Elsevier

Applied Radiation and Isotopes

Volume 68, Issues 7–8, July–August 2010, Pages 1454-1458
Applied Radiation and Isotopes

First measurement of the beta spectrum of 241Pu with a cryogenic detector

https://doi.org/10.1016/j.apradiso.2009.11.054Get rights and content

Abstract

The LNE-LNHB is developing metallic magnetic calorimeters, a specific type of cryogenic detectors, for beta spectrometry. The aim is the determination of the shape factors of beta spectra. Our latest detector has been designed to measure the spectrum of 241Pu, a pure beta emitter with an endpoint energy of 20.8 keV. In this paper, the detection principle of metallic magnetic calorimeters is explained and a detailed description is given of the realization of the detector enclosing a 241Pu source inside the detector absorber. A spectrum resulting from our first measurement is shown and compared with a theoretical spectrum.

Introduction

We are exploring the potential of metallic magnetic calorimeters (Enss et al., 2000; Fleischmann et al., 2005) for beta spectrometry, with the aim of determining the shape factors of beta spectra. We are particularly interested in the spectra of pure beta emitters decaying via forbidden transitions, whose spectra are difficult to calculate and often experimentally not well known. Beta spectra are generally difficult to measure with lithium drifted silicon (Si:Li) detectors or electrostatic or magnetic spectrometers due to low detection efficiency, non-linear response of the detector, or energy loss of the electrons in the source or in the dead layer at the surface of Si:Li detectors.

Metallic magnetic calorimeters (MMCs) are cryogenic detectors (LTD-12, 2008; LTD-13, 2009), usually operated at temperatures between 10 and 50 mK. We embed the beta emitter inside the detector absorber so as to realize a solid angle of 4π. So even if beta particles loose (part of) their energy in the source, the energy remains contained in the detector and will be detected. By choosing the absorber dimensions according to the maximum energy of the beta spectrum, a detection efficiency >99% can be realized over the entire spectrum, starting from a threshold at around 1% of the endpoint energy. Such a low threshold can be achieved with carefully designed MMCs for endpoint energies from a few keV (the lowest existing) up to about 1 MeV. It has been demonstrated that the linearity of MMCs can be better than 0.5% over two orders of magnitude in energy (Rodrigues et al., 2008). We believe that MMCs are a very powerful means to reliably determine the shapes of beta spectra.

Our latest detector has been designed to measure the spectrum of 241Pu, a pure beta emitter with an endpoint energy of 20.8 keV (Audi, 2003). This nuclide is present in nuclear reactor cores and nuclear waste. As for most pure beta emitters, its activity is commonly measured by liquid scintillation counting. This technique requires the precise knowledge of the spectrum shape in order to determine the detection efficiency. The spectrum of 241Pu is not very well known. The results presented here make us confident that MMCs will contribute to a better knowledge of this and other beta spectra. Finally, precise determination of shape factors and comparison with calculated spectra may contribute to a better theoretical understanding of beta decay.

Section snippets

Detection principle of metallic magnetic calorimeters

In a metallic magnetic calorimeter, the energy of a particle is measured in the form of a temperature rise. Within a wide energy range, this temperature rise is proportional to energy. The detector is composed of a metallic absorber (mostly gold) in tight thermal contact with a paramagnetic thermometer (gold containing a dilute concentration of erbium, Au:Er). A magnetic field is applied to magnetize the thermometer. The energy deposited by a particle in the detector absorber is transferred to

Experimental setup

A 241Pu source has been sandwiched between two 12 μm thick gold foils serving as absorber. A much smaller thickness of gold (<1 μm) would be sufficient to stop all beta particles. But in a very thin foil the heat diffusion would be slower. This could lead to a dependence of the pulse rise times on the position of the particle interaction within the detector, and consequently to an inhomogeneous response of the detector. The source was deposited as a small drop (0.5 mg) of a plutonium nitrate

Comparison of experimental and theoretical spectrum

Data were recorded continuously during 61 h at a sampling rate of 200 kHz. All data analysis including triggering was done offline. The pulses had a rise time of about 10 μs and a decay time (1/e) of 3.5 ms. Pulse amplitudes at the output of the SQUID amplifier ranged up to 275 mV for the endpoint of the beta spectrum, at a white noise level of 2.8 μV/Hz1/2. After digital filtering, this translates to an energy resolution, as determined for the Mn Kα line, of 29 eV (FWHM) at 5.9 keV. Applying a

Possible influence of the source quality on the experimental spectrum

In a previous experiment we have carried out an absolute activity measurement of 55Fe (Loidl et al., 2008). In that experiment, where the 55Fe source was also enclosed in the absorber of a MMC, we have observed a substantial difference in the response of the MMC to X-ray photons and Auger electrons. We assumed that X-rays are absorbed in the gold foil forming the detector absorber, whereas Auger electrons loose at least part of their energy in the FeCl3 crystals forming the source. The transfer

Conclusion

To our knowledge, we have performed the first measurement of the beta spectrum of 241Pu with a cryogenic detector. The source was enclosed inside the gold absorber of a MMC. The measured spectrum was compared with a theoretical spectrum. We have observed a shift of the endpoint for which we have no explanation. In a forthcoming measurement we will use additional X-ray lines for energy calibration. We also observe a discrepancy between the two spectra at low energies. We attribute this

References (11)

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

Cited by (51)

View all citing articles on Scopus
View full text