Monte Carlo study of correction factors for Spencer-Attix cavity theory atphoton energies at or above 100 keV

To develop a primary standard for Ir-192 sources, the basic science on whic h this standard is based, i.e., Spencer-Attix cavity theory, must be establ ished. In the present study Monte Carlo techniques are used to investigate the accuracy of this cavity theory for photons in the energy range from 20 to 1300 keV, since it is usually not applied at energies below that of Cs-1 37. Ma and Nahum [Phys. Med. Biol. 36, 413-428 (1991)] found that in low-en ergy photon beams the contribution from electrons caused by photons interac ting in the cavity is substantial. For the average energy of the Ir-192 spe ctrum they found a departure from Bragg-Gray conditions of up to 3% caused by photon interactions in the cavity. When Monte Carlo is used to calculate the response of a graphite ion chamber to an encapsulated Ir-192 source it is found that it differs by less than 0.3% from the value predicted by Spe ncer-Attix cavity theory. Based on these Monte Carlo calculations, for cavi ties in graphite it is concluded that the Spencer-Attix cavity theory with Delta = 10 keV is applicable within 0.5% for photon energies at 300 keV or above despite the breakdown of the assumption that there is no interaction of photons within the cavity. This means that it is possible to use a graph ite ion chamber and Spencer-Attix cavity theory to calibrate an Ir-192 sour ce. It is also found that the use of Delta related to the mean chord length instead of Delta = 10 keV improves the agreement with Spencer-Attix cavity theory at Co-60 from 0.2% to within 0.1% of unity. This is at the level of accuracy of which the Monte Carlo code EGSnrc calculates ion chamber respo nses. In addition, it is shown that the effects of other materials, e.g., i nsulators and holders, have a substantial effect on the ion chamber respons e and should be included in the correction factors for a primary standard o f air kerma. [S0093-2405(00)02008-3].