Journal of Radioanalytical and Nuclear Chemistry

, 282:651

Fission yield measurements by inductively coupled plasma mass-spectrometry


    • Idaho National Laboratory
  • Bruce Hilton
    • Idaho National Laboratory
  • Jeffrey Giglio
    • Idaho National Laboratory
  • Daniel Cummings
    • Idaho National Laboratory

DOI: 10.1007/s10967-009-0209-1

Cite this article as:
Glagolenko, I., Hilton, B., Giglio, J. et al. J Radioanal Nucl Chem (2009) 282: 651. doi:10.1007/s10967-009-0209-1


Correct prediction of the fission products inventory in irradiated nuclear fuels is essential for accurate estimation of fuel burnup, establishing proper requirements for spent fuel transportation and storage, materials accountability and nuclear forensics. Such prediction is impossible without accurate knowledge of neutron induced fission yields. The uncertainty of the fission yields reported in the ENDF/B-VII.0 library is not uniform across all of the data and much of the improvement is desired for certain fissioning isotopes and fission products. We discuss our measurements of cumulative fission yields in nuclear fuels irradiated in thermal and fast reactor spectra using Inductively Coupled Plasma Mass Spectrometry.


Cumulative neutron induced fission yieldsFission productsICP-MSMass-spectrometry


Fission products play essential role in many nuclear reactor and fuel cycle related applications. The most important ones are reactivity and criticality calculations, safety related aspects, determination of fuel burnup, transportation and storage of nuclear materials, benchmarking of reactor physics codes and nuclear data validation, materials accountability and nuclear forensics analysis. Different applications require different levels of accuracy for the fission product inventory.

As a general rule, the quantity of any fission product isotope released in fission is proportional to its neutron induced fission yield:
$$ {\text {\#atoms}}\,of\,{\text {FP}}_{i} = \,\sum\limits_{j,k} {\left[ {({\text {\# fissions}})_{j,k} \times ({\text {fission\,yield}})_{i,j,k} } \right]} $$
where i is fission product isotope, j is fissioning isotope and k is the neutron energy group. Consequently, accurate knowledge of fission product inventory translates into accurate knowledge of fission yields.

The source of the neutron induced fission yields (both direct and cumulative) in the US is Evaluated Nuclear Data File, ENDF/B-VII.0, with the latest version released in 2006 [1]. In it, the best known fission yield values and corresponding one sigma errors are reported for each pair of fissioning isotope and fission product using the internationally adopted format (ENDF-6). The “best” values are calculated based on the independent measurement values that were weighted according to their reported error, with less weight given to the data with larger errors. The majority of the measurements were done in the late 1930s—early 1990s, which were evaluated and compiled by England and Rider into the 1994 report [2]. To the best of our knowledge, no major revisions of the thermal (0.0253 eV) and fast (500 keV) fission yield data have been done in the US since that time.

Our recent examination of the ENDF/B-VII.0 files revealed that the accuracy with which fission yields data are reported is not uniform across the library for various reasons. The yields of certain fission products, like Nd-148 and the yields from certain fissile nuclides, like U-235, were measured more than others. This, in general, resulted in smaller errors of the “best” values because they are averages of a large number of measurements. The “valley” and “wing” isotopes on the typical fission yield curve have higher associated uncertainties due to the lower fission yield values (i.e., higher uncertainties at the lower measurement range, near detection limits) in these regions as compared to the “peak” isotopes values. The accuracy of the analytical chemistry techniques used for the analysis of individual fission yields, for instance, Isotope Dilution Mass Spectrometry (IDMS) versus Radiochemistry, also had some bearing on the overall uncertainty. In addition, higher uncertainties were assigned to those isotopes, whose fission yields could not be measured and were estimated instead. We believe that with the development of the new advanced measurement techniques there exist an opportunity for improvement of the fission yield data, especially for those isotopes with the limited number or complete lack of measurements.

Credit should be given to the extensive fission yield measurement program in thermal and fast reactor spectra that existed in Idaho in 1960–1980s [3], with many of the values incorporated into ENDF. At that time the majority of the non-gaseous fission product isotopes were measured by Isotope Dilution—Thermal Ionization Mass Spectrometry (ID-TIMS), fission gases (like Kr, Xe) by Gas Mass Spectrometry and radioactive isotopes (like Ru-102, Zr-95) by radiochemical methods [4]. Though known for its high accuracy, ID-TIMS is very expensive and time consuming process. Radiochemical methods are associated with higher uncertainties due to ambiguities in the decay constants.

At present we are working on the development of new methods for the measurement of fission products in a variety of fuel matrices irradiated in thermal and fast reactor spectra using the Inductively Coupled Plasma Mass Spectrometry (ICP-MS). The key advantages of the ICP-MS include its ability to perform simultaneous isotopic analysis of multiple elements, less time-consuming sample preparation stage, higher sample throughput capacity and less dose to personnel at an overall cost savings. Multi-Collector ICP-MS can provide similar precision as ID-TIMS method. Our studies used standard ICP-MS, which has a higher uncertainty. Several research groups have already reported their successful results from the analysis of fission products isotopes in irradiated thermal reactor nuclear fuels by ICP-MS [57]. In this paper we focus on the potential application of the ICP-MS as a modern and robust tool for the measurement of the cumulative fission product yields in enriched fuels. These new measurements can add to the overall pool of cumulative fission yields data and potentially result in the refinement and improvement of the old data, especially for those isotopes with a limited number of fission yields measurements.


U-235 thermal reactor fission yield measurements

Six U–Mo metal alloy plates from the RERTR-9 experiment with 44 and 58% U-235 enrichment were selected for measurement of U-235 thermal fission yields. Five of the plates were clad in aluminum and one was clad in zirconium. The plates were irradiated in test position B-11 of the Idaho National Laboratory (INL) thermal Advanced Test Reactor (ATR) as part of the Reduced Enrichment for Research and Test Reactors (RERTR) Program [8]. The peak thermal neutron flux in this position is ~1.1 × 1014 n/cm2 s at 110 MWth power. The midplane segments from irradiated plates had a burnup between 10.6 and 19.6 at.% fission. Based on reactor physics analysis, which was performed using the MCWO method [9] that couples a validated General Monte Carlo N-Particle transport code (MCNP) [10] with a depletion code ORIGEN-2 [11], more than 99% of all fissions in the plates came from U-235 fission.

U-235 fast reactor fission yield measurements

Four U–Zr metal alloy rods with 78% enrichment of U-235 were selected for measurement of U-235 fast fission yields. The rods had stainless steel cladding. They were irradiated in row 5 of the Experimental Breeder Reactor II (EBR-II), which is a sodium cooled fast reactor. These rods were used for EBR-II reactivity control. The peak fast neutron flux at midplane of the rods was ~1.6 × 1015 n/cm2 s. The midplane segments from the irradiated rods had a burnup between 3.8 and 4.0 at.% fission. Based on the reactor physics analysis, which was performed using the validated REBUS-3/RCT [1214] and ORIGEN-RA [15] codes, more than 98% of fissions in the rods came from U-235.

Post-irradiation analysis of fuel elements

After irradiation, a 0.25 in. segment close to midplane was chopped from each fuel element and dissolved together with the cladding. Solutions were analyzed by ICP-MS using VG Plasma Quad 3—Nuclide instrument. More than 40 fission product isotopes were measured for each reactor type, of which 17 were selected to be presented here. Fission product isotopes of molybdenum could not be accurately quantified due to their presence in the U–Mo alloy fuel plates and in the stainless steel cladding of the U–Zr rods. Isotopes of zirconium were present in certain types of cladding used for U–Mo plates and in the U–Zr fuel rods. Isotopes of cesium were disregarded due to their tendency to migrate in the fuel. Other isotopes were not included here due to the presence of isobaric interferences. These interferences can be corrected after separations, which are currently in progress in our laboratory.


All cumulative fission yield values reported here were determined by the conventional relative method (also known as the “ratio method” [16, 17]) with La-139, Y-89 and Nd-148 used as standards:
$$ {\text {FY(FP)}}_{i} = {\frac{{\# {\text{atoms (FP)}}_{i} }}{\# {\text{atoms (Nd-148)}}}} \times {\text{FY (Nd-148)}} $$
where i is fission product isotope.
We have taken several aspects into consideration while selecting isotopes as our standards. Ideally, the “standard” isotopes should have accurately known fission yields, exhibit no significant variation of fission yields with energy and no significant buildup or burnout as a result of neutron capture reaction, plus they should be measured with high accuracy by ICP-MS method. As can be seen from Tables 1 and 2, the yields of Nd-148 from fission of U-235 are known with the highest accuracy among other isotopes and show practically no variation with neutron energy. However, it has been known that at high flux or high fluence irradiations in thermal spectrum, there exists a possibility of production of additional amounts of Nd-148 from Nd-147 via capture reaction, which would require correction. Such capture reactions are less pronounced in fast spectrum due to generally lower capture cross-section values in this region [18]. For this particular reason we have not used Nd-148 as a standard for the fuels irradiated in the ATR, but have used it for the analysis of EBR-II fast neutron spectrum data.
Table 1

Comparison of the U-235 cumulative thermal fission yield data measured by ICP-MS method relative to Y-89 and La-139 with ENDF/B-VII data


U-235 cumulative thermal fission yield measured relative to Y-89a, %

U-235 cumulative thermal fission yield measured relative to La-139a, %

ENDF/B-VII U-235 cumulative thermal fission yield [1], %


2.65 ± 0.08b

2.57 ± 0.07b

2.56 ± 0.03b


4.73 ± 0.14

4.58 ± 0.13

4.73 ± 0.09


5.99 ± 0.18

5.80 ± 0.17

6.11 ± 0.12


5.09 ± 0.15

4.93 ± 0.14

5.17 ± 0.10


4.20 ± 0.13

4.07 ± 0.12

4.30 ± 0.09


2.92 ± 0.09

2.83 ± 0.08

3.03 ± 0.06


1.94 ± 0.06

1.88 ± 0.05

1.88 ± 0.04


0.419 ± 0.014c

0.408 ± 0.013c

0.402 ± 0.008


0.372 ± 0.011

0.360 ± 0.010

0.349 ± 0.020


6.62 ± 0.20

6.41 ± 0.19

6.41 ± 0.09


6.48 ± 0.19

6.27 ± 0.18

6.22 ± 0.06


5.97 ± 0.18

5.78 ± 0.17

5.85 ± 0.12


6.16 ± 0.18

5.96 ± 0.17

5.85 ± 0.06


5.37 ± 0.16

5.20 ± 0.15

5.50 ± 0.04


3.79 ± 0.11

3.67 ± 0.11

3.93 ± 0.03


3.08 ± 0.09

2.99 ± 0.09

3.00 ± 0.02


1.82 ± 0.05

1.76 ± 0.05

1.67 ± 0.01

aAn average value from six U–Mo plates: TUB041, TUB025, TUB021, TUB022, TUB020 and TUB048

bAll standard deviations are presented as two sigma values

cAn average value from five U–Mo plates: TUB041, TUB021, TUB022, TUB020 and TUB048

dNd-148 + Sm-148, pending separation

Table 2

Comparison of the U-235 cumulative fast fission yield data measured by ICP-MS method relative to La-139 and Nd-148 with ENDF/B-VII data


U-235 cumulative fast fission yield measured relative to La-139a, %

U-235 cumulative fast fission yield measured relative to Nd-148a, %

ENDF/B-VII U-235 cumulative fast fission yield [1], %


6.11 ± 0.22b

6.17 ± 0.22b

5.94 ± 0.12b


5.38 ± 0.19

5.43 ± 0.20

5.12 ± 0.14


4.43 ± 0.16

4.47 ± 0.16

4.36 ± 0.09


3.27 ± 0.12

3.30 ± 0.12

3.21 ± 0.09


6.34 ± 0.23

6.40 ± 0.23

6.34 ± 0.06


6.01 ± 0.22

6.06 ± 0.22

5.98 ± 0.08


5.83 ± 0.21

5.89 ± 0.21

5.95 ± 0.17


5.37 ± 0.19

5.43 ± 0.20

5.54 ± 0.11


5.74 ± 0.21

5.80 ± 0.21

5.73 ± 0.06


5.25 ± 0.19

5.30 ± 0.19

5.27 ± 0.07


3.75 ± 0.14

3.79 ± 0.14

3.78 ± 0.04


2.92 ± 0.11

2.95 ± 0.11

2.92 ± 0.03


2.24 ± 0.08

2.26 ± 0.08

2.14 ± 0.03


1.66 ± 0.06

1.68 ± 0.06

1.68 ± 0.02


1.03 ± 0.04

1.04 ± 0.04

1.04 ± 0.01


0.441 ± 0.016

0.436 ± 0.016

0.412 ± 0.004


0.289 ± 0.010

0.286 ± 0.010

0.271 ± 0.008

aAn average value obtained from four EBR-II fuel samples: SADK41, SADK59, SADK43 and SADK09

bAll standard deviations are presented as two sigma values

The yields of La-139 and Y-89 (Tables 1 and 2) are less accurately known and display slight variation with neutron spectra; however there appears to be no significant capture reactions, which can be explained by the fact that these isotopes have magic number of neutrons [19].

Cumulative fission yield values measured in each fuel element by ICP-MS were averaged and compared with ENDF/B-VII values. Thermal fission yields are presented in Table 1 and fast spectrum yields are shown in Table 2, respectively. The ±5% relative 2 sigma errors associated with the individual measurements of fission product isotope by ICP-MS method were propagated accordingly and were included in these tables.


The data presented in Tables 1 and 2 demonstrates a relatively good agreement (within 10%) between cumulative fission yields measured in irradiated nuclear fuels by the relative ICP-MS method and ENDF/B-VII values. It also appears that the shown fission yield values depend on the isotope used as a standard.

In general, the higher than expected measured fission product values can be explained by isobaric interferences with isotopes formed up the chain through neutron capture, contamination with naturally occurred isotopes with the same mass number and/or buildup due to neutron capture reactions in the neighboring atoms. The lower fission product values are usually due to depletion as a result of neutron capture, dissolution problems and migration problems. Capture effects can be significant at higher flux and fluence irradiations and in thermal spectrum as compared to fast spectrum due to higher neutron capture cross-section values at lower neutron energies. We are performing additional separations and other corrections, which will be reported in a separate paper. One example that we would like to note is Nd-148, which has a potential isobaric interference from Sm-148 (produced by neutron capture from Pm-147), but is alleviated by a separation step, which is currently pending in our laboratory [20].

We would like to emphasize that present experiments were not specifically designed for fission yields measurements, rather we have taken advantage of the available experimental information to extract potentially useful fission yield data. For this reason, there were several limitations. For instance, self-shielding effects were not considered because enriched fuel rods or plates were irradiated instead of specially designed thin enriched targets. Certain fission product isotopes could not be measured accurately because they were either constituents of the fuel or cladding or due to isobaric interferences that require separations. None of the fission products were corrected for buildup and burnout by neutron capture at this time, though we were not expecting significant capture effects in fast reactor fuels. In addition, we have not presented results of the measurements of the “valley” and “wings” isotopes as these are still under the development in our laboratory.


We have demonstrated that ICP-MS is a useful and efficient tool in the analysis of selected fission product isotopes in enriched thermal and fast reactor nuclear fuels. Despite several limitations in this analysis, the measured cumulative U-235 fission yields fell within 10% of ENDF/B-VII values. Therefore, even in the absence of the current fission yield measurement and evaluation programs, ICP-MS analysis of enriched fuels irradiated in thermal or fast reactors can provide valuable fission yield information.


The authors would like to thank James Sommers, Marcos Jimenez, Cal Morgan and Michael Rodriguez (INL) for preparation and analysis of the fuel samples; Karl Grimm and Richard McKnight (ANL) for reactor physics analysis of EBR-II experiments; Misti Lillo and Gray Chang (INL) for the reactor physics analysis of the ATR experiments. We would also like to acknowledge RERTR Program and EBR-II Spent Fuel Treatment Program for irradiated fuel samples. This manuscript has been authored by Battelle Energy Alliance, LLC under Contract No. DE-AC07-05ID14517 with the US Department of Energy. The US Government retains and the publisher, by accepting the article for publication, acknowledges that the US Government retains a nonexclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US Government purposes.

Copyright information

© Akadémiai Kiadó, Budapest, Hungary 2009