Investigations of donor and acceptor removal and long term annealing in silicon with different boron/phosphorus ratios

Abstract

Studies of radiation damage of silicon detectors have shown that in most elementary particle physics applications, the major problem caused by non-ionizing interactions in silicon will be the change in the concentration of the electrically active impurity states. For many silicon applications in high radiation environments, the increase in the relative acceptor concentration during long term room temperature annealing is the limiting factor. The first step in developing radiation hardened silicon is to identify the defects which are responsible for this effect. In earlier work, we developed a model which describes the experimentally observed fast neutron induced changes in the net electrically active impurity concentration and provides a method to determine the phosphorus and boron concentration of the starting material. Our previous work has been extended with experiments using silicon with different original boron/phosphorus ratios, larger neutron fluences and to long term room temperature annealing. The donor removal rate of phosphorus was measured and the acceptor removal rate of boron evaluated. Extending our model, we propose that slow restoration of the damage induced boron interstitial back to its original substitutional state is responsible for the acceptor increase observed during long term annealing. This model, which involves only the behavior of the known shallow donors and acceptors present in the starting material, fits our data and the data of other experimenters without the need to invoke unidentified deep levels.