TEM images of (a) 90 nm and (b) 65 nm technology interconnects.
Lifetime distributions for down-flow WSB devices for 90 and 65 nm technologies at , (a) raw data and (b) deconvoluted data.
Arrhenius plots for 90, 65, and 45 nm WSB data.
Scaling behavior of EM early failure lifetimes across 90, 65, and 45 nm technologies. The parameter represents the via diameter and the line height.
Slit-shaped void at the top of an M1 line, connected to contacts. This interconnect did not fail, however, the same void could have lead to failure for the V1/M1 down-flow case.
Trend of critical current density-length products, , as a function of dielectric. With mechanically weaker films, the values decrease. The encircled area encompasses TEOS/FTEOS-based integrations. The 130 nm-hybrid integration uses FTEOS at the via level and SiCOH at the trench level (red datapoints from , single asterisk , double asterisk ).
Down-flow WSB EM failure distributions for eleven current density conditions between 0.5 and . The two lowest current density conditions show a pronounced increase in the lognormal sigma.
Inverse median lifetimes as a function of in the lower current density regime.
Inverse median lifetimes as a function of in the entire current density regime tested, with various current density exponent fits. The case provides the best fit.
Sample size across 90, 65, and 45 nm technology nodes. All samples from part I and part II of this publication are included here.
Article metrics loading...
Full text loading...