HR: 16:15h
AN: V12I-02 [PDF]
TI: Isotope Dilution and LA ICPMS Study of Trace Elements in Garnets: Implications for Sm-Nd and Lu-Hf
Dating.
AU: Anczkiewicz, R
EM: rob@gl.rhul.ac.uk
AF: Department of Earth Sciences, University College London, Gower Street, London, WC1E 6BT
United Kingdom
AU: Anczkiewicz, R
EM: rob@gl.rhul.ac.uk
AF: Department of Geology, Royal Holloway University of London, Egham Hill, Egham, TW20 0EX
United Kingdom
AU: Platt, J P
EM: john.platt@ucl.ac.uk
AF: Department of Earth Sciences, University College London, Gower Street, London, WC1E 6BT
United Kingdom
AU: Thirlwall, M F
EM: m.thirlwall@gl.rhul.ac.uk
AF: Department of Geology, Royal Holloway University of London, Egham Hill, Egham, TW20 0EX
United Kingdom
AU: * Alard, O
EM: o.alard@open.ac.uk
AF: Department of Earth Sciences, Open University, Walton Hall, Milton Keynes, MK7 6AA
United Kingdom
AB:
One of the main advantages of garnet geochronology is a possibility of establishing a direct link between isotopic ages and
PT conditions. However, both Lu-Hf and Sm-Nd garnet dating can strongly be affected by submicroscopic inclusions capable of
dominating the Sm-Nd and Lu-Hf budget. We investigated possible effects of various inclusions on Sm-Nd and Lu-Hf systems by
combined isotope dilution and LA ICPMS studies.\\
Internal isochrons obtained for 6 high-grade blocks of metabasites from the Franciscan complex yielded highly precise Lu-Hf
ages ranging from 114 to 170 Ma, but failed to provide Sm-Nd dates. The main reason for failure of the Sm-Nd dating was a
rather large amount of matrix silicate inclusions, which contained $>$ 100 times more Nd than analyzed "impure" garnet
fractions. The same inclusions had very limited influence on the Lu-Hf budget. This is mainly due to strong enrichment of
garnets in heavy REE and due to much lower Hf concentrations in rock forming silicates. The $^{176}$Lu/$^{177}$Hf ratios
obtained for the analyzed garnets are typically between 1.5 and 8, but for two samples with spessartine-rich garnets, values
range between 21 and 28 and are the highest yet reported. Small amount of zircon and rutile inclusions did not notably
suppress $^{176}$Lu/$^{177}$Hf ratios. This is mainly due to their small size ($<$10 $\mu$m) and low abundance, but also
because of their limited dissolution during sample digestion on a hotplate. Our results demonstrate that even samples with
significant amounts of inclusions can yield precise and accurate Lu-Hf dates.\\
For upper amphibolite facies metapelites from Vietnam, we obtained well defined Sm-Nd ages of 52.1$\pm$2.5 and 31.4$\pm$1.0
Ma, which are grossly discordant with the corresponding Lu-Hf dates of 84.1$\pm$1.6 Ma and 77.9$\pm$1.6 Ma, respectively. We
therefore conducted LA ICP MS measurements on 4 selected samples.\\
Sm/Nd ratios show rather flat zonation profiles throughout most of the crystal, with an up to 30% increase in the rims. The
zonation closely follows Fe and Mn profiles. Lu/Hf ratios, on the other hand, show very large variations with low values in
the rims reaching up to several hundreds in the cores and shows closer association with Y, Ca and Mg. Preservation of very
strong zonation pattern of heavy REE together with Lu-Hf ages significantly older than corresponding Sm-Nd dates suggest
higher closure temperature for the former system and possible inheritance problems in the analyzed samples.\\
Sm/Nd ratios obtained by isotope dilution for garnets leached with sulphuric acid (SAL) were only about 10% lower than those
estimated by LA ICPMS. This demonstrates high efficiency of SAL leaching in purifying garnets from monazite inclusions,
which is of fundamental importance for achieving reliable ages. The same leaching technique had very different influence on
Lu-Hf system. Garnets with $^{176}$Lu/$^{177}$Hf ratios $>$20 after leaching yielded even higher ratios after leachcing. On
the other hand, SAL applied to garnets with more typical $^{176}$Lu/$^{177}$Hf ratios were brought down even by up to 50%
relative to the unleached fractions. The latter case demonstrates removal of inclusions with Lu/Hf ratios higher than those
in garnet. The most likely mineral with such high ratio in analyzed metabasites is apatite and in the case of metapelites-
xenotime.
DE: 1000 GEOCHEMISTRY (New field, replaces Rock Chemistry)
DE: 1035 Geochronology
DE: 1065 Trace elements (3670)
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting