DETERMINATION OF OXYGEN SELF-DIFFUSION IN AKERMANITE, ANORTHITE, DIOPSIDE, AND SPINEL - IMPLICATIONS FOR OXYGEN ISOTOPIC ANOMALIES AND THE THERMAL HISTORIES OF CA-AL-RICH INCLUSIONS
Fj. Ryerson et Kd. Mckeegan, DETERMINATION OF OXYGEN SELF-DIFFUSION IN AKERMANITE, ANORTHITE, DIOPSIDE, AND SPINEL - IMPLICATIONS FOR OXYGEN ISOTOPIC ANOMALIES AND THE THERMAL HISTORIES OF CA-AL-RICH INCLUSIONS, Geochimica et cosmochimica acta, 58(17), 1994, pp. 3713-3734
Oxygen self-diffusion coefficients have been measured for three natura
l diopsidic clinopyroxenes, a natural anorthite, a synthetic magnesium
aluminate spinel. and a synthetic akermanite for oxygen fugacities ra
nging from the NNO to IW buffers. The experiments employed a gas-solid
isotopic exchange technique utilizing 99% O-18-enriched CO-CO2 gas mi
xtures to control both the oxygen fugacity and the isotopic compositio
n of the exchange reservoir. Diffusion profiles of the O-18 tracer wer
e obtained by in-depth analysis with an ion microprobe. The experiment
al results, fit to the Arrhenius relation D = D(0)e((-Q/RT)), yield th
e following: D-degrees(m(2) s(-1)) Q(kJ mol(-1)) diopside 4.3[-3.8 +32
.6] x 10(-4) 457 +/- 26 akermanite 4.7[-4.4 +83.5] x 10(-7) 278 +/- 33
spinel 2.2[-1.8 +8.7] x 10(-7) 404 +/- 21 anorthite 8.4[-8.0 +174] x
10(-13) 162 +/- 36 At a given temperature, oxygen diffuses about 100 t
imes more slowly in diopside than indicated by previous bulk-exchange
experiments (CONNOLLY and MUEHLENBACHS, 1988). Our data for anorthite,
spinel, and akermanite agree well with prior results obtained by gas-
solid exchange and depth profiling methods (ELPHICK et al., 1988; REDD
Y and COOPER, 1981; YURIMOTO et al., 1989, respectively). Since these
other experiments were conducted at different oxygen fugacities, this
agreement indicates that diffusion of oxygen in these nominally Fe-fre
e minerals is not greatly affected by f(o2) in the range between pure
oxygen and the iron-wustite buffer. However, our diffusion coefficient
s for anorthite, melilite, and spinel are also uniformly lower than th
ose obtained by bulk analysis of crushed powders at similar temperatur
es (MUEHLENBACHS and KUSHIRO, 1974; HAYASHI and MUEHLENBACHS, 1986; AN
DO and OISHI, 1974). The oxygen diffusion data are used to evaluate th
e effects of three different types of thermal histories upon the oxyge
n isotopic compositions of minerals found in Type B Ca-Al-rich inclusi
ons (CAIBs) in carbonaceous chondrites: (1) gas-solid exchange during
isothermal heating, (2) gas-solid exchange as a function of cooling ra
te subsequent to instantaneous heating, and (3) isotopic exchange with
a gaseous reservoir during partial melting and recrystallization. Wit
h the assumptions that the mineral compositions within a CAIB were uni
formly enriched in O-16 prior to any thermal processing, that effectiv
e diffusion dimensions may be estimated from observed grain sizes, and
that diffusion in diopside is similar to that in fassaitic clinopyrox
ene, none of the above scenarios can reproduce the relative oxygen iso
topic anomalies observed in CAIBs without improbably long or unrealist
ically intense thermal histories relative to current theoretical model
s of nebular evolution. The failure of these simple models, coupled wi
th recent observations of ''disturbed'' magnesium isotopic abundances
and correlated petrographic features in anorthite and melilite indicat
ive of alteration and recrystallization, suggests that the oxygen isot
opic compositions of these phases may have also been modified by alter
ation and recrystallization possibly interspersed with multiple meltin
g events. Because the modal abundance of spinel remains relatively con
stant for plausible melting scenarios, and its relatively sluggish dif
fusion kinetics prevent substantial equilibration, Mg-AI spinel is the
most reliable indicator of the oxygen isotopic composition of precurs
or material which formed Type B CAIs.