Accurate simulation of ionisation chamber response with the Monte Carlo code PENELOPE

Ionisation chambers (IC) are routinely used in hospitals for the dosimetry of the photon and electron beams used for radiotherapy treatments. The determination of absorbed dose to water from the absorbed dose to the air filling the cavity requires the introduction of stopping power ratios and perturbation factors, which account for the disturbance caused by the presence of the chamber. Although this may seem a problem readily amenable to Monte Carlo simulation, the fact is that the accurate determination of IC response has been, for various decades, one of the most important challenges of the simulation of electromagnetic showers. The main difficulty stems from the use of condensed history techniques for electron and positron transport. This approach, which involves grouping a large number of interactions into a single artificial event, is known to produce the so-called interface effects when particles travel across surfaces separating different media. These effects can be sizeable when the electron step length is not negligible compared to the size of the region being crossed, as it is the case with the cavity of an IC. The artefact, which becomes apparent when the chamber response shows a marked dependence on the adopted step size, can be palliated with the use of sophisticated electron transport algorithms. These topics are discussed in the context of the transport model implemented in the PENELOPE code. The degree of violation of the Fano theorem for a simple, planar geometry, is used as a measure of the stability of the algorithm with respect to variations of the electron step length, thus assessing the 'quality' of its condensed history scheme. It is shown that, with a suitable choice of transport parameters, PENELOPE simulates IC response with an accuracy of the order of 0.1%.