Optimization to improve precision in neutron activation analysis

The level of precision or accuracy required in analysis should be satisfied the general requirements and customer needs. In presenting the results of the analysis, the level of precision is expressed as uncertainty. Requirement general is Horwitz prediction. Factors affecting the uncertainty in the Neutron Activation Analysis (NAA) include the mass of sample, mass standards, concentration in standard, count of sample, count of standard and counting geometry. Therefore, to achieve the expected level of precision, these parameters need to be optimized. A standard concentration of similar materials is applied as a basis of calculation. In the calculation NIST SRM 2704 is applied for sediment samples. Mass of sample, irradiation time and cooling time can be modified to obtain the expected uncertainty. The prediction results show the level of precision for Al, V, Mg, Mn, K, Na, As, Cr, Co, Fe, and Zn eligible the Horwitz. The predictive the count and standard deviation for Mg-27 and Zn-65 were higher than the actual value occurred due to overlapping of Mg-27 and Mn-54 peaks and Zn-65 and Fe-59 peaks. Precision level of Ca is greater than the Horwitz, since the value of microscopic cross section, the probability of radiation emission of Ca-49 and gamma spectrometer efficiency at 3084 keV is relatively small. Increased precision can only be done by extending the counting time and multiply the number of samples, because of the fixed value. The prediction results are in accordance with experimental results. (author)