Search

For a given sample material and experimental arrangement, there is an optimum sample size and shape. Samples must be large enough to give a high capture rate for good counting statistics in a reasonable measurement time but small enough to avoid inaccuracies that creep in with high counting rates, due to broadening and distortion of peaks caused by pulse pileup (Lindstrom and Fleming 1995). In addition, the sample should be small enough that bias from neutron absorption or scattering will be acceptably small. These considerations usually lead preferentially to small samples and long irradiation times. The composition, the experimental conditions, and the required accuracy determine what “small” means numerically.

Twenty-six brain tumor patients have been treated with the boron neutron capture therapy (BNCT) at the JRR-2, Japan Atomic Energy Research Institute (JAERI)1, since 1990. An accurate analysis of ppm levels of 10B content in tissues is essential for determining neutron irradiation dose in the BNCT. Although the ppm level of 10B can be analyzed by destructive analytical methods of inductively coupled plasma atomic emission spectrometry (ICP-AES) and spectrophotometry, due to the necessity of sample decomposition, these methods are not suitable when quick analysis is required for determining the neutron irradiation dose before the BNCT treatment. A neutron-induced prompt γ-ray analysis (PGA)2 is characterized by the capability of nondestructive 10B analysis. On account of large cross section of the reaction 10B(n,α)7*Li (σ = 3837 barn), the Doppler broadened 478 keV photo-peak is observed with the highest sensitivity by the PGA. The 10B in the tissues was determined by measuring the 478 keV peak area against the 2223 keV peak area for the 1H as an internal standard3,4.