We’ll add two new input files for the decay scheme: an “atomic widths” file and a “isotopes” file. Each line of the atomic widths file will be formatted as <z of atom>, <shell>, <width in eV for a vacancy in that shell>. We can fill this file out based on the paper Lijie sent us (Natural widths of atomic K and L levels, Kα X‐ray lines and several KLL Auger lines).

The isotopes file will the mass of the ground state. Currently the entire Q value of alpha and proton decays goes into the alpha/proton, so including the mass will let us properly account for the energy of the recoiling nucleus. The format would be: <isotope name>, <# of protons>, <ground state mass in keV/c^2> < next isotope name>, <# of protons>, <ground state mass in keV/c^2> X-ray weights and energies will be removed from electron capture and internal conversion entries in decay scheme files since they will now be generated by based on Z of the nuclei at the time of the decay. These entries will be replaced by the capture fraction of K, L, and M shell electrons. I initially thought that Geant was able to estimate these capture fractions, but after more digging it turns out that they are simply tabulated in the data files Geant looks at (you can find these in “/mnt/simulations/proton_detector/Geant4/geant4.10.2-install/share/Geant4-10.2.2/data/RadioactiveDecay4.3.2” on fishtank). Instead we can estimate them using the logft tool or BetaShape. The DecayScheme class will be updated to use partial half-lifes rather than “weights” to determine which transition occurs. For each possible decay channel, a “decay time” will be randomly generated based on the partial half-life for that channel, and the channel with the smallest time will be selected. TODO: are these partial half-lives simply those in Rauscher’s “avwi” files? When an EC or internal conversion occurs, an “emit xray” flag will be set. When each subsequent transition occurs, the width of the state will be used to randomly generate a “time in state”, which can then be used to evaluate the probability that an x-ray is emitted prior to the next transition. If an x-ray is emitted, the atomic relaxation routines already built into Geant4 – very similar to what’s done in DecayIt() in “G4ECDecay.cc” in the Geant4 source – will be used to generate x-rays and auger electrons and the “emit xray” flag will be set back to false. The same procedure will be used if there are no further transitions to be performed and the “emit xray” flag is set to true. Note that only one xray will be emitted in the edge case where an EC occurs followed by IC or another EC prior to the first x-ray emission. TODO: internal conversion doesn’t seem to be listed in Rauscher’s files. Do we need to do something about this since internal conversion also makes x-rays? Tasks: • Put our code on FRIB’s gitlab (Alex) • Update DecayScheme code as described above (Alex) • Make input files for updated decay scheme o Generate atomic widths file from the paper Lijie sent (put in seconds) (Joe) o Generate capture fractions using BetaShape, by using the theory calculations to generate ENSDF files and then running BetaShape on these files o Generate isotope file from Alex Brown’s shell calculations + NuDat masses (Joe) o Generate updated decay scheme files based on Brown and Rauschers calculations + output from BetaShape (lifetimes in seconds) • Update beta decay transitions to use BetaDecayIt method of G4NuclearDecayChannel • Update geometry

Longer term tasks: Use Gaussian process emulator