Experimental study of forbidden beta transitions

The precise experimental characterization of non-unique forbidden beta transitions remains both an important and challenging topic. Indeed, only a limited number of reliable studies can be found in the literature. The continuous energy spectra of these transitions are particularly difficult to measure accurately due to several factors: the high diffusivity of electrons in matter and the non-linearity of detection systems; the limited availability of radionuclides and the presence of impurities; their long half-lives and the complexity of the associated decay schemes. From a theoretical standpoint, realistic predictions are also challenging because they require coupling accurate atomic and nuclear structure models through the weak interaction within a consistent, fully relativistic framework. Improving our understanding of non-unique forbidden beta transitions is, however, of fundamental importance in radioactivity metrology, particularly for the realization of the becquerel in the case of pure beta emitters. Furthermore, it has significant implications for nuclear medicine (microdosimetry, targeted radionuclide therapy) and for the nuclear industry (reactor residual power estimation, nuclear waste management). Recent studies have also highlighted the relevance of these transitions to several fundamental physics topics, such as dark matter searches and reactor neutrino physics. In response to the need for high-precision beta spectra, a dedicated beta spectrometer was developed at LNE-LNHB to measure beta spectrum shapes and improve decay data for beta emitters. During this PhD thesis, the spectrometer — featuring an almost 4π detection geometry and operating with an ultrathin radioactive source positioned between two silicon detectors — was extensively upgraded. The original PIPS detectors were replaced by Si(Li) detectors, the entire front-end electronics was renewed, and a new data acquisition system was implemented. The methods for radioactive source preparation were also further optimized. The reconstruction of the emitted beta spectra was performed using a novel coincidence analysis combined with a spectral unfolding approach. Several unfolding techniques were investigated to evaluate potential biases, with Tikhonov regularization providing the most reliable results. The latter was therefore adopted for the analysis. The response matrix of the detection system, required for the unfolding, was obtained using a detailed Geant4 simulation incorporating a highly accurate geometric model of the spectrometer. The quality of the Geant4 modeling was validated against measurements performed with a ²⁰⁷Bi calibration source. The beta spectra of ¹⁴⁷Pm and ⁹⁰Sr/⁹⁰Y were subsequently measured and analyzed to extract the endpoint energy and shape factor. For ¹⁴⁷Pm, an endpoint energy of 226.2(7) keV was obtained, slightly higher than the AME2020 evaluated value. No significant deviation from an allowed shape was observed. For ⁹⁰Sr/⁹⁰Y, simultaneous fitting yielded endpoint energies of 547.7(14) keV for ⁹⁰Sr and 2264.9(35) keV for ⁹⁰Y. The ⁹⁰Sr result is consistent with the AME2020 evaluation, while the ⁹⁰Y value lies about 10 keV below the evaluated one. Due to uncertainties in the energy calibration, no meaningful shape factor could be extracted for the ⁹⁰Sr/⁹⁰Y data. Finally, the properties of the first excited 0⁺ state of ⁹⁰Zr were investigated using coincidence analysis. The half-life of this state was measured to be 53.5(11) ns, differing from the previously reported value of 61.3(25) ns. Its excitation energy was estimated at 1754.9(16) keV, also in disagreement with the ENSDF evaluated value.

Beta spectra, Semiconductor detectors, Monte Carlo simulations, Spectrum unfolding, Shape factor, Q-Value

Full text: https://theses.fr/2025UPASP147