The possible effects on MOS structures of metallization material, method of vacuum deposition, method of electrode delineation, electrode size and geometry, post-metallization heat-treatments, and hydrogen incorporation were investigated. It was found that a minimum in the level of mobile ionic contamination (Qo) was achieved in a time the order of 15 min for annealing at 500~ in dry nitrogen. The minimum Qo values were proportional to either the diameter of circular electrodes or the width of rectangular electrodes in the range of 5-30 mil. Results were similar for both aluminum and aluminumcopper-silicon metallization as well as for either shadow masked or photomasked electrodes. The assumed presence of hydrogen, either arising from that produced in the deposition chamber during filament and electron beam evaporation as monitored by residual gas analysis or from hydrogen anneals prior to metallization, resulted in increased Qo for short anneal times. In interpreting the results, consideration was given mechanisms which would decrease Qo, such as diffusion of contaminant (probably sodium) to and evaporation from the electrode periphery, and increase Qo, such as introduction of contaminant and/or aluminum from the electrodes or substitution of protons for contaminant ions at traps.The importance of minimizing mobile alkali ion contamination (Qo) in silicon device processing is well recognized (1). Such contamination in silicon dioxide can affect the threshold voltage and stability of MOS devices. Adverse effects on bipolar devices, primarily Key words: mobile impurity monitoring, 1ViOS electrode size effects, hydrogen effects, annealing effects, ) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.210.126.199 Downloaded on 2015-06-28 to IP ) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.210.126.199 Downloaded on 2015-06-28 to IP