Correlation of arsenic incorporation and its electrical activation in MBE HgCdTe

The behavior of arsenic for p-type doping of MBE HgCdTe layers has been studied for various annealing temperatures and arsenic doping concentrations. We have demonstrated that arsenic is in-situ incorporated into HgCdTe layers during MBE growth. The carrier concentration has been measured by the Van der Pauw technique, and the total arsenic concentration has been determined by secondary ion mass spectroscopy. After annealing at 250 degrees C under an Hg over pressure, As-doped HgCdTe layers show highly compensated n-type properties and the carrier concentration is approximately constant (similar to mid 10(15) cm(-3)) until the total arsenic concentration in the HgCdTe layers approach mid 10(17) cm(-3). The source of n-type behavior does not appear to be associated with arsenic dopants, such as arsenic atoms occupying Hg vacancy sites, but rather unidentified structural defects acting as donors. When the total arsenic concentration is above mid 10(17) cm(-3), the carrier concentration shows a dependence on the arsenic concentration while remaining n-type. We conjecture that the increase in n-type behavior may be due to donor arsenic tetramers or donor tetramer clusters. Above a total arsenic concentration of 1 similar to 2 x 10(18) cm(-3), after annealing at 300 degrees C, the arsenic acceptor activation ratio rapidly decreases below 100% with increasing arsenic concentration and is smaller than that after annealing at 450 degrees C. The electrically inactive arsenic is inferred to be in the form of neutral arsenic tetramer clusters incorporated during the MBE growth. Annealing at 450 degrees C appears to supply enough thermal energy to break some of the bonds of neutral arsenic tetramer clusters so that the separated arsenic atoms could occupy Te sites and behave as accepters. However, the number of arsenic atoms on Te sites is saturated at similar to 2 x 10(18) cm(-3), possibly due to a limitation of its solid solubility in HgCdTe.