Nuclear Instruments and Methods in Physics Research B 268 (2010) 1179–1184
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B
journal homepage: www.elsevier.com/locate/nimb
A multi-radionuclide approach for in situ produced terrestrial cosmogenic nuclides:
10
Be, 26Al, 36Cl and 41Ca from carbonate rocks
S. Merchel a,b,*, L. Benedetti a, D.L. Bourlès a, R. Braucher a, A. Dewald c, T. Faestermann d, R.C. Finkel a,e,f,
G. Korschinek d, J. Masarik g, M. Poutivtsev d, P. Rochette a, G. Rugel d, K.-O. Zell c
a
CEREGE, CNRS-IRD-Université Aix-Marseille, F-13545 Aix-en-Provence, France
Forschungszentrum Dresden-Rossendorf, D-01314 Dresden, Germany
Institut für Kernphysik, Universität zu Köln, D-50937 Köln, Germany
d
Technische Universität München, D-85748 Garching, Germany
e
CAMS, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA
f
University of California, Berkeley, CA 94720, USA
g
Department of Nuclear Physics, Komensky University, SK-84215 Bratislava, Slovakia
b
c
a r t i c l e
i n f o
Article history:
Available online 12 October 2009
Keywords:
Accelerator mass spectrometry
Terrestrial cosmogenic nuclides (TCN)
Cosmogenic nuclide exposure dating
a b s t r a c t
In contrary to siliceous environments, there is a severe lack of cosmogenic nuclides, that can be used for
in situ dating of calcareous environments. Thus, we have investigated other nuclides than 36Cl as possible
dating tools by cross-calibration. Cosmogenic 10Be is highly contaminated with atmospheric 10Be and
cannot be removed quantitatively, even by a very sophisticated chemical cleaning procedure. Only working on clay-free calcite provides correct 10Be data, giving a 2.7 times higher production rate of 10Be from
CaCO3 than from SiO2. Though, the production rate of 26Al is only 4.6% (CaCO3 relative to SiO2), 26Al can
be easily determined in calcite, as the low intrinsic 27Al concentration yields to nearly as high 26Al/27Al as
within corresponding quartz. The measurement of 41Ca, mainly produced via thermal-neutron-capture, is
hindered by very low 41Ca/Ca:<5 10 15.
Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction
In situ produced cosmogenic nuclides have proved to be valuable tools for quantifying Earth’s surface processes. Here, the
work-horses are 10Be and 26Al in quartz-rich minerals, and 36Cl in
Ca- or K-rich minerals. Several attempts to find new matrix-product-pairs have been yet performed, especially with respect to
broaden the time-scale to both more ancient [1] and more recent
(historic) times.
The ability to analyse 10Be in carbonate rocks would have several advantages: extending the timescale and allowing burial dating in calcareous environments, and as the 10Be production rate
from C is higher than from O [2], better analytical precision, reduced sample mass, enlarged resolution, and investigation of samples with shorter exposure. Beryllium-10 determinations from
calcite samples had been proven to be quite challenging [3], because atmosphere-produced 10Be is absorbed on ubiquitous clay
minerals and cannot be removed by simple leaching techniques.
We have tested new 10Be and clay decontamination schemes, but
we are also focussing on other cosmogenic nuclides from this environment, namely 26Al and 41Ca. Of course, we are aware that both
nuclides have some disadvantages over 10Be or 36Cl, i.e. lower 26Al
production rate or being a pure (n,c)-product yielding low 41Ca/Ca
ratios.
Measuring at least two nuclides from the same sample allows
us to better constrain the thermal neutron field, perform burial
dating and/or reconstruct more precisely erosion rates and irradiation histories. We have, thus, determined 10Be, 26Al, 36Cl, and
41
Ca in accompanied calcite- and quartz-rich samples from Antarctica and Southern France. Ratios between different nuclides from
the same matrix (CaCO3) and ratios of 10Be or 26Al from CaCO3
and SiO2 can be compared with pure physical model calculations
giving us experimental terrestrial production rates for 10Be, 26Al,
and 41Ca from Ca and CaCO3.
2. Experimental
2.1. Samples
* Corresponding author. Address: Forschungszentrum Dresden-Rossendorf,
D-01314 Dresden, Germany. Tel.: +49 351 260 2802; fax: +49 351 260 12802.
E-mail address: [email protected] (S. Merchel).
0168-583X/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.nimb.2009.10.128
As our work is more like a feasibility study, we have looked for
samples with high radionuclide concentrations and coexisting
calcite- and quartz-rich lithologies. Due to generally low erosion
1180
S. Merchel et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 1179–1184
rates and long exposure, samples from Antarctica should be most
suitable to start with. Surface samples have been taken during the
2003 PNRA meteorite recovery expedition [4] from the regions
Dry Valleys (‘‘DV3”: 77.4°S; 163.7°E; altitude: 100 m) and Johannessen Nunataks (‘‘Joh”: 72.8°S; 161.2°E; altitude: 1902 m). The
DV-site (Marble point) corresponds to mixed outcrops of nearly
pure marble- and quartz-rich gneisses on a flat surface. The
Joh-site was sampled on the nunatak crest made of Jurassic Ferrar
dolerite, overlain by the Cretaceous Beacon Sandstones. The wellcrystallized calcite sample comes from a large nodule (former gaz
bubble) in the dolerite. The Joh-calcite sample holds also a magnetic fraction (basaltic fragments) and regions of pure-white
well-crystalline calcite. Based on recent studies on the neighbouring Frontier Mountain [5], it is assumed that the Johannessen
Nunatak crest remained continuously exposed to cosmic rays for
several million years, although temporary thin snow cover is not
excluded. The lithologies being more friable than the Frontier
Mountain granite, a more significant erosion is also expected.
The DV-site, due to its low altitude, may have a more complex
history of thick snow/ice cover, eventual immersion below sea level (prior to isostatic rebound) and significant erosion. It was
unfortunately not possible to recover calcite samples from the
higher altitude Dry Valley site with previously determined exposure history.
Additionally, we have begun to investigate three rock samples exposed at different shielding depths and taken from a core
drilled in 2005 at ‘‘La Ciotat”, France (‘‘Ciot1”@0.085 m,
‘‘Ciot3”@0.965 m, ‘‘Ciot9”@3.53 m: 43.179°N; 5.576°E; altitude:
310 m). All samples contained mixtures of carbonates and siliceous
conglomerates.
2.2. Sample preparation
2.2.1. Pretreatment
We have made considerable effort to test different chemical
cleaning procedures reducing the abundance of clay minerals and
associated atmosphere-produced 10Be in the calcite fraction of the
antarctic samples. Thus, before starting total dissolution for 10Be,
26
Al, 36Cl, and 41Ca isolation, samples have been differently pretreated (see Figs. 1 and 2). The Joh-sample has been roughly freed
from bigger well-crystallized zones. All samples (Joh, DV3, Ciot)
have been sieved into 250–500 and 500–1000 lm fractions. The
magnetic fraction of Joh (500–1000 lm) has been discarded (71%)
by a Frantz magnetic separator. Well-crystallized grains have been
hand-picked from the >1000 lm-fraction of Joh and further crushed
to build a special ‘‘well-crystallized fraction” (Joh-WC-CaCO3).
All 500–1000 lm fractions were washed 10–15 times in H2O
and the suspended particles were discarded each time. Further
purification of the DV3-sample, as seen in Fig. 2, was less complicated as for the Joh-sample due to the absence of magnetic and
well-crystallized fractions. A subset of the Antarctic CaCO3-fractions have been specially treated with the goal of reducing the
amount of associated clay minerals (Figs. 1 and 2). The idea came
from a paper published by Dong et al. [6] using the procedure on
marine calcite, aragonite, and opal. Fortunately, one of the authors
(D. Lal) generously helped us by providing essential details (see
Appendix A).
CIOT samples are not analysed for 10Be and 26Al in calcite, thus,
simple pretreatments to reduce atmospheric 36Cl in calcite and
10
Be in quartz have been applied (see Fig. 3). They are modifications of the work by Stone et al. [7] and Brown et al. [8]. Our main
Joh
well-cryst.
CaCO3
hand-picked
< 250 µm
250-500 µm
500-1000 µm
magnetic
residues >>
10Be & 26 Al
> 1000 µm
nonmagnetic
well-cryst.
CaCO3
hand-picked
bulk
H2O
crushing
3 x 10% HF
§
Joh-SiO2:
BeO & Al2O3
11 g >>
& 26Al
10Be
#
Joh-CaCO3:
BeO & Al2O3
14 g >>
& 41Ca
36Cl
2 x H2O
10% HNO3
*
Joh-CaCO3:
AgCl & CaH2
H2O
34 g >>
& 26Al
10Be
NaOH
H2O
5% HNO3
H2O
10x
(NaO3P)6/H2O2
H2O
#
11 g
(big,clean) >>
10Be & 26Al
#
Joh-WC-CaCO3:
BeO & Al2O3
38 g
(small,dirty) >>
36Cl & 41Ca
2 x H2O
10% HNO3
*
Joh-WC-CaCO3:
AgCl & CaH2
Joh-SpecialCaCO3:
BeO & Al2O3
Fig. 1. Flow-chart presenting the different pretreatments of sample Joh before dissolution. The hand-picked well-crystallized grains have been further separated by size
(bigger/smaller) and optical properties (clear white/dirty brownish) for 10Be & 26Al and 36Cl & 41Ca analyses, respectively. Dissolution and further treatment differently for §, ,
and # (see text).
1181
S. Merchel et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 1179–1184
DV3
250-500 µm
< 250 µm
H2O
residues >>
10Be & 26 Al
> 1000 µm
500-1000 µm
H2O
“100%”
HNO3
11 g >>
& 26Al
15 g >>
& 41Ca
10Be
36Cl
55 g >>
& 26Al
10Be
3 x 10% HF
#
§
2 x H2O
10% HNO3
DV3-CaCO3:
BeO & Al2O3
DV3-SiO2:
BeO & Al2O3
*
DV3-CaCO3:
AgCl & CaH2
NaOH
H2O
5% HNO3
H2O
10x
(NaO3P)6/H2O2
H2O
#
DV3-SpecialCaCO3:
BeO & Al2O3
Fig. 2. Flow-chart presenting pretreatments of sample DV3 before dissolution. The 250–500 lm fraction is only used to produce more quartz-rich-residue for 10Be and 26Al
analyses from SiO2. Dissolution and further treatment differently for §, , and # (see text).
Ciot
< 250 µm
residues >>
10Be & 26 Al
250-500 µm
500-1000 µm
H2O
H2O
>>
4x
HCl/H2SiF6
3 x 10% HF
§
> 1000 µm
36Cl
36Cl
>>
& 41Ca
2 x H2O
2 x H2O
10% HNO3
10% HNO3
*
*
Ciot-S-CaCO3:
AgCl
Ciot-B-CaCO3:
AgCl & CaH2
Ciot-SiO2:
BeO & Al2O3
Fig. 3. Flow-chart presenting pretreatments of samples Ciot1, Ciot3 and Ciot9 before dissolution. Dissolution and further treatment differently for § and (see text).
interest in analysing these samples is the cross-calibration of
vs. 36Cl from CaCO3.
41
Ca
2.2.2. Chemical separation of Al, Be, Ca, and Cl
Calcite fractions for analysing 10Be and 26Al have been spiked
with 0.3 mg 9Be [9] and 0.9 mg 27Al and dissolved in acetic acid
(25%), i.e. marked as # in Figs. 1 and 2. Quartz-rich samples, decon-
taminated from atmospheric 10Be, have been dissolved in HF (48%)
after carrier addition (0.3 mg 9Be, 0.9 mg 27Al), i.e. marked as § in
Figs. 1–3. After separation of the residues by centrifugation, solutions (either CH3COOH or HF) have been evaporated to dryness
and redissolved in HCl. A chemical separation protocol modified
from that originally developed for small meteorite samples (<1 g)
[10] comprising numerous hydroxide precipitations and anion
1182
S. Merchel et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 1179–1184
and cation exchange separation (both with HCl), was used to separate Be and Al from the matrix, the isobar, and other major and
trace elements like Fe and Al. Intrinsic aluminium has been determined in aliquots by inductively coupled plasma optical emission
(ICP-OES). Al2O3 targets were mixed with Ag (Aldrich, 325 mesh,
99.99+%) and BeO targets with Nb (Aldrich, 325 mesh, 99.8%)
powder and pressed with a steel pin into Cu cathodes.
Calcite fractions for analysing 36Cl and 41Ca (i.e. marked as in
Figs. 1–3) have been spiked with 1.5 mg of a 35Cl-enriched carrier
to allow simultaneous natCl determination by isotope dilution AMS
(ID-AMS). The material, cooled in an ice-bath, has been totally dissolved by very slow addition of HNO3 (2 mol/L). Chlorine was further separated and cleaned from sulphur by repeated precipitation
as AgCl and BaSO4. AgCl targets have been pressed into stainless
steel cathodes with a AgBr backing. As Ca is the main target element for 36Cl-production in calcite, it has to be determined in aliquots by any analytical technique. As we have experienced [11]
severe difficulties in getting reliable and concordant calcium data
from ICP-OES (two labs), prompt-gamma (PGAA) and instrumental
neutron activation analyses (INAA), we are still looking for the
method of choice. However, as 41Ca and 36Cl are both normalized
to the same target element, results are not at all needed for this
intercomparison. And as 10Be will be mainly compared relatively,
the assumption of 40% Ca in CaCO3 (validated by gravimetry as
CaC2O4) will give us reasonably adequate data.
For producing a 41Ca-target, a small aliquot of the calcite–HNO3solution has been evaporated to dryness and redissolved in HCl.
Hydroxides have been precipitated by addition of NH3aq and discarded. Calcium oxalate was formed due to addition of a solution
of (NH4)2C2O4 (saturated, pH 9) followed by transformation to CaCO3
by heating to 500 °C. As preliminary results have shown that the
background by measuring CaF2 is too high, we decided to apply
the much more sophisticated two-step chemical preparation as
CaH2-targets: Ca-reduction by heating with Zr-powder at 1500 °C
followed by hydrogenation with H2 at 500 °C. For this purpose, we
slightly modified the existing instruments at the nuclear physics
department at the University of Cologne, which are usually used
for preparing 48CaH2-targets [12,13]. The CaH2-targets have been
mixed with Ag powder (Aldrich, 325 mesh, 99.99+%), pressed in
Cu cathodes, and kept under inert gas until the AMS measurement.
2.3. AMS measurements
Beryllium-10 and 26Al have been routinely measured at the 5 MV
ASTER AMS facility [14]. Normalization has been performed
against NIST SRM 4325 with the certified value of (10Be/9Be) =
(2.68 ± 0.14)x10 11 and the ASTER in-house standard SM-Al-11
with (26Al/27Al) = (7.401 ± 0.064) 10 12, which has been crosscalibrated against the primary standards certified by a round-robin
exercise [15].
Chlorine-36 has been routinely measured at the 10 MV CAMSfacility at LLNL [16]. Results are based on a calibration with
KN500 material.
As we could only perform a single run for 41Ca measurements
at the Maier-Leibnitz-Laboratory, all results can be regarded as
preliminary. A modification of the set-up described by Wallner and
coworkers [17] has been used. Ions have been extracted as CaH3with currents as high as 1.5 lA. The tandem running at 12 MV accelerated Ca10+ to 131.25 MeV. Results are calculated based on the TUM
in-house standard, which had been produced by thermal neutron
activation, with a nominal value of (1.95 ± 0.20) 10 11 41Ca/Ca.
2.4. Model calculations
For comparison with the measurements, we simulated production rates in terrestrial rocks with a computer model described in
detail in [18]. We recount here only its main features. For the primary particle flux, protons with energies between 10 MeV and
100 GeV were considered. The primary flux was set equal to the
long-term average value of 4.56 p cm 2 s 1.
The propagation of primary particles and the production of secondary protons and neutrons by them was calculated. All particles
were followed through the atmosphere and down into the lithosphere. For this purpose, the atmosphere was modelled as a spherical shell with an inner radius of 6378 km and a thickness of
100 km. Atmospheric composition (by weight) was assumed to
be: 75.5% N, 23.2% O, and 1.3% Ar. The total thickness of the atmosphere was taken to be 1033 g cm 2. For the Earth’s surface, we assumed the real composition of investigated samples. The statistical
errors of the calculations, for pure-spallation nuclides, are approximately 5–8%. Having calculated the particle fluxes, the production
rates of particular nuclides were calculated by integrating over
energy the product of these fluxes and cross sections for the nuclear reactions that produce the given nuclide.
3. Results and discussion
Measured isotope ratios and calculated radionuclide concentrations are given in Table 1. Remarkably, the reproducibility for different size fractions (250–500 and 500–1000 lm) for 36Cl (calcite)
of all Ciot-samples is excellent. In contrary to the expectations, antarctic sample DV3 did not contain high radionuclide concentration,
but the Joh-sample did. The upper surface sample Ciot1 seems to
have quite low 36Cl (calcite), but ‘‘normal” high 10Be (quartz), especially if compared to all other samples from the core [19]. Unfortunately, the 26Al-target did not produce adequate current for
quantification to solve this problem. But the 41Ca-value of Ciot1
seems to be comparable to the corresponding low 36Cl.
3.1.
10
Be in calcite and quartz
As shown in Fig. 4, the special cleaning procedures removes a
large part of atmosphere-produced 10Be from both CaCO3 Antarctic
samples. A simple water-leaching procedure as suggested by Braucher et al. [20] would overestimate the cosmogenic 10Be by a factor
of 2.1–2.7 (compared to the special cleaning procedure). However,
the 10Be concentration in the well-crystallized material of Joh (JohWC) is even lower than in the specially treated material of Joh,
leaving the only explanation that the procedure got rid of a good
portion of atmospheric 10Be, but not quantitatively. If assuming
that the well-crystallized material does not contain any atmospheric 10Be at all, we can calculate the production rate for 10Be
in CaCO3 relative to SiO2 to 2.7. This value is in the same range
as those from experimental work of Granger et al. [21] and our
pure physical model calculations, i.e. 2.65, based on [22]. However,
it is incomparable with experimental data from Braucher et al.
[20], which is at the same level as our simple water-leached sample value. If comparing 10Be from Joh-calcite to corresponding 36Clconcentrations, we get experimental values of production rates
(PR), PR36Cl(atoms gCa 1)/PR10Be(atoms g 1) = 1.42–1.48, which
are in very good agreement with a value of 1.49 [3], we determined
earlier working on well-crystallized material from a medieval rockfall sample from the Southern Alps.
3.2.
26
Al in calcite and quartz
Despite the fact that the 26Al production rate from CaCO3 is
much lower than from corresponding SiO2, measured 26Al/27Al ratios are in the same order due to the lower intrinsic 27Al concentration: 8–31 lg/g in CaCO3 compared to about 800 lg/g in
corresponding SiO2. So, if working on total masses of 5–10 g, the
1183
S. Merchel et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 1179–1184
Table 1
Measured isotope ratios (non-blank-corrected) and calculated radionuclide concentrations (blank-corrected). Concentrations of 36Cl are calculated under the assumption of 40%
Ca in the sample. B correspond to size fraction 500–1000 lm and S to 250–500 lm.
Joh-CaCO3
Joh-Special-CaCO3
Joh-WC-CaCO3
Joh-SiO2
DV3-CaCO3
DV3-Special-CaCO3
DV3-SiO2
Ciot1-B-CaCO3
Ciot1-S-CaCO3
Ciot1-SiO2
Ciot3-B-CaCO3
Ciot3-S-CaCO3
Ciot3-SiO2
Ciot9-B-CaCO3
Ciot9-S-CaCO3
Ciot9-SiO2
a
b
10
Be/9Be
[10 13]
10
Be
[105 atoms/g]
26
Al/27Al
[10 13]
26
Al
[105 atoms/g]
36
Cl/35Cl
[10 13]
36
Cl
[105 atoms/g Ca]
41
Ca/Ca
[10 15]
41
Ca
[107 atoms/g Ca]
66.7 ± 5.2
49.9 ± 0.4
40.1 ± 0.4
19.0 ± 0.7
1.15 ± 0.03
0.69 ± 0.02
286 ± 22
107.6 ± 0.9
73.8 ± 0.7
27.0 ± 0.9
2.28 ± 0.06
1.10 ± 0.03
0.41 ± 0.41a
1.65 ± 0.12
2.58 ± 0.26
5.56 ± 0.18
0.10 ± 0.03
0.15 ± 0.04
2.11 ± 2.11a
4.55 ± 0.34
5.05 ± 0.51
105 ± 3
0.13 ± 0.04
0.21 ± 0.05
9.17 ± 0.10
129 ± 1
2.4 (+1.9–1.6)
3.7 (+2.8–2.4)
50.6 ± 0.6
124 ± 2
2.4 (+1.4–1.1)
3.6 (+2.1–1.7)
0.83 ± 0.02
6.05 ± 0.13
0.8 (+0.9–0.5)
1.2 (+1.3–0.8)
2.01 ± 0.05
1.90 ± 0.08
2.86 ± 0.07
2.82 ± 0.12
1.5 (+1.2–1.0)
2.3 (+1.8–1.5)
6.05 ± 0.17
3.28 ± 0.08
8.91 ± 0.25
8.54 ± 0.22
2.8 (+1.9–1.2)
4.2 (+2.9–1.8)
4.32 ± 0.10
3.85 ± 0.09
3.82 ± 0.09
3.82 ± 0.09
0.4 (+0.7–0.3)
0.6 (+1.1–0.4)
b
b
0.060 ± 0.007
b
0.85 ± 0.07
0.36 ± 0.03
1.17 ± 0.17
0.37 ± 0.03
0.178 ± 0.013
0.58 ± 0.05
0.23 ± 0.04
2.73 ± 0.25
0.97 ± 0.16
Sample did not last long enough for having good statistics (1 count).
Problems during chemical processing, current too low for quantification.
7
3x10
CaCO3
SiO2
7
2x10
special
Be [atoms g-1]
7
1x10
5
10
2.0x10
special
5
1.5x10
5
1.0x10
4
5.0x10
DV3
Joh
JohWC
Fig. 4. Beryllium-10 results from differently treated calcite fractions of Antarctic
samples DV3 and Joh with corresponding 10Be data from quartz for the Joh-sample.
The 10Be data for DV3-quartz is doubtful due to problems during chemical
processing resulting in a 9Be current too low for quantification, and, thus, not
shown.
addition of a 27Al carrier is needed to get a reasonable amount of
Al2O3 for an AMS-target. This can be advantageous as there is no
need for very precise 27Al determination, which sometimes can
be problematic. The results on differently treated CaCO3 samples
are very reproducible showing that there is – as expected – not
any influence by atmosphere-produced 26Al. The experimental
production rate of 26Al from CaCO3, calculated from Joh-values,
leaving aside the value for Joh-CaCO3 due to its high uncertainty,
is 4.3–4.8% of the one from SiO2.
This value is in excellent agreement with our pure physical
model calculation value of 4.3%.
3.3.
36
Cl and
41
Ca in calcite
All intrinsic Cl-concentrations determined by ID-AMS are low
(2.2–23.2 lg/g) for all investigated samples, thus, 36Cl is mainly
produced by spallation on Ca and can be used as a good estimate
for normalization of 41Ca production rates. Of course, 41Ca counting
statistics are poor: Measured 41Ca/Ca values, as shown in Table 1,
are based on total counts of 1–5. Nevertheless, the reproducibility
(Joh and Joh-WC) is excellent. All our data is in the range of or lower than the already published one from rock samples, i.e. lower
than the six surface samples (3–63 10 15) of Henning et al.
[23] and Kutschera et al. [24] and lower than the surface and
strongly shielded sample at 11 m depth (3.4–7.6 10 15) of Middleton et al. [25]. As 41Ca in CaCO3 is mainly a thermal neutroncapture product, the production (and its calculation) is highly
depending on the chemical composition of the sample especially
the water-content. Also, external factors like snow cover and vegetation are mostly unknown over long time-scales and influence
the thermal neutron flux, thus having a high impact on the production rate. Our newest calculations for DV3 – taking into account
variable factors – yield values of 40–90% (DV3) and 20–60% (Ciot),
which are far from being consistent with the experimental values
of 36Cl/41Ca of 1–6% (DV3 & Ciot). However, calculations for Joh
and Joh-WC, giving values of 30–50%, are in good agreement with
experimental values 34–36%. So far, it seems debatable, if we need
better calculations or measurements.
4. Conclusions and outlook
The determination of cosmogenic 10Be in calcite suffers from
the inability to remove quantitatively atmospheric 10Be. The presented pretreatment chemistry is well on the way, but working
on clay-free material, as e.g. well-crystallized material, right from
the beginning seems to be still the best choice. On the contrary
measurements of 26Al from calcite might be easier to perform, as
26
Al/27Al ratios can be as high as those form SiO2. Furthermore,
the calculation of a pure high-energy spallation product from a single target element might help to overcome problems involved with
36
Cl data interpretation, especially on natCl-rich samples. The low
41
Ca/Ca ratios make it very unlikely that 41Ca could be generally
used for in situ dating of calcareous environments, especially as
there is little hope that background level for CaF2-targets will improve, thus asking for very sophisticated and time-consuming
CaH2-target preparation and handling. An alternative, we would
like to explore in the very near future, is the determination of
41
Ca in Ti- and Fe-rich minerals. The agreement between calculated
and experimental production rates of pure-spallation nuclides is
impressive. However, there might be potential to improve either
calculations or measurements of thermal neutron build radionuclides like 36Cl and 41Ca.
1184
S. Merchel et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 1179–1184
Encouraged by the CRONUS-EU network, we have tried to expand matrix-product-pair combinations for use of terrestrial cosmogenic nuclides with diverse success. Some might be promising
to transform into work-horses in the future.
Acknowledgments
We appreciate the enormous help of Devendra Lal (UCSD) by
giving us unpublished details on his cleaning procedure. It is truly
his development to reduce atmosphere-produced 10Be from calcite.
The 10Be and 26Al measurements could not have been undertaken
without Maurice Arnold and Georges Aumaître at CEREGE. We also
like to thank J. Lachner and I. Dillmann (TU Munich) for their tremendous support during 41Ca measurements. S. Nardon (ENI
S.p.A., Milan) and J. Borgomano (U Marseille) provided the core
samples from ‘‘La Ciotat”. We are deeply indebted to the PNRA
and IPEV (Italian and French Antarctic programs) for funding the
mission that allowed the present sampling during the 2003–2004
PNRA expedition, which was lead by L. Folco from Museo Nazionale Antartide (Siena). This work was partially funded within the
framework of CRONUS-EU (Marie-Curie Action 6th framework programme; Contract No. 511927). The French AMS Facility ASTER is
supported by INSU/CNRS, the French Ministry of Research and
Higher Education, IRD, and CEA.
Appendix A. Reducing clay minerals and associated
atmospheric 10Be
This procedure is adapted from information of Dong et al. [6] and
D. Lal [D. Lal, pers.com., 2006]: A 1000 ml-HDPE-bottle is filled with
thoroughly washed grains of size 500–1000 lm (34–88 g) and a basic solution of 20 g NaOH in 500 ml H2O. After 3.5 h-exposure at
50 °C part of organic matter can be removed. The grains are five
times washed with H2O, and 5% of the total mass is dissolved by
slowly addition of dilute (1%) HNO3. The grains are again washed
five times with H2O. Further disaggregation of particles and oxidation of organics is performed by 10 times repetition of the following
steps: treatment with a mixture of 220 ml of sodium hexametaphosphate (56 g (NaO3P)n/1000 g) and 22 ml H2O2 (30%) for
15 min in an ultrasonic bath, three times washing with H2O.
References
[1] J.M. Schaefer, T. Faestermann, G.F. Herzog, K. Knie, G. Korschinek, J. Masarik, A.
Meier, M. Poutivtsev, G. Rugel, C. Schlüchter, F. Serifiddin, G. Winckler, Earth
Planet. Sci. Lett. 251 (2006) 334.
[2] I. Leya, H.-J. Lange, S. Neumann, R. Wieler, R. Michel, Meteorit. Planet. Sci. 33
(2008) 259.
[3] S. Merchel, R. Braucher, L. Benedetti, D. Bourlès, O. Grauby, Quat. Geochronol. 3
(2008) 299.
[4] L. Folco, N. Perchiazzi, P. Rochette, Terra Antarctica Reports 14 (2008)
255.
[5] K.W. Welten, L. Folco, K. Nishiizumi, M.W. Caffee, A. Grimberg, M.M.M. Meier,
F. Kober, Earth Planet. Sci. Lett. 270 (2008) 308.
[6] W. Dong, D. Lal, B. Ransom, W. Berger, M.W. Caffee, in: Proc. Indian Acad. Sci.,
Earth Planet. Sci. 110 (2001) 95.
[7] J.O. Stone, G.L. Allan, L.K. Fifield, R.G. Cresswell, Geochim. Cosmochim. Acta 60
(1996) 679.
[8] E.T. Brown, J.M. Edmond, G.M. Raisbeck, F. Yiou, M.D. Kurz, E.J. Brook, Geochim.
Cosmochim. Acta 55 (1991) 2269.
[9] S. Merchel, M. Arnold, G. Aumaître, L. Benedetti, D.L. Bourlès, R. Braucher, V.
Alfimov, S.P.H.T. Freeman, P. Steier, A. Wallner, Nucl. Instr. Meth. B 266 (2008)
4921.
[10] S. Merchel, U. Herpers, Radiochim. Acta 84 (1999) 215.
[11] S. Merchel, M. Bichler, J.H. Sterba, On the reliability of target element data for
cosmogenic nuclide exposure dating, in: Seminar Activation Analysis and
Gamma-Spectroscopy (SAAGAS 22) Vienna, Austria, 25–27 February, 2009.
[12] G. Korschinek, W. Kutschera, Nucl. Instr. Meth. 144 (1977) 343.
[13] H.J. Maier, W. Kutschera, Nucl. Instr. Meth. 167 (1979) 91.
[14] M.G. Klein, A. Gottdang, D.J.W. Mous, D.L. Bourlès, M. Arnold, B. Hamelin, G.
Aumaître, R. Braucher, S. Merchel, F. Chauvet, Nucl. Instr. Meth. B 266 (2008)
1828.
[15] S. Merchel, W. Bremser, Nucl. Instr. Meth. B 223–224 (2004) 393.
[16] J.R. Southon, J.S. Vogel, S.E. Trumbore, J.C. Davis, M.L. Roberts, M.W. Caffee, R.C.
Finkel, I.D. Proctor, D.W. Heikkinen, A.J. Berno, R.S. Hornady, Radiocarbon 34
(1992) 473.
[17] A. Wallner, A. Arazi, T. Faestermann, K. Knie, G. Korschinek, H.J. Maier, N.
Nakamura, W. Rühm, G. Rugel, Nucl. Instr. Meth. B 223–224 (2004) 759.
[18] J. Masarik, J. Beer, J. Geophys. Res. 104 (1999) 12099.
[19] R. Braucher, P. Del Castillo, L. Siame, S. Merchel, L. Benedetti, R.C. Finkel, D.
Nyvlt, D.L.G. Bourlès, in: 11th International Conference on Accelerator Mass
Spectrometry, 2008, p. 67.
[20] R. Braucher, L. Benedetti, D.L. Bourlès, E.T. Brown, D. Chardon, Geochim.
Cosmochim. Acta 69 (2005) 1473.
[21] D.E. Granger, A.J. Cyr, E. Rauh, in: Tenth International Conference on
Accelerator Mass Spectrometry, 2005, p. 1704.
[22] J. Masarik, K.J. Kim, R.C. Reedy, Nucl. Instr. Meth. B 259 (2007) 642.
[23] W. Henning, W.A. Bell, P.J. Billquist, B.G. Glagola, W. Kutschera, Z. Liu, H.F.
Lucas, M. Paul, K.E. Rehm, J.L. Yntema, Science 236 (1987) 725.
[24] W. Kutschera, I. Ahmad, P.J. Billquist, B.G. Glagola, K. Furer, R.C. Pardo, M.
Paul, K.E. Rehm, P.J. Slota Jr., R.E. Taylor, J.L. Yntema, Radiocarbon 31
(1989) 311.
[25] R. Middleton, D. Fink, J. Klein, P. Sharma, Radiocarbon 30 (1989) 305.