Answers are grounded in this paper and cite their sources.
The low-frequency resistance noise in sputtered-deposited magnetic tunnel junctions with MgO barriers has been measured as a function of annealing time at different annealing temperatures. The noise has a 1/f spectrum and it is quantified by a Hooge-like parameter α given in units of μm2. Unannealed devices have the highest noise levels and their α parameters exhibit a pronounced dependence on the voltage bias across the junction. A significant increase in tunneling magnetoresistance (TMR) is observed for short annealing times (on the order of minutes) at high temperatures and it is correlated with a large reduction in noise and in its bias dependence. The maximum TMR and minimum noise levels are reached at a later time that depends on temperature, being shorter at higher annealing temperatures. Devices annealed at 380 and at 430 °C exhibit the same minimum noise levels, α≈2×10−10 μm2. The origin of the resistance noise, its annealing time evolution, and its bias dependence are discussed and they are attributed to vacancy defects in the MgO barriers.
No takes yet. Share an insight, caveat, or question.
Stearrett et al. (2010) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: