On the conversion of NDP energy spectra into depth concentration profiles for thin-films all-solid-state batteries

A three-step numerical procedure has been developed, which facilitates
the conversion of NDP energy spectra into lithium concentration
depth profiles for thin-film Li-ion batteries. The procedure
is based on Monte Carlo modeling of the energy loss of charged
particles (ions) in the solid media, using the publically available
SRIM/TRIM software. For the energy-to-depth conversion, the battery
stack has been split into finite volume elements. Each finite volume
element becomes a source of ions according to the employed
nuclear reaction. Ions loos energy when they move across the battery
stack towards the detector. The as-obtained simulated spectra
have been compared with the experimentally measured spectra. The
thicknesses of the battery stack layers were estimated by minimizing
the deviation between the simulated and measured spectra. Subsequently,
a relation between the average energy of detected ions and
the depth of the corresponding finite volume element, yielding a calibration
function, was used to relate that particular part of the spectra
with the depth of its source. At the final stage, a Bayesian estimator
was used to find the distribution of lithium at a particular depth. The
developed procedure was applied to a practically relevant case study
of Si immobilization in the LPO electrolyte of all-solid-state thin-film
batteries. It is shown that the lithium immobilization process in the
LPO electrolyte is responsible for the battery degradation process.