Experimental verification of EGSnrc Monte Carlo calculated ion chamber response in low energy photon beams

The EGSnrc Monte Carlo system was used to calculate response, defined as the ratio of the dose to the gas in the cavity and air kerma free-in-air at the center of the cavity, normalized at /sup 60/Co. Calculations of the response of a graphite cavity and Exradin A12 air-equivalent plastic chambers were compared with measurements using free-air ionization chambers. For the graphite chamber, the agreement between measured and calculated chamber response diverges up to 8% down to 50 kV. The authors simulated the high-Z contamination, distributed over the inner surface of the wall in the form of iron grains, by introducing an iron ring of 1 /spl mu/m thickness covering 0.45% of the area on the inside wall. These amounts of high-Z contamination are perfectly conceivable considering the machining process of the graphite cavity. Assuming pure C552, for the Exradin A12 the agreement is within 0.5% down to 85 keV but gradually gets worse at lower energies. A mass spectroscopic analysis of C552 showed that the amount of impurities with Z between 10 and 82 amounts to 90 /spl mu/g/g. The calculations with the full composition as resulting from the analysis show agreement within 0.4% down to 70 keV. Below that energy, deviations amount up to 3%. These can be explained by (1) impurities on the inside surface of the wall from the machining process as demonstrated from measurements on other chambers of the same type; (2) accuracy of the spectra used in the calculations; (3) geometrical details of the chamber such as the stem.