A systematic study on discharge-induced GEM-failure phenomena

The gas electron multiplier (GEM), proposed by F. Sauli in 1997, is now a well established concept for electron amplification in gas detectors. One of the most important reason of the success and of the widespread use of GEM-based detectors is the reliability of the electron amplifying structure with respect to discharges, in contrast to other types of micro-pattern detectors. However, even if a GEM foil can tolerate a certain amount of discharges, these are "memorized" in the amplifying structure and after exceeding a certain amount of discharges the GEM foil operation will inevitably fail. This fact has to be careful considered when planning the use of GEM-based detectors in an environment where heavily ionizing particles can considerably increase the GEM discharge probability. In this paper a statistical model predicting the GEM survival probability to discharge events is proposed. This model requires in input the knowledge of the effects of discharges in a single GEM hole, and these effects were measured by means of a specially designed structure where single GEM holes can be individually supplied and studied. Results from these measurements are fed in the model in order to predict the GEM survival probability distribution. This probability distribution is found to be in agreement with measurements performed on standard triple-GEM detectors where discharges are induced by means of an alpha source. The model presented here can be a valuable tool to help designing GEM-based detectors and estimating how many detectors will fail after integrating a certain amount of discharges