SEM photograph showing SIVs formed underneath a buried wire after annealing.
(a) Schematic of a trench before a Cu wire was buried in it. (b) SEM image of the trench with 500 nm height and 100 nm width before plating.
Cross-sectional SEM images of specimens with a 700 nm thick overlayer after annealing at 673 K for 30 min and cooling at 3 K/min. The wire widths were (a) 100, (b) 140, and (c) 220 nm.
SIV occurrence ratio as a function of cooling rate. The overlayer thickness was 700 nm and four wire widths were used.
SIV occurrence ratio as a function of overlayer thickness. The cooling rate was 25 K/min and four wire widths were used.
Schematic of the calculation model.
Specimen setting process.
Atomic structure of the reference system calculated under the conditions shown in Table I.
Effects of the wire width and heat-treatment temperature on void formation.
Effects of the strain and heat-treatment temperature on void formation. A negative sign for strain means compressive strain and a positive sign means tensile strain.
Effects of the thickness of overlayer and annealing temperature on void formation.
Starting temperature of void formation during the cooling process as a function of cooling rate.
Schematics of the void formation model of the buried wire. (a) Large local strain is present at four trench corners in the buried wire before heat treatment. (b) Structural relaxation to strengthen adhesion between wire and substrate. (c) Reduction in surface area to minimize surface energy.
Compression strain suppresses the formation of voids while stretched strain favors their formation.
Schematics to explain the effects of dimensions of buried wire on void formation.
Schematics of the effects of overlayer thickness on void formation.
Calculation conditions for the reference system.
Calculation conditions for the Cu wire/Ti substrate system.
Chemical potentials of the two specimens with thin or thick overlayer.
Article metrics loading...
Full text loading...