J. Longhi et al., PRESSURE EFFECTS, KINETICS, AND RHEOLOGY OF ANORTHOSITIC AND RELATED MAGMAS, The American mineralogist, 78(9-10), 1993, pp. 1016-1030
Anhydrous experiments on natural and synthetic starting materials with
basaltic to anorthositic bulk compositions show a systematic increase
in the albite component of near-liquidus plagioclase and in the Al2O3
content of orthopyroxene with increasing pressure. These results are
consistent with crystallization of the highly aluminous orthopyroxene
megacrysts and most of the plagioclase in massif anorthosite complexes
at lower crustal pressures. Comparison of plagioclase compositions fr
om near-liquidus and subliquidus experiments conducted in this laborat
ory with plagioclase compositions predicted at 1 atm for the experimen
tal temperature and liquid compositions by various empirical models in
dicates that the shift to more albitic plagioclase is predominantly a
pressure effect on the partitioning of albite (Ab) and anorthite (An)
components between plagioclase and liquid. However, even when pressure
terms are added to the models for Ab and An partitioning, there remai
n statistically significant compositional dependencies that are most a
pparent when the liquid composition is nepheline normative. These comp
ositional dependencies probably arise from the absence of highly alumi
nous and nepheline-normative liquids in the data from which the models
were constructed. Accordingly, we present empirical adjustments to th
e plagioclase-liquid models of Drake (1976), Weaver and Langmuir (1990
), and Ariskin and Barmina (1990). The positive pressure dependence of
Al2O3 in orthopyroxene coexisting with plagioclase and liquid is almo
st entirely the result of changes in orthopyroxene-liquid partitioning
and not related to increases in the Al2O3 concentration of the liquid
. Data for Al2O3 partitioning from 46 orthopyroxene-liquid and 45 pige
onite-liquid pairs taken from the literature show that pressure is the
most important control on the simple molar partition coefficient for
Al2O3. Rapid crystal growth is rejected as an alternative explanation
for the high Al2O3 contents of orthopyroxene megacrysts because rapid
growth leads to low Cr2O3 concentrations in orthopyroxene, contrary to
what is observed. These results support polybaric models for massif a
northosite petrogenesis that entail accumulation of plagioclase in evo
lved basaltic magma chambers ponded in the lower crust followed by buo
yant ascent of plagioclase-rich magmatic suspensions that intrude the
upper crust, carrying rafts of orthopyroxene megacrysts. In thick, dec
ompressing suspensions, the interplay of tie-line rotation and mass ba
lance prevents plagioclase from becoming significantly more anorthitic
. Experimental studies suggest that the transition from liquid- to sol
id-state rheology of plagioclase suspensions occurs at approximately 6
0% crystallinity for a homogeneous grain-size distribution and near st
atic conditions. However, both motion of the suspension and uneven gra
in-size distribution shift the transition to higher crystallinities. T
hus the transit of suspensions with leuconoritic composition (65-70% p
lagioclase) may be possible with minimal deformation of the entrained
plagioclase. Formation of deformed anorthosite masses may then occur a
s second-stage buoyant segregations within the upper crustal magma cha
mbers.