Simulation of localized barrier defects in resonant tunneling diodes

Abstract

Much research has been devoted to understanding the resonant tunneling diode in the last two decades, but some aspects of current conduction in this device are still not well understood. At Texas A&M University, Weichold, et al., have proposed defect assisted tunneling as a possible current mechanism. This study attempts to ascertain the effects of defect potentials in the barriers on current in a simulation of a double barrier resonant tunneling diode. Results indicate that these defects could account for lower than expected peak to valley current ratios in real devices. This study examined the characteristics of the cloud-to-ground (CG) lightning in twenty-three storms as they moved from over land to over water or vice versa. The data were categorized according to the storms' generation mechanism and divided into fifteen minute intervals through the storms. The data were then examined for trends in the CG flash rate, median first stroke peak current and mean multiplicity after the storm crossed the coastline. Linear and logarithmic correlations were calculated for different combinations of the parameters before and after the storm crossed the coastline. The positive and negative CG flash rate changed in a more predictable manner for airmass storms than in frontal storms. In the single airmass storm which moved onshore, the flash rate increased, while in the storms which moved offshore, the negative CG flash rates had a decreasing trend in four of the five cases. The positive CG flash rate decreased in all three cases. In the storms caused by the passage of fronts, the negative CG flash rate was more variable. The negative CG flash rate for frontal storms had an increasing trend in 43% of the cases, while it was decreasing in 57% of the storms. The positive CG flash rate had an increasing trend in 25% of the cases, while it was decreasing in 63% of the cases. The positive CG flash rate remained the same in 12% of the cases. The negative CG flash rate and the mean negative multiplicity were positively correlated in 13 of 16 periods. This is consistent with previous findings. An unexpected result was that the negative CG flash rate and first stroke peak current were positively correlated during the early part of the storms and negatively correlated during the latter part of the storms for 11 of the 13 periods. Both airmass and frontal storms which had less than 10% changes in the negative CG flash rate exhibited consistent changes in the mean negative multiplicity. The median first stroke peak current also decreased in 83% of the storms.