High-resolution alpha-particle spectrometry of 238U

https://doi.org/10.1016/j.apradiso.2013.11.075Get rights and content

Highlights

  • Measured alpha-particle emission probabilities in the decay of 238U.

  • Used high-resolution alpha-particle spectrometry on electrodeposited sources.

  • Improved accuracy and precision significantly.

  • Results are Pα,0=77.01 (10)%, Pα,1=22.92 (10)% and Pα,2=0.068 (10)%.

Abstract

The alpha-particle emission probabilities associated with the three main alpha transitions of 238U were measured by high-resolution alpha-particle spectrometry. Highly enriched 238U material was used and its isotopic composition characterised by mass spectrometry. Source production through electrodeposition was optimised to reconcile conflicting demands for good spectral resolution and statistical precision. Measurements were performed at IRMM and CIEMAT for 1–2 years in three different set-ups. A new magnet system was put into use to largely eliminate true coincidence effects with low-energy conversion electrons. Finally the accuracy and precision of the relative emission probabilities for the three transitions – 77.01 (10)%, 22.92 (10)% and 0.068 (10)%, respectively – have been improved significantly.

Introduction

The IAEA CRP F42006 (Coordinated Research Activity on “Updated Decay Data Library for Actinides”) expressed the need for more accurate values of alpha-particle emission probabilities for 238U (Kellet et al., 2011, Kellet, 2012). This task was taken on as part of a work package on improvement of decay data in the JRP ENG08 “Metrology for New Generation Nuclear Power Plants—MetroFission” (Johansson et al., 2011). Due to its extremely low specific activity, 238U is difficult to measure by high-resolution alpha-particle spectrometry. The main difficulty is to combine good spectral resolution, which requires thin sources, with good statistical precision, which requires a relatively high count rate and therefore sufficiently thick sources.

The alpha-particle energy spectrum shows two major peaks (at 4151 keV and 4198 keV) and a small one at lower energy (4038 keV). Only two measurements of 238U alpha-emission probabilities have been reported (Kocharov et al., 1959, García-Toraño, 2000), the most recent values – Pα,0=77.54 (50)%, Pα,1=22.33 (50)% and Pα,2=0.13 (3)% – being recommended by evaluators (Chisté and Bé, 2006). The challenge of the current work was to improve the quality of the existing decay data through better isotopic purity of source material, spectral resolution, statistical precision and more accurate correction for true coincidence effects between alpha-particles and conversion electrons. This cocktail of conflicting demands called for metrological solutions with respect to source preparation, detector stability, optimisation of measurement geometry and suppression of conversion electron signals.

Highly enriched 238U material was obtained and its isotopic composition characterised by mass spectrometry. Extensive investigation went into source preparation by electrodeposition, involving many free parameters such as type of electrolyte, shape and rotation of electrode, surface quality of backing, source radius and thickness (Jobbágy et al., 2011, Jobbágy et al., 2013). High-resolution alpha-particle spectrometry measurements were performed for 1–2 years in three different set-ups, with a different focus on count rate and resolution. A new magnet system was developed (Paepen et al., 2014) and used to prevent conversion electrons from being detected in coincidence with the alpha particles. Finally the accuracy and precision of the relative emission probabilities for the three transitions could be improved significantly. In this paper, the measurements and results are described.

Section snippets

Source preparation

The applied source production procedure is a modification of the electrodeposition method by Dos Santos et al. (2004). Jobbágy et al., 2011, Jobbágy et al., 2013 have described in detail the different source electrodeposition parameters tested at IRMM and CIEMAT to reconcile the inherently opposing requirements of high resolution and sufficient count rate needed to determine the alpha emission probabilities more accurately. Three different electrolytes were tested with respect to attainable

Alpha peak fit

The least-squares fitting software ALPACA (CIEMAT) (García-Toraño, 1996) and ALPHA (IRMM) (Pommé and Sibbens, 2008) were used to deconvolute the 238U alpha spectra. In Fig. 1, Fig. 2, Fig. 3, measured 238U alpha spectra and their spectral deconvolution are shown for three configurations. In all the spectra, the three main alpha-peaks of 238U at 4038 (5) keV, 4151 (5) keV and 4198 (3) keV are easily identified and well separated. The energy resolution and peak-to-valley ratio are noticeably better

Discussion

The uncorrected emission probabilities obtained in configuration CIEMAT1 (G=2%, no magnet) are in good agreement with the previously published measurement results obtained by García-Toraño (2000) in similar conditions in the same laboratory. This time, better statistical precision was achieved, leading to a lower amplitude of the small peak, Pα,2. The new emission probabilities are also consistent with the results obtained with a similar configuration at IRMM (IRMM1, G=1.7%, no magnet).

The

Conclusions

The alpha-particle emission probabilities of 238U have been measured with high-resolution alpha-particle spectrometry. Sources were prepared of highly enriched 238U material by electrodeposition with 18–32 cm diameter and up to 26 μg/cm2 thickness. This was the result of a compromise between spectral resolution and counting rate.

The measured emission probabilities, shown in Table 1, agree within uncertainties with the previously published data but are significantly more precise. The new data are

Acknowledgement

The research leading to these results has received funding from the European Union on the basis of Decision No 912/2009/EC.

References (19)

There are more references available in the full text version of this article.

Cited by (10)

  • Alpha spectrometry and liquid scintillation counting for the measurement of <sup>238</sup>Pu, <sup>239</sup>Pu, <sup>240</sup>Pu, <sup>241</sup>Pu, <sup>242</sup>Pu and age

    2020, Applied Radiation and Isotopes
    Citation Excerpt :

    Different approaches have been applied to minimize such coincidences which, if not low enough, may cause biases in a result from a deconvoluted spectrum. Coincidences between alpha particles and conversion electrons can be avoided, or minimized, using magnets (Pommé et al., 2014) and/or with a large distance between the sample and the detector. The latter will also reduce coincidences with low energy photons (Pöllänen et al., 2012).

  • Alpha spectrometry

    2020, Handbook of Radioactivity Analysis: Volume 1: Radiation Physics and Detectors
  • Direct measurement of alpha emission probabilities in the decay of <sup>226</sup>Ra

    2017, Applied Radiation and Isotopes
    Citation Excerpt :

    Without magnet system, these effects can only be avoided by lowering the effective solid angle. These dilemmas have also been encountered with alpha spectrometry of long-lived nuclides such as 238U (Pommé et al., 2014) and 236U (Marouli et al., 2014). A typical spectrum of the 226Raeq decay series and the corresponding background spectrum obtained at the JRC are shown in Fig. 4.

  • Improved peak shape fitting in alpha spectra

    2015, Applied Radiation and Isotopes
    Citation Excerpt :

    Comparison of fit results with different codes does reveal biases that are not under statistical control (Sibbens et al., 2010). The preferred way of dealing with uncertainties in radionuclide metrology is to make for each peak a detailed uncertainty budget and perform proper uncertainty propagation towards the final result (see e.g. Sibbens et al., 2010; Pommé et al., 2014; Pöllänen and Siiskonen, 2014). Uncertainty components include counting statistics, spectral interferences, impurities and background, residual deviations between fit and measurement, physical effects not included in the model and model dependence of the fit result.

View all citing articles on Scopus
View full text