Group II residual impurities are normally present in InP at low concentrations. They are known to act as acceptors, producing shallow levels in the gap [62]. It has been observed that the presence of hydrogen is able to passivate these impurities, and it is only recently that there have been attempts to fully understand this process [59,53].
Experimental investigations of H- and D-plasma passivated Be, Mg, Cd and Zn have been performed [59]. Infra-red absorption studies show hydrogen local vibrational modes (LVMs) which lie in the frequency range where P-H stretch modes are expected (2200 - 2300 cm-1). This suggests that hydrogen bonds to phosphorus atoms in the lattice instead of the acceptor impurities.
Uniaxial stress measurements show that these acceptor - hydrogen defects have C3v symmetry [59]. Therefore, since these Group II acceptors are known to substitute for indium, hydrogen must lie somewhere along the 111 acceptor-P direction. Passivation with a mixed H,D plasma produces no peak splitting, suggesting that there is only one hydrogen atom present in each defect. It was suggested that passivation occurred through a chemical recombination, allowing a lower coordination for the Group II acceptor atoms and tying up the resultant dangling bond on the neighbouring phosphorus atom [59].
We examine several possible Be-H complexes in InP in order to determine the lowest energy structure and the passivation mechanism of the hydrogen. We then examine how H passivates Be and Mg in InP and GaAs, to examine trends with variation in impurity or host material.