The effect of using air in tissue-equivalent ionization chambers for the measurement of fast neutron absorbed dose distributions in phantoms

Dosimetry of fast neutron beams is undertaken most frequently with tissue-equivalent ionisation chambers filled with tissue-equivalent gas. However, measurements of absorbed dose distributions in a phantom are frequently carried out with air instead of tissue-equivalent gas. Under these conditions the signal from the air-filled chamber relative to that from a chamber filled with tissue-equivalent gas varies with chamber volume, field size, depth in the phantom and neutron beam energy. Consequently isodose distributions measured with air-filled chambers could be in error. For example, with the dose normalised at 5 cm deep, the use of a 0.1 cc chamber would underestimate the surface dose by 1% and overestimate doses at greater depths by up to 1% for a d(16)+Be beam. Larger errors occur for larger chambers but the errors decrease with increasing beam energy. It is suggested that the errors are due to variations in neutron interactions in the gas particularly at low energies