The Effect of Cosmic Ray Irradiation on Platinum and Palladium

46th Lunar and Planetary Science Conference (2015)
IRON METEORITES. A. C. Hunt1, M. Ek1 and M. Schönbächler1. 1ETH Zürich, Institute for Geochemistry and
Petrology, Clausiusstrasse 25, 8092 Zürich, Switzerland. E-mail: [email protected]
Introduction: Dating of early solar system processes such as differentiation and core formation of
planetesimals and planets frequently relies on shortlived decay systems such as 182Hf-182W [e.g. 1] or
Pd-107Ag [e.g. 2,3]. However, these systems are susceptible to the effects of galactic cosmic rays (GCR)
and hence, acquiring correct ages requires knowledge
of the exposure history of the meteorite. Exposure to
GCR causes secondary neutron capture on isotopes
with large neutron cross sections, which can result in
large shifts in the measured isotope ratios [4].
Platinum is a highly siderophile element with six
naturally occurring isotopes (190Pt, 192Pt, 194Pt, 195Pt,
Pt and 198Pt). Platinum isotopes are a powerful neutron dosimeter as a result of the burnout of 195Pt on to
Pt [4]. Platinum isotopes are also affected by neutron
capture by 191Ir, which can result in the production of
Pt [4]. Therefore, depending on the Ir/Pt ratio of the
sample, as well as shielding and exposure age, it is
possible to produce large 192Pt excesses. This relationship is demonstrated for the IVA, IVB, IIAB and IID
iron meteorites [5, 6, 7], highlighting that Pt isotopes
are a powerful tool for correcting the effects of GCR
Palladium is also a highly siderophile element with
six naturally occurring isotopes (102Pd, 104Pd, 105Pd,
Pd, 108Pd and 110Pd). Neutron capture by 106Pd can
potentially lead to disturbances to the 107Pd-107Ag
chronometer (Fig. 1). Large excesses in 104Pd can also
be generated if Rh/Pd ratios of the sample are sufficiently elevated, due to the burnout of 103Rh to 104Rh,
followed by β-decay to 104Pd (Fig. 1). Therefore Pd can
potentially be used as a neutron dosimeter to correct
for cosmogenic Ag.
In this study we aim to collect Pt and Pd isotopes
from the same sample aliquot in order to assess the
effects of neutron capture on both elements in various
iron meteorite groups. Previous studies reported a good
agreement between variations in 104Pd and more established dosimeters such as 192Pt [8, 9]. We present new
data for the IIAB, IIIAB, IVA and IVB iron meteorites,
including the samples North Chile, Sikhote-Alin, Henbury, Cape York, Gibeon, Muonionalusta, Tawallah
Valley and Santa Clara.
Methods: Fusion crust and weathered edges were
removed from all samples before digestion. Additionally, samples were leached in cold 2 M HCl, before dissolution in aqua regia. Primary separation of Pd from
Pt (and general matrix elements) was achieved using
AG1-X8 anion exchange resin following a procedure
modified from [10].
Platinum. A second ion exchange column was developed for further purification and removal of Ir from
Pt [11]. This is important as Ir causes tailing effects
onto Pt isotopes during measurement by MC-ICPMS.
In previous studies Ir tailing resulted in corrections of
between 2 and 15 to ε192Pt/195Pt [5]. Iridium was reduced before the samples were loaded onto anion exchange resin and then separated from Pt. This procedure was repeated in order to achieve a 191Ir/195Pt ratio
of less than 0.16, thus minimizing the effect of tailing
from Ir isotopes onto Pt. Finally, Pt cuts were dried in a
mixture of aqua regia and perchloric acid in order to
volatilize remaining Os, which generates isobaric interferences on Pt isotopes. After chemistry, both Os and Ir
are removed to an adequate level. In particular,
191 195
Ir/ Pt ratios are less than 0.02 such that no additional correction for tailing of Ir onto Pt is necessary,
therefore overcoming a major analytical uncertainty.
Figure 1. Isotopic shifts predicted by [4] for an iron
meteorite with a pre-atmospheric radius of 40 cm and
an exposure age of 625 Ma. Variations in ε107Ag/109Ag
vs. ε104Pd/105Pd for a range of Pd/Ag ratios and Rh/Pd
= 2 are shown.
Palladium. Ruthenium causes isobaric interferences
on 102Pd and 104Pd. Doping experiments show that it is
vital to achieve Ru/Pd ratios of < 0.005 in order to ensure an accurate interference correction when measuring by MC-ICPMS [12]. Separation of Pd from Ru was
achieved by drying the sample in a mixture of aqua
regia and perchloric acid. The sample was then loaded
onto anion exchange resin where Pd was successfully
separated from remaining Ru and other matrix elements. This procedure typically yields Ru/Pd ratios <
46th Lunar and Planetary Science Conference (2015)
For both Pt and Pd, yields after anion exchange
chemistry are 70 % or better, and total procedural
blanks are less than 1 ng.
Isotopic analyses. All Pt and Pd analyses were
made using a Thermo Scientific Neptune Plus MCICPMS with a Cetac Aridus II desolvating nebulizer
and standard H cones, operated in low-resolution
mode. Additionally, the MC-ICPMS is fitted with two
1012 Ω resistors which were employed for collecting
Pt and 188Os, or 101Ru and 111Cd for interference corrections.
For Pt analyses, all isotopes were collected simultaneously. In addition, 188Os and 200Hg were monitored
to correct for isobaric interferences, and 191Ir was monitored to check tailing effects onto Pt isotopes. Analyses were corrected for instrumental mass bias using
the exponential law, and were internally normalized to
both 198Pt/195Pt (‘8/5’) = 0.2145 and 196Pt/195Pt (‘6/5’) =
0.7464 [5, 6]. Samples were measured relative to NIST
SRM 3140 Pt standard solution. The daily external
precision (2 S.D.) of a 260 ppb Pt standard is typically
better than 0.8 for ε192Pt/195Pt (8/5). Reproducibility for
SRM 3140 passed through ion exchange chemistry is
ε192Pt/195Pt (8/5) = 0.2 ± 0.8 (2 S.D., n = 6).
Discussion: Palladium. Our results using the
Aridus II introduction system for the Pd SRM 3138
standard (both untreated and passed through ion exchange chemistry) and the Josephinite yielded identical
Pd isotope compositions (Fig. 2). This shows that our
new analytical method provides accurate and precise
Platinum. New data for the IVA meteorite group
confirm previous analyses by [5, 7], which suggest that
these meteorites were weakly exposed to the effects of
neutron capture due to GCR (Fig. 3). Many IIAB samples were also weakly exposed, but their low Ir/Pt ratios result in a limited spread in ε192Pt/195Pt. IVB samples show larger variations, in particular for
ε192Pt/195Pt, which is due to their higher and more variable Ir/Pt ratios (Fig. 3). Our new data correlate well
with GCR models of [4], and confirm that Pt isotopes
can be used to correct GCR effects without prior
knowledge of exposure time and depth in sample.
Figure 3. ε196Pt/195Pt vs. ε192Pt/195Pt. Modelled GCR
exposure trends from [4] are also shown for Ir/Pt ratios
of the analysed samples. New data for the IVA, IVB
and IIABs overlap with literature data [5, 6, 7].
Figure 2. The 104Pd/105Pd ratios for SRM 3138 measured using a Neptune MC-ICPMS in wet plasma mode,
compared to an Aridus II introduction system. The
Aridus II was tested with both standard H cones and an
X skimmer cone. Also shown are data for SRM 3138
and a terrestrial Pd-bearing mineral (Josephinite) which
had been passed through column chemistry.
All Pd isotopes were simultaneously analysed. Additionally, 101Ru and 111Cd were monitored to correct
for potential isobaric interferences. Instrumental mass
bias was corrected for using the exponential law and
Pd/105Pd = 1.18899 [13]. All samples were measured
relative to the NIST SRM 3138 standard. The daily
external precision (2 S.D.) of a 100 ppb Pd standard
solution is typically better than 15 and 10 ppm for
Pd/105Pd and 106Pd/105Pd, respectively (Fig. 2).
References: [1] Kleine T. et al. (2009) Geochim.
Cosmochim. Acta, 73, 5805-5818 [2] Chen J. H. and
Wasserburg G. J. (1990) Geochim. Cosmochim. Acta,
54, 1729-1743 [3] Theis K. et al. (2013) Earth. Planet.
Sci. Letters, 361, 402-411 [4] Leya I. and Masarik J.
(2013) Meteoritics & Planet. Sci., 48, 665-685 [5]
Kruijer, T. S. et al. (2013) Earth. Planet. Sci. Letters,
361, 162-172 [6] Wittig, N. et al. (2013) Earth. Planet.
Sci. Letters, 361, 152-161 [7] Kruijer et al. (2014) Science, 344, 1150-1154 [8] Wittig N. et al. (2013) LPSC
44, Abstract #2355 [9] Mayer B. et al. (2014) LPSC
45, Abstract #2581 [10] Rehkämper M. and Halliday
A. N. (1997) Talanta, 44, 663-672 [11] Hunt et al.
(2014) 77th Metsoc, Abstract #5260 [12] Ek et al,
(2014) 77th Metsoc, Abstract #5262 [13] Kelly W. R.
and Wasserburg G. J. (1978) Geophys. Res. Lett., 5,