1-s2.0-S0927775701010718-main

Colloids and SurfacesA:Physicochemical and Engineering Aspects202(2002)269–279Metal–metal and metal–semiconductor compositenanoclustersPrashant V.Kamat*,Mark Flumiani1,Amy DawsonNotre Dame Radiation Laboratory,Uni6ersity of Notre Dame,Notre Dame,IN46556-0579,USAReceived5December2000;accepted24August2001AbstractColloidal metal–metal and metal–semiconductor nanocomposites have been synthesized by the chemical reduction of[AuCl4]−on the surface of preformed Silver and TiO2colloids and Ag+on the surface of SnO2and SiO2colloids. Photoinduced bleaching of surface plasmon band has been monitored using a picosecond laserflash photolysis apparatus.In the case of Ag/Au nanoparticles,the magnitude of the surface plasmon bleach of the silver core can be modulated by controlling the capping layer of gold without altering the long term bleaching recovery kinetics.In the case of TiO2/Au nanoclusters,the ratio of[TiO2]/[Au]directly influences the particle size and the stability of these core-shell composite nanoparticles.Aggregation effects were observed for colloids with a TiO2:Au ratio of less than 1:ser pulse excitation(532nm)of these aggregated colloids resulted in the fusion and formation of large nanocomposite clusters.The gold capped TiO2nanoparticles were found to improve the efficiency of interfacial charge transfer process.©2002Elsevier Science B.V.All rights reserved.Keywords:Metals;Semiconductors;Transient bleaching;Composite nanoclusters;Photocatalysis/locate/colsurfa1.IntroductionThe photophysical and photochemical study of single and multicomponent metals and semicon-ductor nanoclusters plays an important role in the development of efficient light energy conversion systems,optical devices and microelectronics(see for example,[1–3]).For example,photoinduced deposition of noble metals such as Pt or Au on semiconductor nanoclusters has often been em-ployed to enhance their photocatalytic activity [4–7].A variety of nanocomposites have been synthesized in recent years.Such composite nanoparticles provides a convenient way to probe the photodynamics of metal–metal and metal–semiconductor interfaces using transient absorp-tion spectroscopy.Examples of core/shell type nanocomposites include,semiconductor/semicon-ductor[2,8–12],semiconductor/metal[4],metal/ semiconductor,[13–15]metal/metal[16–18],and metal/metal oxide[19–21].The surface plasmon absorption band of a metal cluster is dependent both on the cluster*Corresponding author.Fax:+1-219-239-8068;http:/// pkamat.E-mail address:pkamat@(P.V.Kamat).1Mark Flumiani and Amy Dawson are co-op students fromDepartment of Chemistry,University of Waterloo,Canada.0927-7757/02/$-see front matter©2002Elsevier Science B.V.All rights reserved. PII:S0927-7757(01)01071-8P.V.Kamat et al./Colloids and Surfaces A:Physicochem.Eng.Aspects202(2002)269–279 270size/shape and the surrounding media[22–24].Of particular interest are the optical properties of gold and silver nanoclusters that are dependent on the laser pulse excitation.For example,the sur-face plasmon absorption band of gold colloid is bleached in the subpicosecond timescale when excited with a UV or visible laser pulse.The evolution of the optical spectrum of passivated gold clusters in the1.4–3.2nm range( 70–900 atoms)has been demonstrated to be systematic function of size[25].Laser induced fragmentation and fusion of silver and gold clusters has been the topic of interest in many recent studies[26–36]. Depending upon the laser energy,time of irradia-tion and the nature of dispersion controlled the melting of these nanoparticles.Formation of lin-early arrayed gold nanoparticles on gold single-crystal surfaces has been achieved using infrared laser excitation[37].Particle size dependence on the optical limitation properties of gold nanoclus-ters has been investigated by Belloni and cowork-ers[38].The electron dynamics of Au and Ag nanoclusters has also been investigated by tran-sient grating spectroscopy[39]and transient ab-sorption spectroscopy[27,40–42].Recently,in a Au/CdS composite system we were able to induce surface plasmon bleach in the gold core by exciting the external CdS shell[13]. Efforts have been made earlier to synthesize gold capped silver[3]and silver capped gold colloids [17].The absorption spectrum of such composites exhibits spectral features that are significantly dif-ferent than the additive features of the two col-loids.Surface enhanced Raman scattering studies indicated that even submonolayer of silver can modulate specific chemical interactions between the Au and trans-1,2-bis(4-pyridyl)ethylene in Ag capped Au colloids[17].In our continued efforts to explore the optical properties of nanocom-posites we have now investigated the photody-namics of Ag/Au and TiO2/Au nanoclusters (Scheme1).These two examples present two dif-ferent scenarios.In thefirst case we have investi-gated the influence of gold shell on the properties of silver core and in the second case we have probed the influence of a semiconductor(TiO2) core on the properties of a metal(gold)deposit. Synergetics of the metal and semiconductor com-ponents is important for improving the perfor-mance of nanocomposites in photocatalytic applications.2.Experimental section2.1.Preparation of Au and Ag colloidsAu colloids were prepared by the chemical re-duction method of reducing HAuCl4(0.12mM) in water with sodium citrate at near boiling tem-perature[43].Similarly,Ag colloids were prepared using AgNO3as a precursor[27].The concentra-tion of sodium citrate employed for the reduction was kept low(0.42mM)to avoid presence of excess citric acid in the colloidal suspension.2.2.Preparation of Au/Ag colloidsThe composite clusters of Ag/Au were prepared by the sequential reduction of the two metals. Known amount of HAuCl4was added to the preformed silver colloidal suspension and boiled on a hot plate.Additional(equimolar)citric acid was then added and boiled further for about5–10 min.The reduction of gold occurred on the sur-face of silver colloids thus yielding composite metal nanoclusters of core-shell geometry.2.3.Preparation of TiO2colloidal suspensionTiO2colloidal suspension(5mM)was prepared by dropwise addition of2.97ml of10%titanium isopropoxide(Aldrich,99.999%)in1-propanol to Scheme1.Photoexcitation of metal–metal and metal–semi-conductor composite nanoparticles.P.V.Kamat et al./Colloids and Surfaces A:Physicochem.Eng.Aspects202(2002)269–279271 200ml of deionized water with vigorous stirring.Prior to addition of titanium isopropoxide the pHof the water was adjusted to 1.5,with HClO4.The solution was kept stirring in a stoppered glassflask.Freshly prepared colloidal suspension wasused in all the experiments.2.4.Preparation of TiO2/Au colloidsTiO2/Au colloids were prepared by adding thedesired amount of5mM stock HAuCl4(Aldrich)solution to the colloidal TiO2suspension in waterwhile stirring vigorously.The solution was stirredfor an additional5min to allow complete adsorp-tion of[AuCl4]−ions onto the TiO2surface.Reduction of the[AuCl4]−was achieved by thedropwise addition of sodium borohydride(8–10mM)until a color change was observed.Fourdifferent suspensions of TiO2/Au nanocomposite colloids were prepared,by keeping the gold con-centration constant at0.2mM while varying the TiO2concentration from4to0.05mM.These four TiO2/Au suspensions contained[TiO2]:[Au] ratios of20:1,10:1,1:1and0.25:1. Transmission Electron Microscopy(TEM)was carried out by applying a drop of the colloid sample to the carbon-coated copper grid.Particle sizes were determined from the photographs taken at a magnification of150000×using a Hitachi H600transmission electron microscope.Absorp-tion spectra were recorded with a Shimadzu UV-3101PC spectrophotometer.Picosecond laserflash photolysis experiments were performed with355and532nm laser pulses from a mode-locked,Q-switched Continuum YG-501DP Nd:YAG laser system(output1–5mJ per pulse,pulse width 18ps).The white continuum picosecond probe pulse was generated by passing the fundamental output through a D2O/H2O solu-tion.The details of the experimental setup and its operation are described elsewhere[44,45].3.Results and discussion3.1.Ag/Au colloidsSilver colloids prepared by the citrate reduction Fig.1.(A)Ag/Au colloids.Absorption spectra of6×10−4M colloidal Ag in water containing different capping concentra-tions of gold(pathlength2mm).The molar ratio of Ag:Au was maintained at(a)1:0;(b)1:0.0175;(c)1:0.035;(d)1:0.05;(e)1:0.1;and(f)1:0.175.(B)Mixing Ag colloids with Au colloids.Absorption spectra of spectra of2.8×10−4M Ag colloids in water(2mm cell)recorded after the addition of preformed gold particles:(a)0;(b)5×10−6m M;(c)1×10−5 m M;(d)1.5×10−5m M(e)3×10−5m M;and(f)5×10−5m M Au.method produces relatively larger size particles with diameter ranging from40to60nm.The absorption spectrum of silver colloids shows ab-sorption maximum around420nm corresponding to the surface plasmon resonance(Fig.1(A)).This absorption band is rather broad and red-shifted compared with the plasmon absorption band of silver colloids prepared by radiolysis and other methods[46–50].The position and shape of the plasmon absorption of silver nanoclusters are strongly dependent on the particle size,dielectric medium and surface adsorbed species.Efforts have been made earlier to synthesize gold capped silver[3],silver capped gold colloids [17]and bimetallic alloys[51].The absorption spectrum of such composites exhibits spectral fea-tures that are significantly different from the addi-P.V.Kamat et al./Colloids and Surfaces A:Physicochem.Eng.Aspects202(2002)269–279 272tive features of the two colloids.Fig.1(A)shows the absorption spectra of Ag colloids at different Au capping concentrations.As explained in the experimental section,the gold layer was directly deposited on the preformed silver colloids by boil-ing the silver colloid suspension with known amount of HAuCl4and adding equimolar citric acid.We also probed the possibility of interaction between individually stabilized gold and silver colloids by monitoring the absorption properties of mixed systems.The gold and silver colloids were prepared separately using the citrate reduc-tion method.A known amount of gold colloidal suspension was mixed with the Ag colloidal sus-pension to achieve comparable concentrations as in Fig.1(A).The spectra of Ag colloids mixed with different concentration of Au colloids are shown in Fig.1(B).Upon increasing the concen-tration of gold colloids,a small increase in the absorption(500–600nm region)is seen.These spectral differences are merely additive effects and are not the same as those observed with Ag/Au colloids in Fig.1(A).These observations further confirm that the absorption changes observed in Fig.1(A)result from capping of Au shell on Ag colloids.In the case of Ag/Au colloids we also observe a dampening of the surface plasmon absorption band at420nm as the interaction with the gold shell influences the optical properties of the Ag core.Since the Fermi energy of the two metals is similar( 0.5V vs.NHE),we expect a quick equilibration of electrons between the two metals. The surface plasmon absorption band of metal nanoclusters is very sensitive to the surface ad-sorbed species and dielectric of the medium [52,53].Kreibig and coworkers,[22,23,54]who studied the effect of surrounding environment on the plasmon absorption of metal nanoclusters concluded that deviations of the optical properties of metal clusters arise from the predictions of Mie theory.It yields additional information about electronic size effects,about cluster/medium inter-face and about chemical reactions at the interface. For example,one can observe dampening of the surface plasmon band by depositing electrons from radiolytically produced radicals[55]or by exciting the semiconductor deposit in a metal/ semiconductor composite[13].It is interesting to note however,that a thin layer of another noble metal can significantly influence the electronic properties of the bulk metal core.In the present study,we would expect all the gold atoms in thefirst monolayer capping to interact with adjacent silver atoms and form an alloy.Therefore,it is not surprising that we do not observe any surface plasmon absorption cor-responding to gold at this low shell coverage.On the other hand,the small alteration of the surface (alloying with gold)is sufficient enough to dampen the plasmon resonance of silver core. Alloying of Ag and Au systems have also been observed under photoirradiation[56]and room temperature melting conditions.[57]Other inter-esting properties of bimetallic composite particles have also been investigated.For example,a low frequency acoustic breathing mode has been ob-served in the case of Au/Pb colloids[58].Rapid alloy formation has also been noted in the elec-trodeposition of gold onto silver electrodes[59].3.2.TiO2/Au and SnO2/Ag colloids Electrostatic adsorption of metal ions on oppo-sitely charged metal oxide nanoparticles was fol-lowed by the reduction of the metal ions to obtain metal capped semiconductor nanoparticle suspen-sions.The TiO2colloids prepared in acidic medium were positively charged while SnO2col-loids in alkaline medium carried a net negative surface charge.Thus,AuCl4−and Ag+ions bind strongly to the TiO2and SnO2surfaces,respec-tively.Upon reduction with NaBH4we obtain stable TiO2/Au and SnO2/Ag nanocomposite par-ticles in water.The presence of metal oxide col-loid is important for achieving the stability of the suspension.Gold(or silver)reduction carried out in the absence of TiO2(or SnO2)core using same experimental procedure did not produce stable colloids.Fig.2(A)shows the ground state ab-sorbance spectra of TiO2/Au and SnO2/Ag col-loidal solutions at two different ratios of[oxide core]/[metal].We kept the metal ion concentration same and varied the concentration of metal oxide core.P .V .Kamat et al ./Colloids and Surfaces A :Physicochem .Eng .Aspects 202(2002)269–279274Fig.3.Transmission electron micrographs (TEM)of (A)silver (B)gold nanoclusters as prepared by citrate reduction method.(C)Gold capped silver nanoclusters (Ag:Au ratio of 5.7:1).sity in particle size and shape.Most of these silvercolloids exhibit facets,which indicate that thecitric acid reduction method yields fully-grownpared with uncapped silvernanocrystallites,the electron micrograph of Ag /Au sample shows no noticeable increase in size orshape of these crystallites.This in turn indicatesthat the gold layer deposited on the silvernanocrystallites is relatively thin.Even at thehighest gold concentration employed in thepresent study we expect thickness of the gold layerto be of the order of only 5nm.Citric acidreduction of HAuCl 4usually yields 20nm di-ameter spherical particles (Fig.3(B)).Absence ofsuch small spherical particles in the electron mi-crograph of Fig.3(C)rule out formation of sepa-rate gold nanoparticles.As shown in theradiolysis method,the coating of Ag by Au isfacilitated by the fact that their lattice constantsare similar [3].As a result of this lattice matchingthese researchers were not able to distinguish theelectron diffraction pattern of the two metals inAg /Au nanocrystallites.Fig.4shows the electron micrographs of TiO 2/Au and SnO 2/Ag nanoparticles.These compositenanoparticles are spherical in shape.In both thesecases metal oxide core determines the size anddistribution of the nanocomposite.The particlediameter of TiO 2/Au was in the range of 10–15nm while SnO 2/Ag was in the range of 5–10nm.The deposition of the metal layer was relativelysmall since we do not see signi ficant growth in thesize of the composite nanoparticles.3.4.Transient photobleaching of gold capped sil 6er nanoclusters The surface plasmon band of metal particles as explained on the basis of Mie theory involves dipolar oscillations of the free electrons in the conduction band that occupy energy states imme-diately above the Fermi level [22,65].Upon laser pulse excitation,these electrons do not oscillate at the same frequency and thus,cause depletion in the plasmon absorption band.Both Ag and Ag /Au nanoclusters exhibit transient bleaching corre-sponding to the dampening of the plasmon absorption band when subjected to 355or 532nm laser pulse excitation (see for example,spectrum a in Fig.5(A)).This transient bleaching recovers within few nanoseconds.The transient absorption spectra that were recorded 250ps following 355nm laser pulse (pulse width 18ps)excitation are compared in Fig.5(A).The difference absorption spectra of silver and gold capped silver colloids show an intense bleaching of the surface plasmon band at 420nm as they are subjected to laser pulse excita-tion (Fig.5(A),spectrum a).With increasing con-centration of gold capping,the magnitude of transient bleaching increases with a small shift in the bleaching maximum.The maximum bleaching at 440nm is plotted versus the mole fraction of Au in Ag /Au in Fig.5(B).The magnitude of bleaching of the plasmon absorption band in-creases with increasing gold concentration upto the Ag:Au ratio of 1:0.05.Under these conditions,P.V.Kamat et al./Colloids and Surfaces A:Physicochem.Eng.Aspects202(2002)269–279275the bleaching at440nm is nearly twice as that of uncapped silver colloids.At higher gold concen-trations(Ag:Au\1:0.05),the bleaching in the420 nm region decreases and a new bleaching maxi-mum appears around520nm.This new bleaching maximum corresponds to the bleaching of the Au plasmon band.As discussed in the characteriza-tion of ground state spectra of composite clusters, at thicker gold layers the two metal layers attain their individual identity and thus contribute inde-pendently to the plasmon bleach.The results de-scribed here highlight a new approach of modulating transient plasmon bleach of silver col-loids by capping them with another metal shell. It may be noted that the gold colloids at the concentration levels employed here have relatively low absorption compared with Ag colloids at the excitation wavelength.Moreover,the transient spectra recorded following the355nm laser pulse excitation of separately mixed colloidal suspen-sions of Ag and Au colloids did not show any changes in the plasmon bleach.Since these two (Ag and Au)colloids were stabilized separately before mixing,we expect the direct interactions between the two metals to be minimal.The bleaching recovery usually consists of a fast(1–3ps)and a slow( 1ns)components. While the fast recovery corresponds to thermal-ization of hot electrons,the slow recovery has been attributed to electron–phonon and phonon–phonon relaxations.Several studies have dis-cussed in detail the dynamics of laser induced transient bleaching of the plasmon band of Au and Ag nanoclusters[39,41].In the present exper-iments,we have monitored the long term recovery of the transient bleaching at440nm for Ag/Au and510nm for TiO2/Au nanoparticles(Fig.6(A and B)).All the traces in Fig.6(A)had a very similar recovery rate with a time constant in the range of 1.25–1.40ns.Thus,the capping of the gold layer did little or no effect on the long-term bleaching recovery of the surface plasmon bleach in Ag/Au nanoparticles.Following the transient recovery of the surface plasmon band of silver,we observe a growth in the red-IR region.In our previous study[27],we have identified this transient as an aggregate of fragmented silver clusters.We expect such a fragmentation process to dominate under laser excitation conditions.3.5.Transient absorption and photofusion ofTiO2/Au colloidsThe absorption-time profiles recorded at510 nm(Fig.6(B))show the recovery kinetics of theFig.4.Transmission electron micrographs(TEM)of(A)gold capped TiO2(TiO2:Au ratio of10:1)and(B)silver capped SnO2 colloids(SnO2:Ag ratio of50:1).P.V.Kamat et al./Colloids and Surfaces A:Physicochem.Eng.Aspects202(2002)269–279 278spectral feature of this transient match the charac-teristics of the thiocyanate radicals((SCN)2− ) with an absorption maximum at480nm[71,72]. Photocatalytically produced thiocyanate radicals were found to complex with the gold surface(abs. max.390nm),the details of which can be found elsewhere[73].Since the formation of(SCN)2− radicals imme-diately after the laser pulse excitation represents the quantitative estimate of the whole oxidation process,we monitored maximum absorbance at 480nm at different TiO2:Au ratios(the TiO2:Au ratio was varied by varying the gold concentra-tion during sample preparation).The inset shows the dependence of(SCN)2− yield on the gold shell concentration.In the absence of gold capping, TiO2colloids generate(SCN2)− radicals with a quantum yield of0.09.At low concentrations of gold we see an increase in the efficiency of oxida-tion process.For a[Au]:[TiO2]ratio of0.17we see more than40%enhancement in the oxidation efficiency(b(SCN)2− =0.13).As we further in-crease the Au concentration the efficiency of thio-cyanate oxidation at gold capped TiO2 nanoparticles decreases.Inability of the photogen-erated holes to reach the electrolyte interface as well as increased absorption by the gold are the likely reasons for observing lower(SCN)2− yield at higher capping concentrations of gold.On the other hand,when low concentrations of metal is used to cap the semiconductor core we can expect the outer layer to be discontinuous.Such a configuration of core shell particles(i.e.small metal islands deposited on the TiO2core)provide a favorable geometry for facilitating the interfa-cial charge transfer under UV irradiation.The results discussed in this paper highlight some of the interesting properties of the metal–metal and semiconductor–metal nanocomposites. The examples of Ag/Au and Au/TiO2nanoparti-cles show how the composition of the two compo-nents directly influences the photophysical and photocatalytic properties of the metal and semi-conductor nanoclusters.Further studies are un-derway to assess the morphology and surface catalytic properties of nanocomposite materials.AcknowledgementsThe work described herein was supported by the Office of the Basic Energy Sciences of the US Department of Energy.This is contribution No. 4275from the Notre Dame Radiation Laboratory.References[1]A.Henglein,J.Phys.Chem.97(1993)5457.[2]P.V.Kamat,in:P.V.Kamat,D.Meisel(Eds.),Semicon-ductor Nanoclusters—Physical,Chemical and Catalytic Aspects,Elsevier Science,Amsterdam,1997,p.237. [3]P.Mulvaney,M.Giersig,A.Henglein,J.Phys.Chem.97(1993)7061.[4]B.Kraeutler,A.J.Bard,J.Am.Chem.Soc.100(1978)4317.[5]E.Borgarello,R.Harris,N.Serpone,Nouv.J.Chim.9(1985)743.[6]ls,G.Williams,J.Chem.Soc.,Faraday Trans.185(1989)503.[7]E.Amouyal,P.Koffi,J.Photochem.29(1985)227.[8]M.Ocana,W.P.Hsu,E.Matijevic,Langmuir7(1991)2911.[9]A.Haesselbarth,A.Eychmueller,R.Eichberger,M.Gier-sig,A.Mews,H.Weller,J.Phys.Chem.97(1993)5333.[10]I.Bedja,P.V.Kamat,J.Phys.Chem.99(1995)9182.[11]V.F.Kamalov,R.Little,S.L.Logunov,M.A.El-Sayed,J.Phys.Chem.100(1996)6381.[12]A.Mews,A.V.Kadavanich,U.Banin,A.P.Alivisatos,Phys.Rev.B53(1996)R13242.[13]B.Shanghavi,P.V.Kamat,J.Phys.Chem.B101(1997)7675.[14]I.Pastoriza-Santos,D.S.Koktysh,A.A.Mamedov,M.Giersig,N.A.Kotov,L.M.Liz-Marza´n,Langmuir16 (2000)2731.[15]B.Schreder,T.Schmidt,V.Ptatschek,U.Winkler,A.Materny,E.Umbach,M.Lerch,G.Mu¨ller,W.Kiefer,L.Spanhel,J.Phys.Chem.B104(2000)1677.[16]A.Henglein,J.Phys.Chem.B104(2000)6683.[17]R.G.Freeman,M.B.Hommer,K.C.Grabar,M.A.Jack-son,M.J.Natan,J.Phys.Chem.100(1996)718. [18]J.L.Marignier,J.Belloni,M.O.Delcourt,J.P.Chevalier,Nature317(1985)344.[19]wless,S.Kapoor,P.Kennepohl, D.Meisel,N.Serpone,J.Phys.Chem.98(1994)9619.[20]L.M.Liz-Marza´n,M.Giersig,P.Mulvaney,Langmuir12(1996)4329.[21]V.Hardikar,E.Matijevic,J.Coll.Interf.Sci.221(2000)133.[22]U.Kreibig,M.Vollmer,Optical Properties of MetalClusters,Springer,Berlin,1995.P.V.Kamat et al./Colloids and Surfaces A:Physicochem.Eng.Aspects202(2002)269–279279[23]U.Kreibig,M.Gartz,A.Hilger,H.Hovel,in:E.Peliz-zatti(Ed.),Fine Particles Science and Technology, Kluwer Academic Publishers,Boston,1996,p.499. [24]P.Mulvaney,in:P.V.Kamat,D.Meisel(Eds.),Semicon-ductor Nanoclusters—Physical,Chemical and Catalytic Aspects,Elsevier Science,Amsterdam,1997,p.99. [25]M.M.Alvarez,J.T.Khoury,T.G.Schaaff,M.N.Shafig-ullin,I.Vezmar,R.L.Whetten,J.Phys.Chem.B101 (1997)3706.[26]A.Takami,H.Yamada,K.Nakano,S.Koda,Jpn.J.Appl.Phys.35(1996)L781.[27]P.V.Kamat,M.Flumiani,G.Hartland,J.Phys.Chem.B102(1998)3123.[28]H.Kurita,A.Takami,S.Koda,Appl.Phys.Lett.72(1998)789.[29]A.Takami,H.Kurita,S.Koda,J.Phys.Chem.B103(1999)1226.[30]H.Fujiwara,S.Yanagida,P.V.Kamat,J.Phys.Chem.B103(1999)2589.[31]S.Link,C.Burda,M.B.Mohamed,B.Nikoobakht,M.A.El-Sayed,J.Phys.Chem.A103(1999)1165.[32]S.Link,C.Burda,B.Nikoobakht,M.A.El-Sayed,Chem.Phys.Lett.315(1999)12.[33]S.Link,C.Burda,B.Nikoobakht,M.A.El-Sayed,J.Phys.Chem.B104(2000)6152.[34]S.Link,Z.L.Wang,M.A.El-Sayed,J.Phys.Chem.B104(2000)6767.[35]Y.Niidome,A.Hori,T.Sato,S.Yamada,Chem.Lett.,310(2000).[36]C.S.Ah,H.S.Soo Han,K.Kim, D.J.Jang,J.Phys.Chem.B104(2000)8153.[37]K.Murakoshi,Y.Nakato,Adv.Mater.12(2000)791.[38]L.Franc¸ois,M.Mostafavi,J.Belloni,J.-F.Delouis,J.Delaire,P.Feneyrou,J.Phys.Chem.B104,6133(2000).[39]E.J.Heilweil,R.M.Hochstrasser,J.Chem.Phys.82(1985)179.[40]T.W.Roberti,B.A.Smith,J.Z.Zhang,J.Chem.Phys.102(1995)3860.[41]T.S.Ahmadi,S.L.Logunov,M.A.El-Sayed,J.Phys.Chem.100(1996)8053.[42]S.L.Logunov,T.S.Ahmadi,M.A.El-Sayed,J.T.Khoury,R.L.Whetten,J.Phys.Chem.B101(1997) 3713.[43]J.Turkevich,P.L.Stevenson,J.Hillier,Discuss.FaradaySoc.11(1951)55.[44]T.W.Ebbesen,Rev.Sci.Instrum.59(1988)1307.[45]P.V.Kamat,T.W.Ebbesen,N.M.Dimitrijevic, A.J.Nozik,Chem.Phys.Lett.157(1989)384.[46]M.Mostafavi,N.Keghouche,M.-O.Delcourt,J.Belloni,Chem.Phys.Lett.167(1990)193.[47]O.Platzer,J.Amblard,J.L.Marignier,J.Belloni,J.Phys.Chem.96(1992)2334.[48]A.Henglein,Chem.Phys.Lett.154(1989)473.[49]A.Henglein,T.Linnert,P.Mulvaney,Ber.Bunsenges.Phys.Chem.94(1990)1449.[50]T.Linnert,P.Mulvaney,A.Henglein,H.Weller,J.Am.Chem.Soc.112(1990)4657.[51]M.Treguer,C.de Cointet,H.Remita,J.Khatouri,M.Mostafavi,J.Amblard,J.Belloni,R.de Keyzer,J.Bel-loni,J.Phys.Chem.B102(1998)4310.[52]T.Linnert,P.Mulvaney,A.Henglein,J.Phys.Chem.97(1993)679.[53]M.Gutierrez, A.Henglein,J.Phys.Chem.97(1993)11368.[54]B.N.J.Persson,Phys.Rev.B39(1989)8220.[55]A.Henglein,P.Mulvaney,T.Linnert,Faraday Discuss.92(1991)31.[56]J.Hodak,A.Henglein,M.Giersig,G.V.Hartland,J.Phys.Chem.B104(2000)11708.[57]T.Shibata, B.Bunker,Z.Zhang, D.Meisel,J.Phys.Chem.B105,submitted(2001).[58]J.Hodak,A.Henglein,G.V.Hartland,J.Phys.Chem.104(2000)5053.[59]E.Schmidt,S.Stucki,J.Electroanal.Chem.39(1972)63.[60]R.D.Averitt,S.L.Westcott,S.J.Oldenburg,T.R.Lee,N.J.Halas,Langmuir14(1998)5396.[61]H.Giesche,E.Matijevic,J.Mater.Res.9(1994)436.[62]R.D.Averitt,S.L.Westcott,N.J.Halas,J.Opt.Soc.Am.B Opt.Phys.16(1824)1999.[63]S.J.Oldenburg,J.B.Jackson,S.L.Westcott,N.J.Halas,Appl.Phys.Lett.75(1999)2897.[64]O.V.Makarova,A.E.Ostafin,H.Miyoshi,J.R.Norris,D.Meisel,J.Phys.Chem.B103(1999)9080.[65]G.Mie,Ann.Physik B25(1908)377.[66]A.Heller,Pure Appl.Chem.58(1986)1189.[67]Y.Nosaka,Y.Ishizuka,H.Miyama,Ber.Bunsenges.Phys.Chem.90(1986)1199.[68]Y.Chen,Z.Wei,Y.Chen,H.Lin,Z.Hong,H.Liu,Y.Dong,C.Yu,W.Li,J.Mol.Catal.21(1983)275.[69]Y.Nakato,H.Tsubomura,Isr.J.Chem.22(1982)180.[70]N.R.de Tacconi,J.Carmona,K.Rajeshwar,J.Phys.Chem.B101(1997)10151.[71]P.V.Kamat,Langmuir1(1985)608.[72]D.Behar,P.L.T.Bevan,G.Scholes,J.Phys.Chem.76(1972)1537.[73]A.Dawson,P.V.Kamat,J.Phys.Chem.B104(11842)2000.。