Abstract
Motivated by the disagreement and irreproducibility observed by different groups, including ours, on the effects of passivating compound semiconductors with sulfur, we have attempted in this work to see if dissimilarities in the starting solution can account for variations in final electrical results. Specifically we have tried passivating InP with different ammonium sulfide solutions for metal-insulator–semiconductor (MIS) type applications. We have observed that InP treated with a polysulfide solution, prepared by bubbling O2 through ammonium sulfide with excess dissolved sulfur results in excellent interfaces, whereas polysulfide-free solutions have little effect. Interface state densities in the high 1010 cm−2 eV−1 as judged by quasistatic capacitance-voltage measurements were obtained on polysulfide treated MIS structures coated with indirect plasma-enhanced chemical vapor deposited SiO2. Low-temperature photoluminescence spectra show marked differences on polysulfide-treated InP when compared to InP that was treated with commercially available ammonium sulfide.