A. Kudelski, B. PettingerrChemical Physics Letters 321 (2000) 356–362
361
plained by the creation of ‘electromagnetic active
sites’ existing only for a short time. Such temporary
electromagnetic ‘hot spots’, possibly created during
reconstruction of silver or gold surfaces, should be
similarly efficient for different molecules. Hence,
there should also be fluctuations of the pyridine
bands. The absence of spectral fluctuations for pyri-
dine and large spectral fluctuations for carbon chain
segments is striking and allows far-reaching conclu-
sions: Ži. the metallic structures are stable in the
superior surface enhancement for species adsorbed at
so-called ‘hot spots’. Due to their inhomogeneous
structure, the number of carbonaceous groups associ-
ated with them must be limited to produce a low
inhomogeneous spectral broadening. Most impor-
tantly, the spectral fluctuations indicate continuously
progressing surface chemistry at rather small and
discrete locations, which – when averaged in time –
is still representative for the overall chemistry occur-
ring at the sample.
sense that they provide stable electromagnetic field
enhancements varying locally along the surface. Žii.
The pyridine molecules also form and break bonds to
the metal surface. They may adsorb to different sites,
but this has a minor effect on frequencies and inten-
sities and, therefore, their SERS signal is constant.
Žiii. The carbon network can change its local chemi-
cal nature, for instance by incorporating oxygen
atoms. This affects its local electronic configuration
and the observable frequencies. Among the large
manifold of interfacial configurations, some of them
will tune into, others out of an optical resonance.
Given a sufficient resonance, a local carbon chain
segment behaves like an adsorbed dye. It will exhibit
an extremely large cross-section for Raman scatter-
ing and show up in the SERS spectrum.
Acknowledgements
The authors thank Prof. G. Ertl for helpful discus-
sion and financial support. A.K. gratefully acknowl-
edges financial support by the Max-Planck-Gesel-
lschaft.
References
w1x B. Pettinger, in: J. Lipkowski, P.N. Ross ŽEds.., Adsorption
of Molecules at Metal Electrodes, Chap. 6, VCH, New York,
1
992.
w2x J.E. Pemberton, in: H.D. Abruna ŽEd.., Electrochemical In-
terfaces: Modern Techniques for In-Situ Interface Characteri-
By decreasing the incident laser power by a factor
of 100 only the general SERS intensity level is
reduced, but not the rate of fluctuations for experi-
ments carried with 50= and 100= objectives. This
intensity reduction is higher than for a 10= objec-
tive using 100% incident laser power. In the latter
case, the fluctuations are substantially diminished,
but the intensity at the sample drops only by a factor
of 25. Obviously, the power density and a possibly
correlated local increase in the surface temperature
are not the key factors that determine the occurrence
of fluctuations. In fact, our preliminary studies on
the sample temperature influence on the rate of
fluctuations reveal only a minor or negligible effect
in the investigated region Ž297–490 K.. However,
zation, Chap. 5, VCH, New York, 1991.
w3x J.A. Creighton, M.G. Albrecht, R.E. Hester, J.A.D. Matthew,
Chem. Phys. Lett. 55 Ž1978. 55.
w4x R. Lombardi, R.L. Birke, L.A. Sanchez, I. Bernard, S.C.
Sun, Chem. Phys. Lett. 104 Ž1984. 240.
w5x B. Pettinger, J. Chem. Phys. 85 Ž1986. 7442.
w x
6
B. Pettinger, K. Krischer, G. Ertl, Chem. Phys. Lett. 151
Ž1988. 151.
w7x L. Moerl, B. Pettinger, Solid State Commun. 43 Ž1982. 315.
w8x B. Pettinger, L. Moerl, J. Electron. Spectrosc. Relat. Phenom.
2
9 Ž1983. 383.
w9x D.P. Tsai, J. Kovacs, Z. Wang, M. Moskovits, V.M. Shalaev,
J.S. Suh, R. Botet, Phys. Rev. Lett. 72 1994 4149.
w10x P. Zhang, T.L. Haslett, C. Douketis, M. Moskovits, Phys.
Rev. B 57 Ž1998. 15513.
Ž
.
w11x S.I. Bozhevolnyi, B. Vohnsen, I.I. Smolyaninov, A.V. Zay-
ats, Opt. Commun. 117 Ž1995. 417.
w12x S. Nie, S.R. Emory, Science 275 Ž1997. 1102.
because of the random character of this phenomenon
a more advanced statistical analysis is required.
13 J.T. Krug, G.D. Wang, S.R. Emory, S. Nie, J. Am. Chem.
w x
Soc. 121 Ž1999. 9208.
w14x S.R. Emory, W.E. Haskins, S. Nie, J. Am. Chem. Soc. 120
Ž1998. 8009.
4
. Conclusion
w15x K. Kneipp, Y. Wang, H. Kneipp, L.T. Perlman, I. Itzkan,
R.R. Dasari, M.S. Feld, Phys. Rev. Lett. 78 Ž1997. 1667.
The appearance and variance of a small number
w16x
K. Kneipp, Y. Wang, R.R. Dasari, M.S. Feld, Appl. Spec-
of narrow Raman lines can only be explained by a
trosc. 49 Ž1995. 780.