate that the Jan Mayen magmas were
generally high in these elements, especially F. If the primary magmas were
volatile rich, these volatiles would have prevented polymerization of tetra-
hedral complexes, and if this volatile content was relatively anion rich, it
would follow that the total amount of free tetrahedral complexes in the
liquid was high, allowing Cr3+ to occur, at least partly, in the complex form.
The decreasing ÍD2 of the remaining liquid which occurred during the
crystallization of the pre-silicate early spinels of the ankaramites, results in a
decrease in the total anion to Si+Al ratio of the liquid. This decreasing
anion to Si+Al ratio would break up tetrahedral complexes in the liquid,
starting with Cr3+ complexes rather than Al3+ complexes, and thus increase
the relative amount of free Cr of the liquid. This means that the crystallizing
spinel has an easier access to Cr relative to A1 as the crystallization
continues. Whether the compositional and physicochemical conditions of
the primary magmas of the Jan Mayen ankaramites actually were in favour
of this process, or whether the changes of the fCU are enough to cause the
necessary changes in the anion to Si+Al ratio described above, is not
known. Some kind of structural changes in the liquid, causing increased
availability of Cr relative to A1 seems, though, to be a more likely explana-
tion, for the increased Cr/(Cr+Al) ratio, shown by the pre-silicate early
spinels, rather than a direct increase in the Cr content of the liquid
crystallizing.
Huebner et al. (1976) state: “In terrestrial environments, in which most
chromium is present as the Cr3+ species, the order of partitioning is oxide >
Ca rich pyroxenes íS Ca poor pyroxene > melt > olivine.” These authors
give 0.9 and 2.5 as average partition coeíficients of Cr203 for olivine-liquid
and clinopyroxene-liquid respectively. They also note that: “Sparse data
suggest, that at pressures to at least 20 kbars, the partitioning of chromium
follows the same pattern as at low pressures.” And further: “DCr is not
related to temperature ofour runs in any simple manner” (DCr=wt. per cent
Cr203 cryst/Cr203 Hq). Using these average values as partition coefficients in
the Jan Mayen case, one gets 0.33—0.47 wt. per cent Cr203 in the liquid at
the time when the chromian diopside started to crystallize and 0.04—0.06