Content uploaded by Sergiy Dubchak
Author content
All content in this area was uploaded by Sergiy Dubchak on Oct 04, 2021
Content may be subject to copyright.
Detection of the Fission Product Palladium-107 in a Pond Sediment
Sample from Chernobyl
Anica Weller, Tim Ramaker, Felix Stäger, Tobias Blenke, Manuel Raiwa, Ihor
Chyzhevskyi, Serhii Kirieiev, Sergiy Dubchak, and Georg Steinhauser
Environ. Sci. Technol. Lett. 2021, 8, 656−661
ABSTRACT: Radiometric or mass spectrometric analysis of the long-lived
fission product 107Pd is notoriously difficult. We developed and optimized a chemical
separation protocol for minute amounts of radiopalladium with a subsequent
measurement by inductively coupled plasma triple quadrupole mass spectrometry
with propane as the collision gas. This allows for detection limits of <2 ng of 107Pd/kg,
which makes the method suitable for environmental samples with low levels of 107Pd.
For testing of this method, a sample of sediment from the Chernobyl cooling pond was
analyzed. Indeed, it could be shown that the cooling pond sediment exhibits a uniquely
increased 107Pd/105Pd ratio (0.08 ± 0.02), thus strongly indicating detectable levels of
107Pd using this method.
INTRODUCTION
Long-lived radionuclides are of special importance for the long-term safety
assessments of nuclear waste repositories. Given their high radiotoxicity, most research
focused on α-emitting actinides. However, also long-lived β−-emitting fission products
such as 79Se, 99Tc, 107Pd, 129I, and 135Cs as well as activation products (59Ni, 93Zr, and
41Ca) can also be dose relevant.1 Because most of these radionuclides are pure
β−emitters, any radiometric detection reaches its limit due to the nuclides’ long half-
lives and the need for chemical separation of trace amounts of these elements.1−3
Despite this analytical shortcoming, modeling data illustrate the importance of long-
lived fission products. Calculations for spent nuclear fuel of a boiling water reactor
(BWR) with UO2 fuel and burnup of 50 GWd/t show an activity in the range from 109
to ∼1012 Bq per metric ton of heavy metal after 1000 years of cooling.4 The fission
product 107Pd falls in the lower range with 7.2 × 109 Bq t−1. It is a β emitter with a low
β end point energy of 34 keV and a half-life of 6.5(3) × 106 years.5 With increasing
burnup of commercial nuclear fuel, the onset of 107Pd increases due to the breeding and
subsequent fission of 239Pu in the fuel. The yield of 107Pd through thermal neutron
fission of 235U is relatively low (0.14%), but it is much higher for 239Pu (3.2%).6 Buck
et al. reported that fission-produced radiopalladium accumulates in spent nuclear fuel
in Ag−Pd-rich phases.7
Palladium-107 is a radionuclide of potential concern in final repositories. In any
case, fission-produced Pd may become a potentially interesting source of industrial Pd
due to rising Pd prices. Spent fuel consists of ≤0.1% Pd per ton of heavy metal.4 The
use of such “artificial Pd” is presently a topic of discussion with respect to its
application in catalytic products.8−10 However, in such artificial Pd, 107Pd would have
to be considered, as well. Once mixed into the Pd recycling stream, the radioisotope
107Pd could never again be removed from the stream by chemical means.
The examples presented above obviate the need for sensitive analytical methods
for 107Pd. However, radioanalytical methods face limitations when it comes to
analyzing 107Pd. The low β energy of 107Pd results in a weak radiation detector response
and a low detection efficiency as the analytical signal is partly obscured from cosmic
radiation. Liquid scintillation counting (LSC) suffers from low count rates due to
absorption effects in solution.11,12 Additionally, the low specific activity of pure 107Pd
(1.9 × 107 Bq g−1) drastically increases the detection limits. Lastly, all β− measurement
techniques are susceptible to interference from other β− emitters. Stable Pd can be
analyzed by mass spectrometry (MS) with high sensitivity (detection limits with a
direct injection nebulizer or flow injection in the lower picograms per gram range).13,14
This detection limit corresponds to approximately 102 mBq (g of pure 107Pd)−1. Only a
few analytical studies have specifically targeted 107Pd so far. This includes meteorites,
where 107Pd was measured indirectly from 107Ag/109Ag isotopic ratios via either
TIMS15,16 or MC-ICP-MS17,18 under the assumption that the cosmogonic 107Pd had
completely decayed to its daughter nuclide, 107Ag. Due to the long half-life of 107Pd,
such indirect measurement is not feasible for terrestrial materials. Andris et al.19 and
Dulansk et al.20 developed chemical separation techniques and measured radioactive
waste for 107Pd by LSC. In both studies, the 107Pd activities were all below their
determined detection limits for radioactive waste samples. Asai et al. first reported the
determination of 239 ng of 107Pd in 1 mg of U in spent nuclear fuel with single
quadrupole inductively coupled plasma mass spectrometry (ICP-MS),21 which is the
same order of magnitude like the calculated value for spent nuclear UO2 fuel.4 Further
MS studies targeted 107Pd in Zircaloy fuel cladding22 and radioactive waste from the
Fukushima Daiichi Nuclear Power Plant (FDNPP).23 These studies dealt with higher
concentrations of 107Pd, which would typically be found in spent nuclear fuel but hardly
in the environment. Due to several types of interference as well as lack of sensitivity,
some of these methods cannot be used for environmental samples.
The objective of this study was the development of an analytical protocol for the
separation of minute amounts of radiopalladium in environmental samples followed by
MS measurement. The method should subsequently be tested on a sample possibly
contaminated with 107Pd. Because no certified reference materials are available for
107Pd, we extracted fuel particles from the sediment of the Chernobyl cooling pond and
tested them for the detectable occurrence of environmental 107Pd. The Chernobyl
cooling pond is notorious for its contamination with fuel particles,24 which contain,
among other radionuclides, 107Pd. Thirty-five years after the Chernobyl nuclear
accident (April 26, 1986), to the best of our knowledge, the detection of low 107Pd
concentrations has never been successful in a Chernobyl sample, let alone in any other
environmental sample. This motivated us to develop a method with enhanced
sensitivity for this difficult-to-measure radionuclide. The method included a multistep
chemical separation with subsequent triple quadrupole ICP-MS (ICPQQQ-MS)
measurement in reaction mode with propane/He gas. We aimed at detection limits in
the range of microbecquerels per kilogram or nanograms per kilogram. For quality
assurance purposes, we monitored the major types of MS interference, which are silver
isotopes, zirconium and yttrium oxides, and palladium hydride species, throughout the
analytical procedure.