Constraints on partial half-lives of Ce136 and Ce138 double electron captures

The $\ensuremath{\gamma}$-ray emissions from a radiopure cerium-bromide crystal with a mass of 4381 g were measured for a total of 497.4 d by means of high-resolution $\ensuremath{\gamma}$-ray spectrometry in the HADES underground laboratory at a depth of 500 m.w.e. A search for $0/2\ensuremath{\nu}\ensuremath{\epsilon}\ensuremath{\epsilon}$ and $0/2\ensuremath{\nu}\ensuremath{\epsilon}{\ensuremath{\beta}}^{+}$ double beta decay transitions of $^{136}\mathrm{Ce}$ and $^{138}\mathrm{Ce}$ was performed using Bayesian analysis techniques. No signals were observed for a total of 35 investigated decay modes. 90% credibility limits were set in the order of ${10}^{18}$--${10}^{19}$ yr. Existing constraints from a cerium oxide powder measurement were tested with a different cerium compound and half-life limits could be improved for most of the decay modes. The most likely accessible decay mode of the $^{136}\mathrm{Ce}$ $2\ensuremath{\nu}\ensuremath{\epsilon}\ensuremath{\epsilon}$ transition into the ${0}_{1}^{+}$ state of $^{136}\mathrm{Ba}$ results in a new best 90% credibility limit of $5.0\ifmmode\times\else\texttimes\fi{}{10}^{18}$ yr.