We have examined the formation of copper films alloyed with small quan
tities of palladium. Independent control studies of palladium and copp
er deposition from palladium bis-(hexafluoroacetylacetonate) [Pd(hfac)
(2)] and (hexafluoroacetylacetonato)copper(I)(viny trimethylsilane) [(
hfac)Cu(I)(vtms)], both in the presence and absence of H-2, were carri
ed out. The growth kinetics for both metals were feed-rate-limited und
er similar reactor conditions. No significant variation in deposition
rate (100 nm/min, Pd; 100-500 nm/min, Cu), morphology, resistivity, an
d purity of the copper films was observed due to the addition of H-2.
Simultaneous introduction of both precursors yielded Cu-Pd alloy films
. The absence of pure palladium grains was confirmed by X-ray diffract
ion analysis which showed binary solid solutions (Cu99.5Pd0.5-Cu80Pd20
) as the only crystalline phases. Auger electron spectroscopy analysis
showed a significant reduction in the palladium content of the films
as compared to that expected on the basis of the growth rates obtained
during independent palladium deposition. Go-deposition of copper and
palladium also resulted in a change of the palladium growth kinetics f
rom a feed-rate-limited to a surface-reaction-limited regime (E-a = 16
kcal/mol). The Cu/Pd stoichiometry could be varied by controlling bot
h the Pd-(hfac)(2) partial pressure and substrate temperature. Experim
ents to investigate the cause of the change in Pd CVD kinetics showed
that vinyltrimethylsilane (vtms) severely inhibits Pd growth during in
dependent Pd deposition. The films were also contaminated with C when
vtms was added. This study has shown that alloy CVD kinetics can be dr
astically different from the independent metal deposition kinetics.