Page 168 - 00. Complete Version - Progress Report IPEN 2014-2016

168 Nuclear Science and Technology | Progress Report

aiming to improve IPEN’s NAA results, either ports the results to an Excel-compatible spread-
by focusing in the technique itself or in the sheet was developed. The software, written
data analysis. in the Pascal computer language, makes use
of several tools shipped with Canberra’s Ge-
In this sense, Sm determination in urani- nie-2000 to import all the spectra found in a
um-rich samples was studied, focusing not given folder, import energy and FWHM calibra-
only in the interference of fission-produced tions from a saved reference spectrum (chosen
Sm, but also in the spectral interference in the by the operator according to the detector sys-
103keV gamma-ray line that arises from the tem from which the spectra came) and analyze
x-rays of Pu, which is produced by sequential the resulting spectra; an in-house procedure
239
238
neutron captures on U and Np. The result- then reads Genie’s output file and produces a
ing time-depending correction factor, a novel Comma-Separated Values (CSV) file converted
concept that arose from this work, proved to to the Brazilian language standards (commas
properly address the problem, rendering re- as decimal separators and semicolons separat-
liable results for Sm even in the presence of ing fields) that can be opened in a typical in-
high amounts of uranium. stallation of Microsoft Excel.

A thorough investigation on sources of uncer- Preliminary studies aiming to implement coin-
tainty in comparative NAA measurements has cidence comparative NAA in IPEN were started,
also been made, focusing not only in the ex- in collaboration with NIST and UTexas-Austin
plicit uncertainty factors, but also in factors (USA), and a fully-digital gamma-gamma co-
that are hidden in the simplification made in incidence system is under development at IP-
the comparative method. In this work, the pos- EN’s Applied Nuclear Physics Laboratory. This
sible variation in the HPGe detector efficiency technique aims to provide an extra tool for
between measurements was studied in great NAA analyses when regular NAA can’t give
detail, and it was shown that for dead-times good results because of strong spectral inter-
below 10%, the efficiency variation won’t con- ferences, either from very similar gamma-ray
tribute with more than 0.05% to the final un- peaks or from strong Compton or bremsstrah-
certainty. A study on the explicit sources of lung continua.
uncertainty has shown that, for long-lived
isotopes (half-lives greater than one day), the Basic Nuclear Physics
time-related parameters (decay constant and
time) contribute with no more than 0.5% of The half-life of short-lived 27Mg was studied
the final uncertainty; the counting statistics by following the activity of samples after they
and comparator concentration contribute with were irradiated in the IEA-R1 reactor. The re-
most of the uncertainty (usually around 80- sulting value of 564.5(7) s is compatible with
90% of the total), while the uncertainties in most of the literature values, but not with the
the masses typically contribute with less than ENSDF compilation value, accepted as a ref-
20% of the final uncertainty value. erence value for most experiments, suggest-
ing that further precise measurements of this
In an attempt to promote automation of the half-life should be made, and that the ENSDF
NAA data analysis procedures, a simple soft- value should probably be reassessed.
ware that automatically analyses spectra in
different formats (CHN, CNF and MCA) and ex- In a collaboration with the Pelletron laborato-

Instituto de Pesquisas Energéticas e Nucleares