Geology Reference
In-Depth Information
of K-bearing MORB is de
ned by incongruent melting of phengite through a
reaction, phengite + coesite + omphacite = garnet + melt.
Schmidt (1996) considered that the subduction history of highly soluble ele-
ments such as potassium is directly related to
fluid content. The quantity of free
fluid phase present in the descending oceanic crust is expected to be small and most
of the
fluid should ascend. In his experiments with 1 wt% H
2
O, Schmidt observed
that in case of the MORB bulk composition, almost pure KAlSi
3
O
8
(K-hollandite
structure) instead of phengite appears. In experiment with >1 wt% water K-hol-
landite or other K-bearing crystalline phase was not observed as a breakdown
product of phengite instead K-bearing silicate phases (submicroscopic to 10
m
appear between grain boundaries between garnet and clinopyroxene. These phases
probably precipitated from the
µ
fluid phase and is interpreted as the breakdown
product of phengite to omphacite and garnet and a K-rich
fluid.
The experiments of Schmidt suggest that K was strongly partitioned in the
fluid
phase and may be interpreted as evidence of breakdown of phengite to ompha-
cite + garnet and a K-rich
fluid. These experiments suggest that phengites (at low
pressures), and K-hollandite and also omphacite, can carry K to great depths in the
subduction zones, but that any
fluid resulting from phengite breakdown will be
strongly enriched in K and other such mobile elements as Rb, U, Pb and B. Thus,
fluids passing through or arising from subducted oceanic crust at great depths may
leach some of the K from the crust during metamorphism.
Tatsumi (1989, cited by Schmidt 1996) proposed that a descending oceanic crust
should release all its potassium when amphiboles break down at 2.5 Gpa. This
should lead to subsequent metasomatism of the overlying mantle wedge and
transportation of K
2
O through such mantle phases as phlogopite or K-richterite. The
experiments of Schmidt suggest an alternative mechanism for storing K
2
O in the
subduction zones. Minor amounts of K-rich sediments and abundant K-poor
basaltic and gabbroic crust will loose a small amount of K into
fluids, produced
during initial partial dehydration at shallow depths (<1.0 GPa). According to
Schmidt the presence of minor amounts of phengite in metabasalts and major
amounts of phengites in metasediments from coesite-bearing eclogite terrains,
documents that K is stored in subducted crust to conditions beyond depths of
90 km. He suggests that at depths of 75
300 km with the increasing pressure during
-
subduction, omphacite (40
55 %) could retain almost all potassium present in a
MORB bulk composition. He further suggested that the phengites
-
finally break
down at depths of 300 km and would release only minor K and H
2
O from MORB
composition. However, in the depth range of 100
300 km when K-omphacite is
formed continuously from phengite through the reaction, Phengite-
-
→
K-feld-
spar + pyroxene + enstatite + coesite + K-rich
fluid. Potassium will then be released
continuously from the subducted crust through the solution in
fluids. Schmidt
(1996) suggested that greywackes with higher K
2
O contents (1.4
-
4 wt%) and lower
omphacite abundances (20
-
30 vol%) might still have signi
cant amount of
phengite (10
35 vol%) at 300 km depth (10 GPa). He thought that a pulse of K-rich
-
fluid may be generated in sediments at a depth of
300 km, when K-hollan-
*
dite + K-rich
fluid is formed from phengite. At still greater depths the descending
Search WWH ::
Custom Search