magnesium, silicon and calcium isotopes in central european tektites

46th Lunar and Planetary Science Conference (2015)
WITH THE RIES AREA SEDIMENTS. T. Magna1, J. Farkaš1, Z. Rodovská1,2, J. Trubač1, R.B. Georg3, C.
Holmden4, K. Žák5, 1Czech Geological Survey, Prague, Czech Republic. E-mail: [email protected]
Faculty of Science, Charles University in Prague, Czech Republic. 3Water Quality Centre, Trent University, Peterborough, Canada. 4University of Saskatchewan, Saskatoon, Canada. 5Institute of Geology, v.v.i., Academy of Science of the Czech Republic, Prague, Czech Republic.
Introduction: Isotopes of Mg and Ca appear to be
sensitive indicators of protolith chemistry considering
variations of their isotope abundances in silicate and
carbonate lithologies [1–2] although Ca as well as Si
isotopes may also indicate biologically mediated processes [3]. However, a handful of studies exists that
deals with isotope fractionation of these major elements at magmatic temperatures (see review by [4] for
Si) and extreme events like impact processes remain
largely unexplored [5]. At then-achievable analytical
precision, [6] did not discover measurable Mg isotope
offsets in tektites from common silicate reservoirs and
implied limited loss of Mg that could impart Mg isotope fractionation during impact events. Silicon isotopes are thought to fractionate during core segregation [7] but the response to extreme temperatures remains unknown although it could bear new information on conditions taking place shortly after the
giant collisions. Because specific isotope fractionations exist between the vegetation and soils, and due to
the fact that some organic materials have been embedded in impact-related glasses [8] and that a minor contribution of ashes from local vegetation into impactrelated materials cannot be excluded [9], the highprecision analyses could provide a fingerprint for determining the former presence of flora at impact sites.
Impacts are major surface-shaping processes in
early stages of planetary evolution and despite the decreasing rate of large impacts throughout the history,
the Earth’s surface has constantly been reworked by
small to large impacts in its modern era. During the
bombardment, natural silica-rich glasses are produced
both within the craters and around them, and also as
distal ejecta (tektites). Tektites have been recognized
in four known strewn fields, which could be assigned
to a specific crater or at least the area of origin: North
American, Central European, Ivory Coast and Australasian. However, a direct link to their respective protoliths is unclear.
For this study, we selected a range of Central European tektites (moldavites) with distinctive SiO2 contents and collected from five different strewn subfields
in order to cover both the maximum range in chemistry
of moldavite and the range in ballistic transport trajectories. In addition, chemically variable Miocene sedi-
ments (mainly the Upper Freshwater Molasse) from
the Ries target area in south-eastern Germany (centered close to Nördlingen) that could have contributed
to the finite moldavite melt pool were also analyzed.
On the basis of the correspondence of the ages [10],
the Ries impact structure has been accepted as the parent crater to moldavites.
Results and discussion:
Magnesium. Sediments from the Ries area span a
26/24MgDSM3 range of >3‰ (–3.0 to 0.1‰) with two
distinct trends, apparently related to major element
chemistry (MgO, FeOtot) and volatile contents (F,
P2O5, H2O). This suggests major lithological contrasts
determining the Mg isotope fractionation (carbonateversus clay- versus quartz-rich sediments) although
clay minerals isolated from three sediments do not
show a uniform sense of Mg isotope fractionation relative to bulk sediments. Overall, carbonate-rich samples
tend towards significantly lower 26/24Mg values than
silicate-rich samples. The moldavites span a significantly more constrained range of 26/24Mg values (–1.6
to 0.4‰) and no particular difference is found for
moldavites from distinct strewn subfields. The
26/24Mg values overlap with most of the sediments
although for the isotopically heaviest moldavite no
sedimentary equivalent with similarly heavy Mg isotope composition has been found in this study. These
findings may reflect either incomplete sampling of
sedimentary precursors or a minor Mg isotope fractionation related to modest Mg loss during the impact
event. But with respect to extremely variable Mg contents in the source sediments, any Mg loss during moldavite formation cannot be tested. While both the moldavites and sediments follow mass-dependent isotope
fractionation in triple-Mg isotope plot, subtle differences between equilibrium (slope ~0.529) and kinetic
(slope ~0.511) effects are recognized that may be related to low-temperature versus high-temperature conditions. A single sample of a residual glass produced in
a power plant during cobmustion of organic matter
(“straw glass”) has a common Mg isotope composition
but low Mg content largely precluding significant organic component in the source of moldavites.
46th Lunar and Planetary Science Conference (2015)
Silicon. The 30SiNBS28 values vary greatly for the
Ries area sediments (–1.4 to 0.3‰) but no particular
difference is observed for silica-rich versus silica-poor
samples. A straw glass and phytolith samples isolated
from Miscanthus sp. carry isotopically heavy Si signature while phytoliths from Equisetum sp. carry 30Si at
the lower range of values reported for sediments in this
study. However, the cumulative effect of plants on the
bulk Si systematics is rendered insignificant given the
Si-rich nature of target lithologies. Moldavites show
more homogeneous 30Si values (–0.5 to –0.1‰) with
no difference between the individual strewn subfields.
Two moldavites show a systematic and reproducible
mass-independent offset in 30Si fractionation (with a
slope of ~0.56), the origin of which remains presently
unclear, but does not appear to be related to peculiar
sample chemistry and/or analytical procedures. A
slightly wider scatter among the moldavite Si data does
not allow to distinguish between different fractionation
Calcium. Contrary to Mg, Ca isotopes are largely
invariant in sedimentary samples from the Ries area,
with 44/40CaSRM915a values ranging between 0.58 and
0.85‰. Two separate positive trends emerge in relation to the carbonate content. A sample of straw glass
has a distinctly low 44/40Ca of –0.4‰ which implies a
very limited contribution of Ca from vegetation in the
moldavites (see the data below). Also, given the low
Ca and Mg contents in plants coupled with high
Ca/Mg (Ca/Mg in tektites is ~1), significant proportion
of ashes (several tens percent by volume) would be
required to provide detectable changes to Ca–Mg isotope compositions while these amounts would violate
other major element constraints. The moldavites display a limited range in 44/40Ca values (0.60 to 0.83‰),
within the isotope variability of sediments. Slight differences exist between the different strewn subfields of
the Central European moldavites, with the Moravian
moldavites having lighter Ca isotope compositions
compared with the rest of the suite. Unlike sediments,
moldavites do not form a dichotomy but are displaced
between low-CaO sedimentary components with the
end member 44/40Ca values, i.e., carbonate-rich sediments represent only a marginal addition to moldavite
Several observations can be inferred from the cumulative Mg–Si–Ca isotope dataset. The order of decreasing half-mass condensation temperatures [11] is
from Ca to Mg to Si. Calcium shows the highest degree of homogeneity (~0.06‰/amu) in moldavites
while both Si and, in particular, Mg display a higher
extent of variability (~0.23‰/amu and ~1.0‰/amu,
respectively). Therefore, larger degree of variability
discovered for Mg may be, to a certain extent, surprising but this may perhaps be explained by the large
range in 26/24Mg values in the possible sedimentary
parentage, compared with that found for Si.
Acknowledgements: The study is a part of the
Czech Science Foundation project 13-22351S.
References: [1] Saenger & Wang (2014) Quater
Sci Rev 90, 1-21. [2] Holmden & Bélanger (2010)
GCA 74, 995-1015. [3] Ding et al. (2005) Chem Geol
218, 41-50. [4] Savage et al. (2014) Lithos 190-191,
500-519. [5] Davis et al. (1990) Nature 347, 655-658.
[6] Esat (1988) GCA 52, 1409-1424. [7] Georg et al.
(2007) Nature 447, 1102-1106. [8] Howard et al.
(2013) Nature Geosci 6, 1018-1022. [9] Řanda et al.
(2008) MAPS 43, 461-477. [10] Gentner et al. (1963)
GCA 27, 191-200. [11] Lodders (2003) ApJ 591,