 |
| Titel |
QEMSCAN+LA-ICP-MS: a ‘big data’ generator for sedimentary provenance analysis |
| VerfasserIn |
Pieter Vermeesch, Martin Rittner, Eduardo Garzanti |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250152344
|
| Publikation (Nr.) |
 EGU/EGU2017-17171.pdf |
|
|
|
|
|
| Zusammenfassung |
| Sedimentary provenance may be traced by ‘fingerprinting’ sediments with chemical,
mineralogical or isotopic means. Normally, each of these provenance proxies is characterised
on a separate aliquot of the same sample. For example, the chemical composition of the bulk
sample may be analysed by X-ray fluorescence (XRF) on one aliquot, framework
petrography on another, heavy mineral analysis on a density separate of a third split, and
zircon U-Pb dating on a further density separate of the heavy mineral fraction. The labour
intensity of this procedure holds back the widespread application of multi-method
provenance studies. We here present a new method to solve this problem and avoid mineral
separation by coupling a QEMSCAN electron microscope to an LA-ICP-MS instrument and
thereby generate all four aforementioned provenance datasets as part of the same workflow.
Given a polished hand specimen, a petrographic thin section, or a grain mount, the
QEMSCAN+LA-ICP-MS method produces chemical and mineralogical maps from
which the X-Y coordinates of the datable mineral are extracted. These coordinates
are subsequently passed on to the laser ablation system for isotopic and, hence,
geochronological analysis. In the process of finding all the zircons in a sediment grain
mount, the QEMSCAN yields the compositional and mineralogical compositions as
byproducts.
We have applied the new QEMSCAN+LA-ICP-MS instrument suite to over 100 samples
from three large sediment routing systems: (1) the Tigris-Euphrates river catchments and
Rub’ Al Khali desert in Arabia; (2) the Nile catchment in northeast Africa and (3) desert
and beach sands between the Orange and Congo rivers in southwest Africa. These
studies reveal (1) that Rub’ Al Khali sand is predominantly derived from the Arabian
Shield and not from Mesopotamia; (2) that the Blue Nile is the principal source
of Nile sand; and (3) that Orange River sand is carried northward by longshore
drift nearly 1,800km from South Africa to southern Angola. In addition to these
geological findings, the first applications of QEMSCAN+LA-ICP-MS highlight
some key advantages of the new workflow over traditional provenance analysis:
(a) the new method not only increases sample throughput but also improves data
quality by reducing significant biases associated with mineral separation and grain
selection; (b) the three case studies highlight the importance of zircon ‘fertility’ for
interpreting detrital zircon U-Pb datasets, and the ability of QEMSCAN to quantify this
crucial parameter semi-automatically; (c) QEMSCAN+LA-ICP-MS provides an
opportunity to add textural information to detrital geochronology and, for example,
quantify possible grain-size dependence of U-Pb age distributions. But besides these
advantages, the three case studies also reveal a number of limitations: (a) mineral
identification by QEMSCAN is not as reliable as commonly achieved by human
observers; (b) heavy mineral compositions obtained by QEMSCAN cannot easily be
compared with conventional point counting data; and (c) apparent grain sizes can be
greatly affected by polishing artefacts. In conclusion, QEMSCAN+LA-ICP-MS is a
transformational new technique for provenance analysis but should be used with
caution, in combination with conventional petrographic and heavy mineral techniques. |
|
|
|
|
|
|