9220 J. Am. Chem. Soc., Vol. 121, No. 39, 1999
Communications to the Editor
the SER spectrum acquired upon dosing 100 cm3 min-1 of TaEtO-
saturated N2 on Pt at 100 °C. Two SERS bands are discernible
immediately at 1040 and 1165 cm-1. These features were retained
after the reactor was flushed with pure N2 for 10 min to remove
the gas-phase TaEtO. Subsequently dosing with O2, however,
yielded marked spectral changes. The resulting spectrum (Figure
1B) shows the disappearance of the 1040 and 1165 cm-1 SERS
bands, and the advent of features at 575 and 270 cm-1, the latter
peaks growing in within 1-2 min.
The vibrational assignments are facilitated by comparison with
the normal Raman spectra of liquid TaEtO as well as Ta2O5
powder as shown in Figure 1, parts C and D, respectively. The
broadness and ca. 20 cm-1 down-shift exhibited by the SERS
features in Figure 1A upon comparison with the Raman bands of
bulk TaEtO (Figure 1C) imply that the former originates from a
chemisorbed form of ethoxide. While the Raman spectrum of
liquid TaEtO has not apparently been analyzed in detail, it is
similar to that of liquid ethanol.11 The 1040 and 1165 cm-1 SERS
features are tentatively correlated with the corresponding skeletal
ethanol modes at 1065 and 1180 cm-1, thereby providing evidence
for nondissociated ethoxide adsorbate. These SER spectral features
in Figure 1A are also reminiscent of published vibrational spectra
of ethoxide adsorbed on transition metals.12 At least on a
qualitative level, then, ethoxide species are evidently chemisorbed
on platinum, although it is not clear if there is Ta-ethoxide bond
rupture.
1B), obtained upon subsequent O2 dosing at 100 °C in relation
to the bulk-phase Ta2O5 spectrum (Figure 1D) provides strong
evidence that the ethoxide precursor is being transformed into
Ta2O5 under these conditions. It is important to note that no such
spectral features were obtained upon dosing O2 onto unmodified
platinum at 100 °C, heating to 200 °C or above being required to
form PtO at this surface.13 Also, neither adsorbed ethoxide nor
its oxidative conversion to Ta2O5 were evident when performing
parallel experiments on unmodified gold, no discernible SERS
bands being obtained at either stage. This finding clearly
implicates the chemisorbing role of platinum in assisting the
oxidative formation of Ta2O5. It is plausible that the Pt substrate
acts to “trap” the tantalum by triggering partial ethoxide ligand
dissociation, yielding nonvolatile charged metal complexes. The
inability of gold to incur such surface chemistry is unsurprising
in view of its inert character.
Overall, the significance of the present study to the elucidation
of CVD reaction pathways lies in the demonstration that SERS
offers a versatile means of examining both precursor-substrate
interactions and the initial transformation into a monolayer film.14
Indeed, the present SERS results indicate that at least the initial
deposition of Ta2O5 on platinum can occur via a purely “surface”
mechanism, thereby involving adsorbed precursor and oxidant.
Consequently, the elementary chemical steps responsible for such
binary CVD processes do not necessarily occur in the gas phase
prior to deposition, but rather after one or both of the reactants is
adsorbed. However, at present it is not clear that such a surface
reaction as deduced here represents the dominant pathway under
“technological” CVD conditions, whereby both the tantalum
precursor and the oxidant are introduced together. Nevertheless,
it is evident that the capability of imparting SERS activity to a
variety of solid surfaces, achieved by ultrathin film deposition
onto an inert gold template, provides broad-based opportunities
for exploring the fundamental surface chemistry underlying
processes of technological relevance, including microelectronic
materials preparation.
Most significantly, the appearance of the broad SERS envelope
around 500-600 cm-1, along with the 270 cm-1 features (Figure
(6) (a) Zou, S.; Williams, C. T.; Chen, E. K.-Y.; Weaver, M. J. J. Am.
Chem. Soc. 1998, 120, 3811. (b) Zou, S.; WIlliams, C. T.; Chen, E. K.-Y.;
Weaver, M. J. J. Phys. Chem. B 1998, 102, 9039, 9743.
(7) (a) Wilke. T.; Gao., X.; Takoudis, C. G.; Weaver, M. J. J. Catal. 1991,
130, 62. (b) Williams C. T.; Chen, E. K.-Y.; Takoudis, C. G.; Weaver, M. J.
J. Phys. Chem. B 1998, 102, 4785. (c) Chan, H. Y. H.; Williams, C. T.;
Weaver, M. J.; Takoudis, C. G. J. Catal. 1998, 174, 191.
(8) (a) Lo´pez-R´ıos, T.; Sandre´, EÄ .; Leclercq, S.; Sauvain, EÄ . Phys. ReV.
Lett. 1996, 76, 4935. (b) Motte, P.; Lo´pez-R´ıos, T. Thin Solid Films 1997,
295, 5.
(9) Gao, P.; Gosztola, D.; Leung, L.-W. H.; Weaver, M. J. J. Electroanal.
Chem. 1987, 233, 211.
Acknowledgment. This work was supported by the National Science
Foundation.
(10) For example: (a) Kukli, K.; Aarik, J.; Aidla, A.; Siimon, H.; Ritala,
M.; Leskela, M. Appl. Surf. Sci. 1997, 112, 236. (b) Kukli, K.; Ritala, M.;
Leskela, M. J. Electrochem. Soc. 1995, 142, 1670.
JA991707S
(11) Perchard, J.-P.; Josien, M.-L. J. Chim. Phys. Chim. Biol. 1969, 65,
1834, 1856.
(12) (a) Street, S. C.; Gellman, A. J. J. Chem. Phys. 1996, 105, 7158. (b)
Gao, P.; Lin, C.-H.; Shannon, C.; Salaita, G. N.; White, J. H.; Chaffins, S.
A.; Hubbard, A. T. Langmuir 1991, 7, 1515.
(13) Chan, H. T. H.; Zou, S.; Weaver, M. J. J. Phys. Chem. B, submitted.
(14) Note also that SERS is intrinsically sensitive preferentially to the first
few monolayers in the case where thicker films are formed.