The measurement of gamma emitters natural radionuclides is a very relevant subject in environmental radioactivity in many fields. For this, a valid calibration based on the full-energy peak efficiency (FEPE) is essential, which can be carried out for gamma-ray spectrometry using Monte Carlo codes (Geant4, LabSOCS, MNCP, etc.) without needing a source preparation and with much less time consuming compared to experimental methodologies. Therefore, this work aims to develop a methodology for the FEPE determination of uncharacterized Ge detectors by using characterized Ge detectors with known FEPE. For this, a general relationship was found between the simulated FEPE (εsim) for characterized Ge detectors by LabSOCS and the experimental one (εexp) for uncharacterized Ge detectors obtained for a problem sample (phosphate rock (PR) in our case), where εsim and εexp were factorized using three correction factors: geometrical, photon self-absorption and true coincidence summing (TCS) effects. Thus, the εsim/εexp ratio (FEPER) was obtained for two couples of Ge detectors (2 characterized and 1 uncharacterized detectors), using 4 different cylindrical geometries (C) and 2 Marinelli (M) ones, and varying the sample thickness (h), and gamma energy (Eγ) (from 46 keV (210Pb) to 1765 keV (214Bi)). The consistency obtained for the FEPERs for C geometries suggested a constant geometric factor, finding some deviations for M geometries only at Eγ < 150 keV. Given a pair of detectors, the behavior of the FEPERs was found to be very similar for all geometries, and h at each specific Eγ. Moreover, a constat relationship between the FEPERs of both detector couples versus Eγ was found. Then, the influence of the TCS effects was also studied varying the distance between the geometry bottom and detector window. The developed methodology was also validated confirming the consistency of the FEPERs, which were found to be independent on the chemical composition and apparent density of the sample, supporting our theoretical hypothesis. Consequently, given that the FEPER was found to be independent on the sample type and h, using our developed methodology it is possible to obtain εexp for any sample, geometry and h using the FEPER, previously obtained at each Eγ for the sample selected for methodology development (PR in our case), and the εsim that is obtained for the desired case.
Read full abstract