Environmental Engineering Reference
In-Depth Information
basaltic rocks commonly dissolve completely during this process. However, if a
slightly lower temperature is used for the partial dissolution step (12 hours at
180
C vs. 210
C), some zircon survives, although often as skeletal remnants,
with clear evidence of having been partially dissolved (e.g.
Figure 4.3e
-1,2,3).
In cases where the grains contain elongate inclusions, chemical abrasion com-
pletely dissolves the inclusions, leaving thin, blade-like fragments of zircon, that
presumably had overgrown the inclusion. These shards are then completely dis-
solved and usually result in high-precision concordant analyses. In cases where
chemical abrasion of a single grain yields a single large fragment and many smaller
fragments, dates on both populations are identical within uncertainty.
4.2.4 Correction for initial Th exclusion
While the U
-
Pb zircon system may yield the most precise results for LIP eruptions,
the accuracy of these measured dates is in part controlled by the necessary
correction for the initial preferential exclusion of
230
Th (and corresponding
206
Pb
de
ciency) when zircon crystallizes. The magnitude of the correction can be
signi
cant when the highest levels of precision are achieved. Accurate correction
requires knowledge of the relative partitioning of Th and U for a zircon and
the magma from which it crystallized. The bulk partition coef
cient for Th/U is
determined using the Th/U of both the zircon and magma it crystallized from
(Schärer,
1984
). The zircon Th/U value is often determined from the measured
208
206
232
208
238
206
Pb
systems. Zircons from flood basalts often have high Th/U values compared with
those from silicic rocks (e.g. Palisades sill
~
1.2). A Th/U for the magma may be
measured directly from melt inclusions or inferred from whole-rock composition.
It is unlikely that whole-rock data reliably represent the liquid from which a zircon
in a
Pb
-
Pb, assuming concordance between the
Th
-
Pb and
U
-
flood basalt crystallized, as the magmatic processes leading to the concen-
tration of incompatible elements, such as Zr, could very likely fractionate
Th from U. Efforts to map the Th and U concentrations within sills reveals that
although the concentrations of U and Th mimic that of Zr, the Th/U value remains
nearly constant (
first observed by Gottfried
et al
.(
1968
) in diabase
sills from CAMP and Tasmania, and is shown here, using data from Shirley (
1987
)
for the Palisades sill (
Figure 4.2
).
The magnitude of the
238
U
-
~
4.0). This was
206
Pb Th-correction is correlated with the difference
between Th/U of the zircon crystal and that of the magma from which it crystal-
lizes. There exists, however, a threshold at which this correction will reach
a maximum, resulting in a near constant
238
206
Pb date for some plausible
Th/U magma compositions. For the Palisades sill, a series of Th-corrected
weighted mean
U
-
238
206
U
-
Pb dates, calculated over a range of assumed Th/U
