surface science ELSEVIER
Surface Science 386 (1997) 89-92
Highly-sensitive Raman spectroscopy to characterize adsorbates on the electrode M. F u t a m a t a * Joint Research Centerfor Atom Technology (JRCAT), National Institute for Advanced Interdisciplinary Research
Received 5 November 1996; accepted for publication 26 February 1997
Abstract Theoretical calculation shows that ATR-surface plasmon-polariton Raman spectroscopy is quite useful to increase Raman scattering intensity from adsorbates even on transition metals with higher intrinsic damping than silver. In fact, a remarkably large enhancement factor of ca. 300 was obtained for Raman band intensity from copper phthalocyanine (CuPc) on platinum, compatible with theoretical values only by assuming parallel orientation of the CuPc molecules on the platinum surface. Parallel orientation was also confirmed by a decreased depolarization ratio of the 1530cm -l (Aag) band for thinner CuPc adlayers. Two different adsorbed states for pyridine on optically flat silver surfaces were observed in an electrochemical environment for the first time. © 1997 Elsevier Science B.V. Keywords: Electrochemical methods; Ion-solid interaction; Polycrystalline surfaces: Raman scattering spectroscopy; Silver;
Solid-liquid interfaces; Vibrations of adsorbed molecules
1. Introduction R a m a n spectroscopy might be the most prominent tool to elucidate the dynamic properties o f m o n o l a y e r adsorbates on a well-defined electrode. However, due to intrinsic limitation in the scattering process, it is in general quite difficult to detect a R a m a n signal from m o n o l a y e r adsorbates [1 ]. Therefore, we have been studying to improve the sensitivity level o f classical R a m a n scattering using the s u r f a c e - p l a s m o n - p o l a r i t o n (SPP) on s m o o t h metal surfaces [2, 3] as well as increasing the optical efficiency in a spectrometer. This technique does not have the disadvantages in the so-called "first * Tel. : (+ 81 ) 298.542725; fax: (+ 81 ) 298 542786; e-mail: [email protected]
0039-6028/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PH S0039-6028 (97) 00333-6
layer" surface enhanced R a m a n scattering, which results from a certain atomic level roughness or from a specific charge transfer interaction. It can be applied relatively easily to well-defined metal surfaces, provided they are optically flat. Theoretical evaluation predicted that even intrinsically strong d a m p i n g metals give a promising enhancement o f the R a m a n b a n d intensity f r o m adsorbates. Therefore, we applied the attenuatedtotal-reflection ( A T R ) m e t h o d to platinum and nickel. Pyridine on a s m o o t h silver electrode was also studied using this under potential control.
2. Experimental The SPP is excited by adjusting the m o m e n t u m o f the incident p h o t o n in the A T R m e t h o d so as
M. Futamata / Surface Science 386 (1997) 89-92
to be in accord with that of the SPP. Fig. 1 shows the optical arrangement for measuring the reflectivity and the Raman spectra of adsorbates on a metal in an electrochemical environment. The Weierstrass prism, made of sapphire or zinc selenide, which is longer by r/n (r, radius; n, refractive
index) than a hemispherical prism, was used in the Otto  or Kretschmann configuration.
3. Results and discussion 3.1. Theoretical calculation f o r various dielectric constants o f the metal
~oL WE RE
Fig. 1. Schematic light scattering configuration under the SPP resonance with the Weierstrass prism (WP). Aplanatic pair of points F and F ' of a sphere of radius r and a central ray for ct~= ~tsPa(~L) is depicted. The platinum surface is irradiated with a laser of 514.5 nm wavelength (1 m W ) through a 1:0.7 (L1) objective. For variation of ct~ and fl~, the position (y) of the prism (P) is adjusted. The scattered light including the SPP cone is focused on the entrance slit (ES) of the monochromator with the second objective (L2).
The local field intensity at the air side of the metal surface was evaluated for various dielectric constants in a prism/air gap/metal configuration. Fig. 2 shows the squared absolute part of the parallel (x) and perpendicular (z) components of the fields under the SPP resonance with respect to the incident field: (1) for various Re £rnetal with a fixed Im emetal=0.342; and (2) for various imaginary parts Im £metal with a fixed real part Re £metal=--10.60. In conclusion, [gzl2 is approximately proportional to IRe Emetall and is inversely proportional to Im emctal for fRe emetad> 5 or Im emct~l>2. In addition, IEzJ2 is approximately proportional to Im emet~ for Re em,t~]=--l.0. Accordingly, the promising enhancement within the incident channel of > 20 is predicted even for highly damping metals, if ]Re emet~l[ or Im em,t~] is sufficiently large. Moreover, out-coupling by the
_- ~400 l--~verall-o i 2 0 0 ~
Ag'Au,Cu Im e
Fig. 2. Absolute local fields squared at metal surfaces (IEil2,j= x, parallel; z, normal to the surface or overall: Igxl2 +lEd 2) under the SPP resonance with respect to that of incident light (tEol2): (a) for various real part values at Im emit,a=0.342, (b) for various imaginary part values at Re e,.~t,~ = - 10.60.
M. Futamata / Surface Science 386 (1997) 89-92
ATR configuration yields a further enhancement of ca. 15 with respect to emission within the external reflection configuration. These results encourage us to apply the ATR method to various metals.
3.2. ATR reflectivity and Raman band intensity for the Weierstrass prism/air gap/CuPc/Pt configuration at various angles of incidence and gap sizes The dielectric constants of the metal and of the solid CuPc, and the thicknesses of the air gap and the CuPc film determine the dispersion of the SPP and the enhancement of the electric field at the metal surface. The in-coupling of the SPP was clearly observed as a sharp minimum of the reflectivity only for p-polarized light as shown in Fig. 3. The observed resonance angle ~sPP°bS=34.5°+_ 0.2 ° and optimum gap size dopt=2.5,~L/4 for tc, p¢=3.5 nm are in good agreement with the theoretical values, ~sPPCalc=34.6°,and dopt= 1.5~L/4, within experimental uncertainty. Raman bands from CuPc, for example, at 1530cm -1 [Alg, v ( C ~ - N~) + ~ ( C ~ - N i l - Ca)], display the maximum intensity at the reflectivity minimum (34.5°). Consequently, the Stokes shifted light of CuPc on platinum is prominently enhanced by the excitation of the SPP using p-polarized light at c
I 1.0- (e)exp.
Table 1 Enhancement factor for Raman band intensity of CuPc on various metals under SPP resonance Wavenumber
Au 6.0b (nm)
1535 Alg 49.9 Theory Random 24.7 Parallel 44.0 Perpendicular 12.6 1457 BI~ 41.6 1346 A1~ 5 5 . 4
23.6 37.0 13.7 20.5 22.5
354 5710 28.9 288 286
1082 9340 52.7 1305 1322
40.0 251 7.5 291 261
26.5 270 4.5 23.7 21.7
aNormalized to the intensity for the external scattering geometry at the same angle of incidence.~l'hickness of CuPc.Cln H20; other case in air.
800 .~ ¢
the same as that for the silver system , despite extremely larger internal damping than in the case of silver: Im ept ( = 1 4 . 5 0 ) > I m eA8 (=0.342) at 514.5 nm. The values are compatible with theoretical prediction only if we assume preferred orientation of CuPc molecules parallel to the Pt surface as shown in Table 1. The depolarization ratio for the 1530 cm- 1 band diminishes with decreasing CuPc thickness: from 0.50 at 17.1 nm to 0.13 at 3.5 nm. Theoretically, the depolarization ratio for this band is 1.00, 0.56 and 0.0 for perpendicular, random and parallel orientation of the molecular
30 40 Angle (deg.)
3.0 - 2.4
, 30 40 Angle (deg.)
_ • 50
Fig. 3. Reflectivity R at p-polarization and unpolarized Raman band intensity I from CuPc at 1530 cm- 1 in the Weierstrass (sapphire) prism/air gap/CuPc (8.6 nm)/platinum system for various incident angles: (a) observed (©, R; e , I) and (b) calculated curves, where IEI2 is the same as in Fig. 2. The gap size dopt is 2.52L/4 (obsd) or 1.52L/4 (calcd).
M. Futamata / Surface Science 386 (1997) 89-92
" i ......:......... ..........................
1040 1000 Wavenumber(cm"1)
Fig. 4. Raman spectra of pyridine on a smooth silverelectrode in 0.1 M LiC104 under potential control, excited at 568.2nm (versus Ag/AgCI). plane with respect to the surface, respectively. Namely, CuPc films become increasingly wellordered with decreasing thickness, in contrast to CuPc films on evaporated silver  or gold films. Thus, the large enhancement factor probably arises from the parallel orientation of CuPc on the platinum surface. 3.3. Application o f A T R - S P P Raman method to Ag/Py/electrolyte interface
Pyridine adsorbed onto a smooth Ag surface on a ZnSe prism shows two distinct adsorbed states,
that is, 1004 and 1014cm -1 bands for a totally symmetric ring breathing mode were observed in 0.1 M LiC104 solution as shown in Fig. 4. The Raman band at 1004 cm-1 is enhanced by charge transfer resonance (electron transfer from Ag to the L U M O (Lowest Unoccupied Molecular Orbital) of Py), which is confirmed by the positive shift of the peak potential, which gives the maximum band intensity, with a shorter wavelength of excitation: from - 0 . 7 0 V (versus Ag/AgC1, at 647.1 nm) to - 0 . 4 5 V (at 568.2 nm). In contrast, the peak potential for the 1014cm -1 band at - 1 . 2 V is due to an increase in the Py adsorbate as evidenced by the cathodic current peak in cyclic voltammogram, and thereby does not depend on excitation wavelength. This suggests that coadsorption of CIO4 ion is critical to decrease the L U M O level of Py so as to become resonant with excitation light, because C104 ions desorb from Ag at ca. - 0 . 7 to - 0 . 8 V. It was supported by similar phenomena observed for different electrolytes, such as KPF6 or KSCN. Details on this point are now under investigation. In conclusion, the ATR-SPP Raman spectroscopy combined with an efficient spectrometer substantially improves the detection limit and provides essential information on the dynamic properties of adsorbates.
 T. Maeda, Y. Sasaki, C. Horie, M. Osawa, J. Electron Spectrosc. Relat. Phenom. 64/65 (1993) 381.  M. Futamata, P. Borthen, J. Thomassen, D. Schumacher, A. Otto, Appl. Spectrosc. 48 (1994) 252 and references therein.  M. Futamata, J. Phys. Chem. 99 (1995) 11901; Langmuir 11 (1995) 3894.