Energy resolution enhancement in ion beam experiments with Bayesian probability theory

In experiments with MeV ion beams the energy resolution of the detector and the energy spread of the incident ion beam restricts the energy and depth resolution. The deconvolution of the measured spectra with the energy transfer function is an ill-posed problem which is tackled consistently by the Bayesian probability theory. Along with the entropic prior and the adaptive kernel method we determine the most non-committal energy distribution compatible with the measured data together with its confidence interval. An enhancement of the resolution by a factor of 6 in Rutherford backscattering analysis of thin films, resulting in an energy resolution of about 3 keV with semiconductor detectors, is obtained. The uncertainty of the transfer function is considered properly. The deconvolution allows the separation of the isotopes of Cu and Fe. The resolution enhancement of depth effects of energy loss straggling and multiple scattering is demonstrated for a CoAu multi-layer film.