Photo-induced changes in the charge state of the divacancy in neutron and electron irradiated silicon

Abstract

An IR study of the absorption near 0.36 eV of the negative divacancy (V2-) in silicon has been made under additional optical excitation for neutron-irradiated silicon as a function of the initial type and doping, and also of the neutron fluence. For some n-type samples, comparisons were made with electron irradiation and with the IR absorption of the P impurity near 0.04 eV. When the ratio of the neutron fluence to the initial carrier concentration rn is 2 cm, the bands are not seen under equilibrium, but they can be made to appear in both n- and p-type Si under pumping of the sample with photons in the 0.8 to 1 eV range; the bleaching out of the bands is obtained by pumping near 0.7 eV. The pump energy dependence for the appearance of the bands indicates that at least two mechanisms can turn V20 into V2-: (i) the direct ionisation of V20 in the valence band; (ii) an indirect process in which electrons are released by some defects and trapped by V20. The threshold for the vanishing of the band is connected with the photoionisation of V2- in the conduction band. Under equilibrium conditions, the authors have measured an optical cross section of V2- at 0.343 eV of 1.45*10-15 cm2 in n-type Si at 8K. This is a limiting value corresponding to a reduced neutron fluence of less than 0.4 cm, related to Gossick's simple model of neutral damages in semiconductors where a core, made intrinsic by high defect concentration, is separated from the undisturbed crystal by a space charge region. An empirical determination of the fraction of the crystal outside the space charge region was obtained by comparing the equilibrium P concentration, measured by IR absorption, before and after neutron irradiation.