In order to aid process development and address extendibility of ionized ph
ysical vapor deposition (IPVD) technology to future integrated circuit gene
rations, an integrated model capable of simulating phenomena across the var
ious length scales characteristic of these systems has been developed. The
model is comprised of a two-dimensional equipment simulation, which relates
process variables to characteristics of material fluxes to the wafer, and
a three-dimensional Monte Carlo based feature scale model. The ion-surface
interaction data required to model the surface processes is generated by a
molecular dynamics based simulation. The integrated model is used to study
the effect of various IPVD process parameters such as wafer bias, coil powe
r, target power, and buffer gas composition on copper film profile inside a
trench. Variations in film profile across the wafer are also examined. It
is found that increasing the wafer bias results in an increase in the mean
ion energy and the amount of sputtering inside the feature. This results in
material transfer from the bottom of the feature to the sidewalls and face
ting of the upper corners of the trench. Two variables, namely the total io
n to Cu flux ratio (R-I/N) and the mean ion energy, are found to play a cru
cial role in determining the effects of coil power and target power. Increa
sing the coil power enhances R-I/N and slightly decreases the mean ion ener
gy. This leads to more sputtering, and therefore a thicker film on the side
walls relative to that on the bottom. Increase in target power causes R-I/N
to decrease, which decreases sputtering within the feature. Film profiles
generally show evidence of enhanced sputtering as buffer gas ionization thr
eshold decreases (He --> Ne --> Ar --> Xe) for the gases considered. These
variations can be explained in terms of two factors: Cu flux ionization fra
ction, which decreases with buffer gas ionization threshold, and mean ion e
nergy, which increases with ionization threshold. (C) 2001 American Institu
te of Physics.