16 Adv Mat Review

REVIEW

Exploitation of Localized Surface Plasmon Resonance** By Eliza Hutter and Janos H. Fendler* Recent advances in the exploitation of localized surface plasmons (charge density oscillations confined to metallic nanoparticles and nanostructures) in nanoscale optics and photonics, as well as in the construction of sensors and biosensors, are reviewed here. In particular, subsequent to brief surveys of the most-commonly used methods of preparation and arraying of materials with localized surface plasmon resonance (LSPR), and of the optical manifestations of LSPR, attention will be focused on the exploitation of metallic nanostructures as waveguides; as optical transmission, information storage, and nanophotonic devices; as switches; as resonant light scatterers (employed in the different near-field scanning optical microscopies); and finally as sensors and biosensors.

1. Introduction Localized surface plasmons (LSPs) are charge density oscillations confined to metallic nanoparticles (sometimes referred to as metal clusters) and metallic nanostructures (Fig. 1).[1] Excitation of LSPs by an electric field (light) at an incident wavelength where resonance occurs (Fig. 1) results in strong light scattering, in the appearance of intense surface plasmon (SP) absorption bands, and an enhancement of the local electromagnetic fields. The frequency (i.e., absorption maxima or color) and intensity of the SP absorption bands are characteristic of the type of material (typically, gold, silver, or platinum), and are highly sensitive to the size, size distribution, and shape of the nanostructures, as well as to the environments which surround them.[2,3] These are the precise properties which have prompted the ongoing intense interest in LSPs and fueled the construction of LSP-based sensors and devices in ever increasing variety. The intense red color of aqueous dispersions of colloidal gold particles is, of course, a manifestation of localized surface plasmon resonance. Colloidal gold has fascinated people for centuries. It provided color for medieval cathedral windows and until the eighteenth century was believed to have life-

±

[*] Prof. J. H. Fendler, Dr. E. Hutter, Department of Chemistry and Center for Advanced Materials Processing Clarkson University, Potsdam, NY 13699 (USA) E-mail: [email protected] [**] We thank the US National Science Foundation (E. H. Grant No. INT-0206923) and the Department of Energy (JHF) for their financial support.

Adv. Mater. 2004, 16, No. 19, October 4

Figure 1. Top: Schematics for plasmon oscillation for a sphere, showing the displacement of the conduction electron charge cloud relative to the nuclei. Reproduced, with permission, from reference [106]; copyright 2003, the American Chemical Society. Bottom: Field lines of the Poynting vector (excluding that scattered) around a small aluminum sphere illuminated by light of energy 8.8 eV where resonance occurs (left hand side) and 5 eV where there is no resonance (right hand side). Reproduced with permission from [109].

prolonging and rejuvenating benefits if taken internally as aurum potabilis.[4] In our times, there has been an exponential growth in the exploitation of colloidal gold for biological labels, markers, and stains for various microscopies.[5] More recently, metallic nanoparticles and nanostructures have been fruitfully employed as molecular-recognition elements and amplifiers in sensors and biosensors, in addition to serving as components in nanoscale optical devices.[6] It is these more recent and significant developments which will be detailed in the present review.

DOI: 10.1002/adma.200400271

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1685

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Following brief surveys of the most commonly used current methods of preparation and arraying materials with localized surface plasmon resonance (Sec. 2), and of the optical manifestations of localized surface plasmon resonance (Sec. 3) we shall discuss recent advances in the construction of nanoscale optics and photonics and the exploitation of metallic nanostructures as waveguides, switches, superlenses, information storage devices, and as resonant light scatterers in the different near-field scanning optical microscopies, NSOMs (Sec. 4). Finally, we shall summarize the use of metallic nanoparticles and nanostructures as sensors and biosensors (Sec. 5). We shall not cover the vast area of propagating surface plasmons, PSPs, usually referred to as just surface plasmons, SP (i.e., the longitudinal charge density oscillations at a flat 40±200 nmthick metal film, typically gold or silver, deposited onto a dielectric such as glass) and their applications in surface plasmon resonance, SPR, spectroscopy. There are several recent reviews which cover these topics.[7±9] Attention in this review is focused on LSPs as manifested by isolated, coupled, and arrayed metallic nanoparticles and nanostructures. We shall, however, summarize the beneficial amplification of SPR signals by LSPs.

2. Preparation and Arraying of Materials with Localized Surface Plasmon Resonance As the title of this section indicates, we consider in our review all materials that, upon suitable excitation, demon-

strate localized surface plasmon resonance. Accordingly, we shall survey the recent advances in the preparation and stabilization of metallic nanoparticles by colloid chemical methods (Sec. 2.1); fabrication of ultrathin metallic nanoislands by vacuum evaporation; lithographically nanopatterned structures, including the arraying of nanoparticles and the construction of periodic particle arrays, PPAs (Sec. 2.2); and nanostructures prepared by laser ablation (Sec. 2.3) and electrodeposition (Sec. 2.4).

2.1. Recent Advances in the Colloid Chemical Preparations of Metallic Nanoparticles Traditionally, metallic nanoparticles are prepared by the controlled precipitation and concurrent stabilization of the incipient colloids.[10,11] In homogeneous solution this is accomplished by the judicious adjustment of the precipitating conditions (type, concentration, order and rate of additions of the reagents and stabilizers, temperature, and solvent). The approach is illustrated by the classical synthesis of 12 nm diameter gold nanoparticles by the sodium citrate reduction of gold chloride in an aqueous solution.[12] The nanoparticles formed are coated by negatively charged electrical double layers (composed of bulky citrate ions, chloride ions, and the cations attracted to them) which provide required stabilization. Recipes for the preparation of colloidal gold nanoparticles in diameters ranging from 0.82 nm to 64 nm are provided by Handley.[13]

Eliza Hutter was born and raised in Hungary. She obtained an MD degree at the Medical Institute in Leningrad (now St. Petersburg) in 1991 and launched her scientific career in the Department of Virology at the B. Johan National Institute of Public Health in Budapest, Hungary. In 1995 she obtained a postdoctoral position at SUNY Health Science Center, Syracuse, NY, in a research group whose attention was focused on the influence of the enzymes of pentose±phosphate pathway on programmed cell death. Subsequently, she obtained a Ph.D. degree in Chemistry at Clarkson University (2001). Her thesis work consisted of construction of ordered two-dimensional and threedimensional nanostructures and studying the molecular interactions therein. She is currently an NSF International Research Fellow. Her research interests include DNA films, DNA-derivatized nanoparticles, and the development of more sensitive surface-plasmon-resonance-based detection devices.

Janos H. Fendler is the Distinguished CAMP Professor of Chemistry at Clarkson University, New York, USA. He received his B.Sc. from the University of Leicester (England) and his Ph.D. and his D.Sc. from the University of London. He used surfactant assembliesÐaqueous and reversed micelles, Langmuir monolayers, Langmuir±Blodgett films, vesicles and polymerized vesicles, and bilayer lipid membranesÐfor molecular organization and compartmentalization. His research has been reported in more than 306 primary publications and summarized in 96 review articles and three books. He has received several awards, including the Langmuir Distinguished Lecturer award and the ACS National Award for Colloid and Surface Chemistry. 1686

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4

Advantage has also been taken of ultrasonic power to form metallic nanoparticles, complex nanoparticles, and metallic nanoparticles deposited on silica spheres.[14±19] Recent reports on controlling sizes and size distributions of gold and silver nanoparticles are particularly significant.[20,21] Stabilization of metallic nanoparticles by covalently attached thiol- and disulfide-functionalized monolayers represented a significant milestone.[22±24] These nanoparticles, stabilized by covalent metal±sulfur bonds, became known as monolayer-protected clusters, MPCs.[25] The unique feature of MPCs is that they behave as if they were large individual macromolecules. Indeed, MPCs can be isolated from their dispersions as solids and redispersed in a solvent without structural or chemical alterations. Importantly, highly monodispersed populations of MPCs can be prepared by fractional crystallization.[26] MPCs can also be functionalized and subjected to all the synthetic procedures (without breaking the Au±S bonds) known to chemists. The relative ease of preparation of MPCs in substantial quantities is equally important, since it greatly facilitates characterizations by such standard methods as elemental analysis, differential scanning calorimetry, NMR spectroscopy (both solution and solid-state), absorption, and Fourier-transform infrared (FTIR) spectroscopy.[27,28] A single-phase, one-pot approach has been recently reported for the gram-scale preparation of gold MPCs in controllable sizes and size distributions.[29,30] Thiol-functionalized metallic (Au, Ag, Pt, Pd) and composite metallic (Au/Ag, Au/Co, Ag/Co, Au/Pt, Au/Pd, and more complex bi- and tri-metallic core± shell-like) particles have also been reported.[25,31±34] Slow evaporation of gold MPC dispersions have been shown to lead to the spontaneous formation of highly ordered three-dimensional superlattices.[35,36] Covalent stabilization of metallic nanoparticles is generally restricted to the use of metal±thiol bond. Alternatively, they can be stabilized by enclosing them in non-metallic shells (latex, silica, polystyrene), thus providing a chemically different surface for the immobilization of molecules without losing the spectral features of the metal core.[37,38] Metallic nanoparticles are also stabilized if they are in-situ generated, trapped in (or on) templates, or confined in the restricted volumes of nanoreactors.[39] Such surfactant aggregates as aqueous micelles, reversed micelles, microemulsions (both oil-in-water, O/W, and water-in-oil, W/O), surfactant vesicles, and polymerized surfactant vesicles have been fruitfully employed as templates (and/or nanoreactors) for the insitu generation of nanoparticles.[39,40] Alternatively, preformed nanoparticles have also been incorporated into surfactant aggregates. Some degree of size and shape control has been achieved by generating metallic nanoparticles in reversed micelles.[41±43] Ultimate stabilization is achieved, of course, by the prompt covalent capping of the nanoparticles generated in reversed micelles. Polymers have also been used as templates for the in-situ generation or incorporation of metallic nanoparticles. Particularly significant is the versatile use of dendrimers[44,45] and di- and triblock copolymers.[46±53] Judicious selection of appro-

Adv. Mater. 2004, 16, No. 19, October 4

priate block copolymers in appropriate concentrations have resulted in the formation of large variety of self-assembled structures including those which resemble aqueous micelles, reversed micelles, microemulsions, and surfactant vesicles. All of these structures can, in principle, be used as templates for nanoparticle generation or incorporation. It should be remembered that partial entrapment of nanoparticles in rigid compartments obviates the need for surface coverage (i.e., modification). Metallic nanoparticles can be, therefore, examined (and employed) in their pristine states. Preparation of metallic nanoparticles by soft solution processing (i.e., direct fabrication in aqueous solution at moderate temperatures and pressures) has gained increasing acceptance. Hydrothermal synthesis (hydrothermal technique) is a soft solution processing which involves the heterogeneous reaction between powdered solids and water above ambient temperature and at pressure greater than 1 atm in a closed system.[54] The relative ease (appropriate powdered reagents and water are placed in a teflon-lined autoclave and heated without stirring at moderate to high temperatures and pressures for the desired time) and the possibility of predicting optimum reaction conditions by electrolyte thermodynamics (in terms of phase diagrams) are the advantages of the hydrothermal synthesis.[55] Spherical metallic (mostly gold) nanoparticles have been converted to nanorods, with controllable aspect ratios, by a seed-mediated sequential growth process.[56±60] Citrate-ionstabilized nanoparticles were used as seeds in a series of gold chloride solutions in a hexadecyltrimethylammonium chloride aqueous micellar solution containing ascorbic acid as a reducing agent.[56±60] The technique was also adapted for growing gold nanorods on mica surfaces (to several hundred nanometers length, with diameters in the range of 10±20 nm).[61] Gold nanoshells have been synthesized by reacting aqueous HAuCl4 solutions with solid templates such as silver nanoparticles. The morphology, void space, and wall thickness of these hollow nanostructures were all determined by the templates, which were completely converted into soluble species during the replacement reaction.[62] Replacement reactions between silver nanocubes and aqueous gold chloride provided a versatile method for generating nanostructures with hollow interiors.[63]

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

2.2. Vacuum Deposition and Electron-Beam Lithography Vacuum deposition or sputtering involves the condensation of atoms produced by evaporation of a metal shot placed on a heating element.[64] The thickness of the metal film is monitored by an in-situ quartz crystal microbalance. Ultrathin (less than 10 nm nominal thickness) evaporated gold films have well-defined island structures and optical properties similar to those of colloidal gold nanoparticles.[65±67] Figure 2 shows typical transmission electron microscopy (TEM) images of gold nanoislands obtained by the slow (0.0014±0.0028 nm s±1) evap-

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1687

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Figure 2. A typical TEM image of ultrathin gold nanoislands (2.5 nm nominal thickness), evaporated onto Formvar-covered, carbon-coated, 200-mesh copper grids at a rate of 0.0014±0.0028 nm s±1. The gold nanoislands were found to be slightly elongated; therefore the two axes are separately shown in the inserted histogram. Reproduced with permission from [68]; copyright 2003, the American Chemical Society.

oration of gold onto Formvar-covered, carbon-coated, 200 mesh copper grids.[68] Preparation of arrayed, triangular silver nanoparticles (referred to as periodic particle arrays, PPAs) is a combined colloid chemical and vacuum deposition technique. The approach involves the drop coating of a substrate by a monolayer of monodispersed polystyrene particles, which serves as a deposition mask through which the silver is evaporated (in a vacuum evaporator) to a desired thickness. Removal of the polystyrene particles by sonication in ethanol leaves the triangular silver nanoparticles patterned on the substrate surface.[69±73] Electron-beam (EB) lithography is physicists' first choice for fabricating arrays consisting of monodispersed (typically better than 99 %) metallic nanoparticles in desired sizes, shapes, patterns, and interparticle distances on glass substrates. Typically, this process involves the following steps: 1) spin-coating of a conducting substrate (either an indium tin oxide (ITO) glass or a glass slide coated by a ca. 10 nm thick gold film) with a 70±100 nm electron-sensitive resist, usually poly(methyl methacrylate), PMMA); 2) forming the desired pattern by EB lithography (which burns off the polymer and the thin gold film); 3) chemical development of the exposed PMMA; 4) thermal evaporation (at a rate of 1±2 nm s±1) of the gold (or silver) film; 4) removal of the remaining PMMA film by acetone (and the 10 nm gold film, if present, by etching using an aqueous solution of KI and I2 (4.0 g and 1 g respectively in 15 mL water); 5) imaging the pattern by scanning electron microscopy (SEM) and/or atomic force microscopy (AFM).[74±77] Figure 3 shows a typical pair-wise interacting gold nanoparticles fabricated by EB lithography.[76] 1688

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Figure 3. SEM images of identical 150 nm diameter gold nanoparticles deposited by EB-lithography to heights of 17 nm at 450 nm (a), 300 nm (b), and 150 nm (c) center-to-center distance. Reproduced with permission from [76]; copyright 2003, Elsevier.

Difficulty of scale-up, beyond a few microscopic slides at a time, and the expenses involved are the disadvantages of EB lithography.

2.3. Laser Ablation Formation of metallic nanoparticles by the photo- and radiolytic reduction of metal ions has been the subject of extensive investigations since the early 1960s.[39,78±80] In fact, much of our understanding of how metallic clusters evolve has been obtained by the pulse-radiolytic investigations of the sequential transient absorbances formed in the reduction of silver cations.[80,81] Absorbances appearing initially were ascribed to Ag0, Ag2+, and Ag4+; longer irradiation led to formation of non-metallic clusters and ultimately to the appearance of a surface plasmon absorption band indicating the transformation to metallic silver nanoparticles.[80,81] Laser ablation (usually using a pulsed laser with controlled energy and pulse duration) of metal plate targets immersed in aqueous solution, containing NaCl or surfactants or cyclodextrin, resulted in the formation of metallic nanoparticles[82±90] and composite alloyed nanoparticles (Au±Ag,[91] for example).

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Transformation of spherical metallic nanoparticles to nanorods,[92,93] nanodisks,[94] nanoprisms,[95] and nanonetworks[86] by light (and/or laser beams) at selected wavelength is quite interesting.[96] The shape control has been rationalized by assuming the fastest growth of a given shape of nanoparticle that has the largest absorption coefficient at the rotationally averaged irradiation wavelength.[94]

2.4. Electrodeposition Electrochemical deposition of nanoparticles (or nanostructured films) is attractive since a high degree of control can be achieved by the judicious employment of Faraday's laws (using 96 500n Coulombs of electricity results in the deposition of 1 g mol±1 of materials, where n is the number of electrons passed in depositing one mole of the deposit). Furthermore, composition, defect chemistry, and shape (along with thickness) are controllable at a reasonable cost. Anodic depositions are limited to metallic substrates, since the process involves the electrolytic reduction of metal in the presence of the appropriate anions. Gold and palladium MPCs were prepared on silicon surfaces, for example, by the galvanostatic reduction (using 0.5 mA cm±2 current) of HAuCl4 and PdNO3 in methanol, in the presence of dodecanethiol.[97,98] Galvanic replacement reactions have been used for alloying and dealloying silver and gold nanoparticles as well as for preparing nanoshells.[99] Sequential electrodeposition within a porous template was used to prepare multimetal microrods intrinsically encoded with submicrometer stripes. Variations in composition along the length of the wire permit the incorporation of electrical functionality, optical contrast, and/or desired surface chemistry.[100,101]

3. Optical Manifestations of Localized Surface Plasmon Resonance Well-separated metallic nanoparticles and nanostructures, with dimensions significantly smaller than the wavelength of the exciting light, are characterized by a broad, intense absorption band in the visible range of the spectrum. The bandwidth, the peak height, and the position of the absorption maximum depend markedly, as stated in the Introduction, on the size, size distribution, surface state, surface coverage, and surrounding environment of the given nanoparticles and nanostructures.[2,3,102] The significant difference in the absorption spectra of the same concentration of ethylenediaminetetraacetic acid (EDTA)-stabilized, 22.8 ± 5.8 nm and 37.8 ± 9.0 nm diameter silver nanoparticles illustrates the point well (Fig. 4).[103] Varying the dielectric constant of the media introduced between and above silver nanoislands, deposited onto quartz and sapphire substrates, changed the absorption maximum of the LSPR band from 482 nm to 1310 nm.[104] Adv. Mater. 2004, 16, No. 19, October 4

Figure 4. Absorption spectra of 22.8 ± 5.8 nm and 37.8 ± 9.0 nm silver nanoparticles showing characteristic surface plasmon bands at 406 and 428 nm, respectively. Reproduced with permission from [103].

It was Gustav Mie who first rationalized the observed intense color of dispersed colloidal gold particles in 1908.[105] Relative simplicity and versatility are the advantages of the classical Mie theory, which assumes that the particle and the surrounding medium are homogeneous and can be described by bulk optical dielectric functions.[1,106] Solving Maxwell's equations lead to a relationship for the extinction cross-section, rext (rext = rabs + rsca = absorption cross-section + scattering cross-section) for metallic nanoparticles as a summation over all electric and magnetic oscillations. For nanoparticles which are significantly smaller than the wavelength of the exciting light (k >> 2R, where R is the nanoparticle radius) the Mie theory is reduced to:[1]

rext …x†ˆ 9

e2 …x† x e V c 3=2 O ‰e …x†‡2em Š2 ‡e …x†2 2 1

(1)

where VO = (4p/3)R3, x is the angular frequency of the exciting radiation, em is the dielectric function of the medium surrounding (or embedding) the metallic nanoparticles, and e1 and e2 are the real and imaginary part of the dielectric function of the metallic nanoparticles, respectively. As indicated in Equation 1, the surface plasmon absorption band appears when e1(x) » ±2em if e2(x) is small or if it is only weakly dependent on x. The bandwidth and peak height are well approximated by e2(x); however, contrary to the experimental evidence, no size dependency of the peak position is predicted by Equation 1. The size dependency of the position of the surface plasmon absorption band of metallic nanoparticles and nanostructures can be accommodated by assuming size-

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1689

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

1690

dependent dielectric functions.[26,107] The observed spectral changes that accompany the electrochemical charging and discharging of metallic nanoparticles, as well as metal adatom deposition, and alloying and formation of bimetallic core± shell type nanoparticles, have been rationalized by applying Drude's theory for describing free-electron behavior in the Mie equation.[3,108] Equation 1 does not adequately describe the optical behavior of larger (2R ³ 30 nm) metallic nanoparticles. For such nanoparticles the extinction cross-section is dominated by higher-order multipole absorption and scattering, and the full Mie equation needs to be used to model the absorption spectra.[109,110] Surface plasmons are unevenly distributed around nonspherical metallic nanoparticles and nanostructures and this manifests in shape-dependent LSPR absorption spectra.[111] In particular, for metallic nanorods the plasmon resonance splits into low- and high-energy absorption bands. The high-energy, or transverse, absorption band corresponds to the electron oscillations perpendicular to the major axis; while the low-energy, or longitudinal, absorption band results from the oscillation of the electrons along the major axis. As the aspect ratio of the nanorods increases, separation between the two plasmon bands becomes more pronounced (Fig. 5).[96,112] Significantly, triangular particles exhibit multiple plasmon resonance, a longitudinal (bulk) plasmon mode and a very large localized enhancement at their sharp tips.[113,114] Mie theory has been extended to cylindrical and needle-like nanoparticles,[96] and using discrete dipole approximations optical properties (absorption, extinction, and scattering efficiencies) have been calculated for metallic nanoparticles having arbitrary shapes.[106,115,116] Examination of the scattering spectrum of differently shaped (spherical, rod-like, triangular, pentagons, tetrahedrons) individual nanoparticles (of silver, gold, and nickel), by total internal reflection (pseudo-dark-field) spectroscopy (Fig. 6), has obviated the need for producing highly monodispersed populations of nanoparticles in desired shapes, and provided meaningful correlations between shape, environment, and spectra.[117] Mie theory is only applicable to non-interacting nanoparticles that are well separated in their solid state or are present at low concentrations in dispersions. For interacting particles the plasmon resonance red-shifts with the concomitant appearance of a lower energy absorption band (similar to the longitudinal absorption band observed for nanorods). Two types of interactions prevail between arrayed metallic nanoparticles: near-field coupling (between particles that nearly touch each other) and far-field dipolar interactions.[118±120] Dipolar interactions are electrodynamic in nature; the dipole fields, resulting from a plasmon oscillation of a single particle, induce surface plasmon oscillation in the adjacent particle(s).[121] The effect of changing the interparticle distance between pairs of electron-beam deposited gold nanoparticles (150 nm diameter, 15 nm height, see Fig. 3) on their s- and p-polarized extinction spectra is shown in Figure 7.[76] The observed spectra were rationalized in terms of a simple di-

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Figure 5. a) TEM image of gold nanorods synthesized electrochemically in micellar solutions. b) Size distribution of the nanorods. c) Absorption spectrum of the nanorods. d) Simulated absorption spectra of the nanorods with different aspect ratios. Reproduced with permission from [235]; copyright 2001, the American Chemical Society.

pole±dipole interaction model.[76] Significantly, these far-field dipolar interactions permit the use of aligned metallic nanoparticles as waveguides (see Sec. 4.2).

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

4. Nanoscale Optical and Photonic Devices

Figure 6. Schematics of the total internal reflection (pseudo-dark-field) spectroscope, TIRS. The angle of incident light (tungsten halogen lamp and a high throughput monochromator and polarizer) is chosen for total internal reflection (i.e., all the light is reflected). The nanorods placed in the evanescent field above the substrate surface scatter the light out of this field. Only the scattered light is collected by a conventional microscope objective and focused onto a spectrophotometer or onto a CCD camera (for microscopic images, not shown).

Near-field coupling of localized surface plasmon resonance is increasingly being exploited for the construction of nanoscale optical and photonic devices.[122±124] This highly exciting and potentially rewarding development is based upon the substantial enhancement of the electric field around and between the metallic nanoparticles, subwavelength holes, corrugations, and textures in metallic films. Fabricating these nanostructures with an unprecedented (nanometer-level precision) control over their sizes, shapes, and spacings is an essential requirement for the construction of nanoscale optical and photonic devices. In essence, advantage is taken in nano-optical devices of converting photons to electromagnetic modes of radiation (i.e., to localized surface plasmons, LSPs) that are not diffraction limited (i.e., they can be focused more tightly than k/2) and can be employed as optical elements (mirrors, lenses, switches, or waveguides, for example) within their propagation lengths (up to 150 lm at 700 nm for Ag[125]). Furthermore, decoupling of surface plasmons to light has been recently demonstrated.[126] Propagation of light is, of course, not limited by distance. It is fair to state in the introduction to this section that while extensive numerical simulations have been performed in these systems, theoretical understanding of the complex relationship between the nature of surface plasmon modes and the surface states of metallic nanostructures is only in its infancy. Here we shall survey the propagation of surface plasmons, the extraordinary optical light transmission, superlenses, and optical data storage (in Sec. 4.1); plasmon waveguides and nano-optics (in Sec. 4.2); the different near-field scanning optical microscopies; the mapping of localized and propagating surface plasmons (in Sec. 4.3); and nanophotonics (in Sec. 4.4). Table 1 summarizes the most pertinent publications.

4.1. Surface Plasmon Propagation, Extraordinary Optical Transmission of Light, Superlenses, and Optical Data Storage Conversion of photons to surface plasmons allows the subwavelength manipulation of electromagnetic radiation. Optimally this conversion can reach 100 % efficiency. However, the intensity of surface plasmons, propagating along the metallic surface, gradually decays over the 10±200 micrometer range. This propagation length (the distance after which the intensity of the surface plasmons decreases to <E1-e>, LSP,) can be theoretically modeled and is given by:[127] LSP ˆ Figure 7. Extinction spectra of 150 nm diameter gold nanoparticles deposited by EB-lithography to heights of 17 nm at 450 nm (a), 300 nm (b) and 150 nm (c) center-to-center distance. The orthogonal particle separation is kept constant, as can be seen in Figure 3. The polarization direction of the excitation light is parallel (top) and orthogonal (bottom) to the long axis of the particle pairs. Reproduced with permission from [76]; copyright 2003, Elsevier.

Adv. Mater. 2004, 16, No. 19, October 4

1 2kSP

c e0m ‡ed ˆ x e0m ed

!3=2 

e0m

2

em

(2)

where kSP² is the imaginary part of the complex surface plasmon wavevector, em¢, and em²are the real and imaginary parts of the dielectric function of the metal, c is the speed of the light and x is the optical excitation frequency. Equation 2 predicts that longer propagation lengths can be realized with

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1691

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Table 1. Nanoscale optical and photonic devices based on localized surface plasmon resonance. Localized Surface Plasmon Resonance Material

Principle

Ligand coated silver nanoparticles arrayed on water (metal-liquid-like films, MELLFS)

High reflectivity versatile and inexpensive surfaces are formed which can be readily manipulated by magnetic and electromagnetic fields 20 nm diameter, 1 ± 15 lm long Au, Ag, and Au/Ag nanowires Metal- and wavelength dependent unidirectional (prepared by template directed electrosynthesis) plasmon propagation over distances > 10 lm are exposed through prism total internal reflection illumination demonstrated Periodic arrays of 600 ± 1800 nm diameter holes in 200 nm Coupled LSPR enhanced optical transmission silver film on quartz substrate A single 40 nm-diameter 10 lm long slit, surrounded by Coupled LSPR enhanced optical transmission by: a finite array of grooves, made on a 350 nm-thick silver film (i) groove cavity mode excitation (controlled by the depths of the grooves); (ii) in-phase groove re-emission (controlled by the period in the groove array); (iii) slit waveguide mode (controlled by the thickness of the film) A single subwavelength aperture surrounded by an equally Lensing controlled by the output grooves (transmission spaced finite array of grooves, made on a metallic film controlled by the input grooves, see above) Periodic arrays of 150 nm diameter holes Resonant interaction of the coupled in 200 ± 500 nm thick silver films LSPR with the incident radiation Periodic arrays of 250 ± 400 nm diameter holes Resonant interaction of the coupled in 200 ± 300 nm thick silver and nickel films LSPR with the incident radiation Nickel microarrays of subwavelength apertures Resonant interaction of the coupled LSPR with the incident radiation 1-dodecanethiol monolayer self-assembled onto Resonant interaction of the coupled nickel microarrays of subwavelength apertures LSPR with the incident radiation

A bigger (41”41 holes) and a smaller (11”11 holes) periodic arrays of 250 nm diameter holes in 150 nm thick gold films placed 30 lm apart Electron beam deposited 35 nm thick, 30 nm diameter, and 150 nm long elongated silver nanoparticles are organized into different configurations (i) 50 nm diameter silver or gold nanoparticles are patterned (with 75 nm center-to-center distance) on ITO substrates by EB lithography (ii) 30 nm diameter gold nanoparticles are manipulated by the AFM-tip to desired pattern on ITO substrates EB-lithographic generation of eighty 50 nm diameter gold nanoparticles with 75, 100, and 125 nm center-to-center spacing 40 nm thick silver film (deposited onto a glass substrate) textured with a hexagonal array of 100 nm diameter dots (composed of a photoresist) with a 300 nm periodicity A two-dimensional array of repeated unit cells of copper strips and split ring resonators on interlocking strips of standard circuit-board material 50 nm thick silver film, deposited onto the flat side of a semi-spherical glass prism 40 nm high, 100 nm diameter oblate spheroidal gold arrays, with 650®20 nm periods, prepared by EB-lithography on a glass prism (substrate) A bull's eye structure surrounding a cylindrical hole in a suspended silver film (groove periodicity, 500 nm; groove depth, 60 nm; hole diameter, 250 nm; film thickness, 300 nm) A vertical-cavity surface-emitting laser array, VCSEL, is fabricated. It consists of 12 pairs of p-type distributed Bragg reflectors, DBRs, and 34.5 pairs of n-type DBRs with a 2.8 lm oxide aperture which is covered by a layer (Y2N thick) of SiO2 and a 100 nm thick gold film (with a single aperture of 200 nm) diameter 40 nm-high and 480 nm-wide silver wire gratings, with a 577 nm period fabricated on a glass slide by soft-lithography

1692

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Focusing a Ti-sapphire laser onto smaller array (the source) lit up the bigger array (the probe) Scattering spectra of the LSPR band for a given polarization direction have maxima at 680 nm and 480 nm for vertically and horizontally oriented particles, respectively Dipole fields, resulting from LSPS of a single nanoparticle, induce plasmon oscillations in neighboring particles leading to coherent modes with a wavevector k along the nanoparticle arrays Optical properties, determined by far-field polarization spectroscopy, indicate the presence of longitudinal and transverse plasmon modes Reflectivity from TM polarized (scanned between 400 nm and 800 nm), coupled to the surface plasmon was collected at the full range of propagation directions The negative refractivity material (above 20 Ghz) enhanced transmissivity of the evanescent field Negative permittivity broadened the surface plasmon band and enhances the transmissivity of the evanescent field Appropriate coupling of the localized surface plasmon fields gave rise to subwavelength spots between the gold structures (measured by an apertureless near field scanning optical microscope) Light is transmitted with low divergence, directionality and high efficiency

(Potential) Function, Application High quality parabolic mirrors [219]

Nanoscale optical coupling, plasmon waveguides [130] Extraordinary optical transmission through subwavelength holes [131] 40-fold light transmission enhancement has been experimentally observed [138]

Theoretical analysis demonstrates the display of a lensing effect [139] Wavelength selected transmission with 1000 times increased efficiencies [132] Transmission enhancement depend on the dielectric properties of the metal [220] Microarray transmit about 3 times the intensity of light incident upon the holes [221] Intensity of the absorption of the methyl CH2 band (in the 2800 ± 3000 cm-1 region) is enhanced by 100-fold with respect to traditional IRRAS [221] Surface plasmons launched at the source array appears as a streak between the two arrays and then decouple to light[126] Optical data storage[145]

Waveguides below the diffraction limit of light [122]

Optical information transfer below the diffraction limit [119] Light propagation is prohibited in all directions between 1.9 eV and 2.0 eV [166] Construction of diffraction-free collecting and focusing superlenses [143,229] Construction of diffraction-free collecting and focusing superlenses [144] Subwavelength patterning [230]

Miniature phase-array antennas in the optical regime which can transmit or receive light for a given wavelength [231]

Optical near field is strongly enhanced by the excitation High density optical data storage[232] of localized surface plasmon around the metal aperture. The far-field power radiated from the apertures is enhanced by sixteen-fold (with respect to a single aperture VCSEL). Peak power density = 2.5 mW/lm2 with a 260 nm spot size. Dispersion of surface plasmons show a clear energy gap when the two plasmons (at the air/silver and glass/silver interface) interfere

http://www.advmat.de

Metallo-dielectric gratings with subwavelength gaps in the thin metal limit [233]

Adv. Mater. 2004, 16, No. 19, October 4

silver than with gold, and that the propagation length increases with increasing wavelength. Indeed propagation lengths of up to 150 lm were observed for silver at 700 nm, while gold under the same conditions only propagated surface plasmons for up to 30 lm.[123,128] Experimental probing of the surface plasmon propagation length involved optical and photonic scanning microscopies in the near field; and darkfield or total internal reflection microscopy and/or spectroscopy in the far field.[122,125,128,129] For example, gold and silver waveguides (typically 30±60 nm thick, 1±2.5 lm wide metallic stripes deposited onto an optically transparent glass) were excited by light generated by total internal reflection via a prism, and the scattered light was visualized by photon scanning microscopy.[125] Plasmon propagation has also been examined in electrochemically generated 1±15 lm long, 20 nm diameter gold, silver, and bimetallic (with distinct gold and silver segments, i.e., gold and silver heterojunctions) nanorods.[128] Arrangements were made in this work to selectively launch the surface plasmons only at one end of the nanorods. Scattered light from gold nanorods was observed upon 820 nm irradiation, but not on using a 532 nm light as the excitation source. Conversely, silver nanorods propagated both the 532 nm and the 820 nm irradiation (Fig. 8).[128] Bimetallic nanorods were found to be directionally selective. Surface plasmons were found to propagate the 532 nm light from the silver end of the rod, whereas the 820 nm light could be propagated from both the silver and the gold ends of the bimetallic nanorods. These observations pave the way to nano-optical devices in which optically encoded information is transmitted with nanoscale accuracy over distances of 10± 15 lm.[128] Surface plasmon propagation has been investigated in several additional systems.[130] Significantly, surface plasmons were reconverted to light by two arrays of 250 nm diameter holes in a 150 nm thick gold film, placed 30 lm apart.[126] One array contained 41 ” 41 holes and the other array 11 ” 11 holes. Surface plasmons launched by a Ti:sapphire laser at the smaller array (the source) propagated to the sink, the larger array, where they reconverted to light (Fig. 9).[126] This possibility of reversibly converting light to surface plasmons opens the door to the extension of nano-optical devices to the millimeter scale. In 1998, Ebbesen and co-workers reported an extraordinary light transmission through 150 nm diameter holes (with a regular periodicity between 600 and 1800 nm) formed by a focused ion beam through 200 nm thick silver films (on quartz substrates).[131] The extraordinary light transmission manifested itself in the appearance of much more light (with only a small angular divergence!) than expected from the diffraction theory of isolated holes.[131] The transmitted light was characterized by a narrow silver surface plasmon peak (at 326 nm) and a set of well-defined minima and maxima related to the geometry of the hole array (Fig. 10).[131,132] This behavior is in contrast with the classical theory which predicts transmission efficiencies in the order of 10±3 through a 150 nm diameter hole and a monotonic decrease (instead of the observed increase) of the light transmission with increasing wavelength.

Adv. Mater. 2004, 16, No. 19, October 4

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Figure 8. a) Photograph of 20 nm diameter Au and Ag rods. The two labeled rods are typical of the ones used for these experiments. b) Optical microscope image of a 4.7 lm long Au rod exposed to through-prism TIR illumination at 532 nm. Note the strong scattering at the rod input and the absence of scattering at distal tip. c) Image of the same Au rod under TIR illumination at 820 nm. Under these conditions, both ends of the rod exhibit strong emission/scattering. d,e) A 4.7 lm long Ag rod illuminated at 532 and 820 nm, respectively. Emission is observed from both tips at both wavelengths in the case of Ag. Note: In panels b±e), the incident laser is focused to include a distribution of wave vectors, but the primary direction of propagation points vertically on the page. The scale bar in panel (b) is equivalent to 1 lm and is valid for images b±e). Dotted outlines of the rods are meant only as visual guides to the reader. Reproduced with permission from [128]; copyright 1998, the American Chemical Society.

Clearly, the array is actively involved in light transmission, most likely via resonant interactions with surface plasmons, localized in and around the holes in the metallic film. Extraordinary light transmission has been modeled by several approaches.[133±137] Appropriately, it was soon realized that the model employed should represent the experimental structure (arrayed nanoholes, gratings on one or both sides of the metallic films, with or without slits, etc.). For example, modeling a narrow-grooved zeroth-order (i.e., the pitches are smaller than k/2) silver grating which had the same periods on both sides indicated the resonance tunneling of p-polarized photons through the metal film via surface plasmons localized in the grooves of the opposite surfaces.[135] Conversely, theory predicted the complete reflection of s-polarized photons by

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1693

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

efficient transmission of the p-polarized light through the film and the complete reflection of the s-polarized light. Interestingly, the different plasmon resonance modes were calculated to have different features. Thus the propagating surface plasmons had a narrow resonance peak corresponding to a large phase lag and long time delay. In contrast, the localized surface plasmon modes had a broad resonance peak and their electric field highly concentrated in the grooves at both sides of the film.[137] Experimental verification of these models are awaiting the interested scientists. Substantial transmission enhancements have also been observed on using a single subwavelength aperture surrounded by an equally spaced finite array of grooves.[138,139] Images of metallic nanostructures, taken in the near field by photon scanning tunneling microscopy (see Fig. 11), contain all the optical information about the scatterer. Moving to the far field, however, the evanescent field decays exponentially, and information on subwavelength features are lost. Pendry Figure 9. Top: SEM image of coupled arrays of subwavelength holes. The arrays have a period of 760 nm and are separated by 30 lm. The hole diameter is 250 nm. The small array on the right (11 ” 11 holes) is the ªsourceº array whereas the big array on the left (41 ” 41 holes) is not illuminated by the laser light and acts as a ªprobeº array. Inset: zoom in on the holes of one array. Bottom: Transmission spectrum recorded in the far field when the sample is illuminated by a collimated source of white light. Reproduced with permission from [126]; copyright 1998, the American Institute of Physics.

Figure 11. Schematic of the optics used in photon scanning tunneling microscopy of gold and silver nanoparticles [157]. The incident light is coupled to the surface plasmon modes of the gold nanoparticles in the total internal reflection mode (Kretschmann configuration). Near-field images are probed by a sensitive photomultiplier (not shown) placed at the end of a sharply elongated uncoated optical fiber (the tip) which is used to scan the sample. Using the same fiber to deliver laser pulses through the optical fiber onto the sample (a thin metal film, rather than metallic nanoparticles) provides a means for lithographically introducing surface defects which can, of course be imaged in the scanning tunneling microscopic mode (i.e., replacing the photomultiplier at the end of the optical fiber) [155]. The light reflected from the prism is monitored by a photodiode (PD).

Figure 10. Zero-order transmission spectrum of an Ag array (ao = 0.9 lm, d = 150 nm, f = 200 nm). Reproduced with permission from[131]; copyright 1998, Nature Publishing Group.

these gratings. This polarization selectivity could lead to the construction of nanoscale polarizers. Two different mechanisms were proposed for the extraordinary light transmission through thin metal films with narrow grooves of different periodicities on both sides.[137] Both propagating and localized surface plasmon modes were found to be responsible for the 1694

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

recently proposed an ingenious way to overcome this problem by passing the light through a material with a negative refractive index, which enhances the evanescent field and hence permits subwavelength resolution.[140] Negative refractive index materials are also referred to as ªleft-handed materialsº, LHMs, (since the electric and magnetic vectors lie along the direction of ±k for propagating plane waves and thus form a left-handed system, as opposed to the right-handed system which prevails for ordinary materials) or more generally as metamaterials (implying properties and functionalities which are unattainable in naturally occurring materials, but which can be tailor made). Numerical simulations have indicated

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4

that judiciously arrayed metallic (silver, gold, or copper) nanowires and their composites can be used as LHMs in the near-infrared and visible part of the spectrum.[141,142] Experimentally, negative index of refraction has only been demonstrated in the microwave region to date.[143] A two-dimensional (2D) array of repeated unit cells of copper strips and split ring resonators on interlocking strips of standard circuit board material was shown to have negative index of refraction above 10.2 GHz.[141] Enhanced transmissivity of the evanescent field was demonstrated by taking advantage of the natural roughness of a 50 nm thick silver film deposited onto the flat surface of a hemispherical glass prism in a reversed attenuated total reflection system (i.e., the collimated Ar-ion laser beam is incident normal to the center of the silver-coated flat side of the prism, and the angle distribution of the relative light intensity is determined by a charge-coupled detector, CCD, camera placed 5 cm from the center of the hemispheric side of the prism).[144] The negative permittivity of the silver film (approaching ±1) was considered to be responsible for the observed broadened surface-plasmon bandwidth and enhanced transmissivity of evanescent waves.[144] These results open the door to the construction of diffraction-free image collecting and focusing of subwavelength superlenses. Optical data storage has been demonstrated by the spectral coding of differently organized 35 nm thick, 30 nm diameter, and 150 nm long silver nanorods (Fig. 12).[145] Three different arrangements of the nanorods, A, B, and C, constitute the building blocks of the data storage system. In A, a single silver nanorod was aligned with its long axis parallel to the propagation of the incident light. B consisted of two silver nanoparti-

cles, both aligned with their long axis perpendicular to the propagation of the incident light. In C, in addition to the two silver nanoparticles which were aligned as those in B, a third silver nanoparticle was placed parallel to the propagation of the incident light close to one end of the other two particles. B and A had a single absorption maximum at 680 nm and 480 nm, respectively, while the absorption spectra of C was characterized by two maxima. Organizing a large number of A, B, and C structures in a given area, 500 nm apart, provided a data storage system in which the information could be read by irradiation with blue (488 nm) and red (633 nm) lasers.[145,146] Using this approach the information density is limited by a) the area observable by the objective of the total internal reflection spectroscope, TIRS, b) the number of nanorods which manifest different scattering spectra in that observable area, and c) the number of distinctly resolvable resonance lines originating in the scattering spectra in the observable area. The far-field resolution of a microscope is roughly a 500 nm diameter circular area. This is, in fact, the area typically occupied by a single pit in a conventional compact disk. Smaller areas can be observed by photon scanning tunneling microscopy (by using a sharply elongated optical fiber, instead of the scanning tunneling microscope tip, and placing a sensitive photomultiplier at the end of this fiber to observe the surface plasmons generated on the platform of the TIRSÐFig. 6). Limiting the spectral range to the visible range a sixfold enhancement of information content of a compact disk pit is achievable.[145]

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

4.2. Plasmon Waveguides and Nano-optics

Figure 12. Top: SEM images of the nano-optical letters. The double arrow indicates the polarization direction of the incident light. Bottom: Scattering spectra corresponding to the nano-optical letters. Reproduced with permission from [145]; copyright 2000, Optical Society of America.

Adv. Mater. 2004, 16, No. 19, October 4

Construction of optical devices at the nanoscale requires the efficient guiding of light well below the diffraction limit, even around 90 bends, without radiation losses. Traditional metallic pipe waveguides only transmit microwaves without losses and the dielectric guides, operating in the infrared and visible range of the spectra, are restricted to moderate curvature bends by radiation losses. In contrast, aligned gold and silver nanoparticles, by virtue of their coupled surface plasmon resonance modes, have been shown to propagate electromagnetic energy well below the diffraction limit in a coherent fashion, with negligible radiation losses, at velocities approximating 10 % of the speed of the light.[122] Furthermore, transmission losses for nanoparticle chain arrays, 90 corners, and T-structures were negligible.[122] Far-field polarization spectroscopy of 50 nm gold nanoparticles, aligned with 75 nm, 100 nm, and 150 nm center-to-center spacing, indicated the presence of longitudinal and transverse plasmon modes.[119] Polarization dependence of the extinction spectrum of aligned metallic particles is in accord with the point dipole model of these structures and permits the exploration of the bandwidth dispersion relationship, which in turn, will aid the design of improved plasmon waveguides. Plasmon waveguides and metallic nanostructures can therefore lead to the construction of integrated nano-optical de-

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1695

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

vices incorporating reflectors, mirrors, switches, and other components.[124]

beam

splitters,

4.3. Apertureless Near-Field Scanning Optical Microscopy, Scanning Plasmon Near-Field Microscopy, Photon Scanning Tunneling Microscopy, and Mapping Localized, Coupled, and Propagating Surface Plasmons Propagating and localized surface plasmons have significantly aided the development of high-resolution near-field scanning optical microscopies (NSOMs). These microscopies, in turn, have provided an experimental means for imaging and mapping localized, coupled and propagating surface plasmons. Near-field scanning optical microscopy is not diffraction limited; optical resolution is primarily determined by the interior diameter of the tip of a sharply elongated metal-coated optical fiber through which illumination is introduced. The tip-to-sample distance is a few nanometers (i.e., it is a near field optical measurement).[147,148] The tip (or the sample) is raster-scanned by a piezoelectric drive, typically employing a commercially available scanning force microscope, to obtain nanoscale optical images (either in the transmission or reflection mode). In addition, topographic images can be recorded using the tip as an atomic force microscopy (AFM) probe. Problems of preparing uniformly pulled and metal-coated fibers with a nanoscale interior diameters at their tips, through which laser light of appropriate energy (just sufficient to illuminate the sample, but not too much to photo-damage it) can be introduced, and intensity decays in the tunneling region impose fundamental sensitivity limits to NSOM. These problems have been overcome by the development of apertureless NSOM, in which a sharp metallic probe is illuminated and the scattered light is detected at far field. Resolution of a few nanometers with apertureless NSOM is a consequence of the enhanced optical field at and near the tip±sample region.[149±151] An alternative approach is to take advantage of propagating surface plasmons, generated from a thin gold or silver film, deposited on a prism.[152] The incident light is coupled to the surface plasmon modes of the metallic film in the total internal reflection mode by a prism (Kretschmann configuration) and the reflected light is detected by a photodiode. As the sharp scanning tunneling microscopy (STM) tip scans the substrate close to the surface it slightly increases intensity of the reflected light (by reducing the destructive interference) and thus produces topography-dependent signals. This approach, termed scanning plasmon near-field microscopy, permits up to 3 nm resolution.[153] Using a sharply elongated uncoated optical fiber, instead of the STM tip to scan the sample (at a constant tip-to-surface distance), and placing a sensitive photomultiplier at the end of this fiber allow the near-field optical imaging of the sample. This configuration is referred to as photon scanning tunneling microscopy.[154,155] Photon scanning tunneling microscopy directly measures the propagating sur-

1696

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

face plasmons near the surface. Photon scanning tunneling microscopy has also been employed to create surface defects lithographically (by focusing an excimer laser through the uncoated optical fiber onto the silver film, evaporated onto the hypotenuse face of a right angle prism).[156] The same fiber was then used to obtain near-field images of the defects created.[156] Additionally, the probe±sample interactions can be observed at far field (for example an angle of 75 from the probe axis in the backscattering direction with respect to the incident wave vector[157]) by a photomultiplier via a microscope objective, band-pass filter, polarizer, and lens (see Fig. 11). More recently, silver and gold nanoparticles have been investigated by photon scanning tunneling microscopy (Fig. 11).[157] The normal component of the scattered electric field was examined as functions of incident light polarization and tip-tosample distance. Resonant excitation (at 415.4 nm), using transverse magnetic (TM) illumination, indicated spatial confinement and local enhancement of the electric field around the silver nanoparticles. Under the same conditions, off-resonance excitation of gold nanoparticles indicated a substantial contribution of the propagating components of the scattered field. Transverse electric (TE) illumination resulted in a much larger scattering cross-section for silver than for gold nanoparticles.[157] These results significantly impact our understanding of surface plasmon propagation. Single photon tunneling (analogous to single electron tunneling) has also been demonstrated recently by measuring the intensity of the light traveling through individual subwavelength pinholes in a gold film (deposited on a glass prism in an optical arrangement similar to that shown in Fig. 11) covered by a layer of polydiacetylene.[158]

4.4. Nanophotonics Photonic bandgap materials, wavelength scale insulators, and semiconductors which are periodically structured affect the properties of photons in much the same way as a semiconductor affects the properties of an electron.[159±162] That is to say, photons can have band structures, bandgaps (i.e., frequencies between which light propagation is prohibited), localized defect states, and surface states. Consequently, photonic bandgap materials can mold and guide light. Photonic bandgap materials can be prepared by chemical approaches. Synthesis of barium titanate inverted opals illustrates one of the most common methods. The procedure involves: i) the introduction of barium titanate into the interstices of the polystyrene opal template; ii) polycondensation; and iii) removal of the polystyrene opal by solvent extraction or calcination.[163] The three-dimensional network of ferroelectric barium titanate surrounding the spherical air packets (whose dimensions can be controlled by selecting the size of polystyrene particles which form the three dimensional opal template) were demonstrated to function as a photonic crystal.[163] Alternatively, three-dimensional semiconductor photonic crystals can be di-

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4

rectly self-assembled.[164] Chemical approaches provide a viable economic alternative to EB lithography. Photonic bandgap material has also been prepared by surface (instead of bulk) modulation of thin metallic films.[124,165± 168] Characterization of a full photonic bandgap, between 1.91 and 2.00 eV, was first reported on a 40 nm thick silver film (deposited onto a glass substrate) which was textured with a hexagonal array of 100 nm diameter dots (composed of a photoresist) with a 300 nm periodicity (Fig. 13).[166] The incident TM polarized light (scanned between 400 nm and 800 nm) was coupled to the surface plasmon modes of the sil-

5. Sensors and Biosensors

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Localized surface plasmon resonance has been exploited in several ways for sensing applications (see Tables 2±5 for a summary).[169] First, the sensitivity of surface plasmon band to its immediate environment offers, in itself, an opportunity to detect attached molecules and environmental changes (assays based on changes of the localized surface plasmon resonance absorbance, Sec. 5.1). Second, the reversible aggregation of plasmon resonant particles through specific linkers (oligonucleotides, for example) provides an excellent means for colorimetric assays (assays based on the aggregation of metallic nanoparticles, Sec. 5.2). Third, the ultrabright light scattering from each plasmon resonant particle makes the optical detection of a single molecular target possible (labeling by metallic nanoparticles manifesting localized surface plasmon resonance, Sec. 5.3). Fourth, localized surface plasmon resonance amplifies the signal of several optical techniques to give rise to surface-enhanced Raman scattering, SERS, surface-enhanced resonance Raman scattering, SERRS, surface-enhanced infrared reflection absorption spectroscopy, SEIRRAS, localized surface plasmon (LSP)-enhanced surface plasmon resonance spectroscopy (SPR), and enhanced fluorescence spectroscopy. Since there are many recent reviews on SERS and SERRS,[170,171] as well as on SEIRRAS,[172] and surface-plasmon enhanced fluorescence spectroscopy[96,173±175] we shall focus our discussion only to the LSP-enhanced SPR (in Sec. 5.4).

5.1. Assays Based on Changes of the Localized Surface Plasmon Resonance Absorbance

Figure 13. a) A scanning electron micrograph of the hexagonal array of dots. The dots are composed of photoresist on a glass substrate. The surface has been coated with a thin film of silver to support the propagation of SPPs. b) The energies of the upper and lower branches of the SPP energy gap plotted as a function of the propagation direction w. The angles are accurate to ±0.2and the energies to ±0.01 eV. There is a full gap between 1.91 and 2.00 eV. Reproduced with permission from [166]; copyright 1996, the American Physical Society.

ver film in the total internal reflection mode by a prism (in the Kretschmann arrangement, similar to that shown in Fig. 12) and the reflectivity data were collected as a function of photon energy and photon wave vector in the plane of the silver±air interface at the full range of propagation direction. Dispersion curves for each propagation direction exhibited a clear energy gap between 1.91 eV and 2.0 eV (Fig. 13), indicating that there is no propagation of energy in this range in any direction at the silver/air interface.[166] Emission from a laser dye 4-dicyanomethylene-2-methyl-6-(p-(dimethylamino)styryl-4H-pyran (DMC), spin-coated on a textured silver film, was also found to be inhibited in the bandgap region.[165]

Adv. Mater. 2004, 16, No. 19, October 4

The fact that the color of metallic nanoparticles and nanoislands depends markedly on the refractive index of the surrounding medium has been exploited for sensing applications. Significant amounts of work have been published on the sensing, detection, and quantification of molecules attached to metallic nanoparticles and nanostructures by the shift they inflict upon the surface plasmon resonance absorption band. These type of assays were conducted in dispersions,[62,176±179] on metallic nanoparticles immobilized on substrates,[180±182] on metallic nanoislands,[65±68,183,184] and periodic particle arrays (PPAs) deposited onto substrates.[70,72,73,185,186] (see Table 2). Functionalization of colloidal gold nanoparticles by monoclonal antibodies, specific for single or multiple epitopes (sites on antigens which are recognized by the antibody) and monitoring the shift of their SPR absorbance peak upon interacting with their ligand illustrate the approach in dispersions.[176,177] The shift of the surface plasmon absorption band was found to be proportional to the concentration of the ligand and related to the kinetics of the interaction.[177] Several similar approaches in dispersions have been reported (see Table 2). Most assays have been carried out, however, on surfaceimmobilized sensing elements. Confinement of the metallic nanoislands or PPAs onto substrates has manifold advantages.

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1697

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Table 2. Assays based on changes of the localized surface plasmon resonance absorbance. Sensing Element

Principle

Analyte Detected

Gold nanoparticles, functionalized by monoclonal antibody, in dispersions Mannose-functionalized gold nanoparticles in dispersions Amide-functionalized gold nanoparticles in dispersions Gold nanoshell dispersions

Absorption maximum of the LSPR in the sensing element is redshifted by the molecularly recognized ligand The LSPR spectrum of gold nanoparticles changes when the mannose interacts with lectin The LSPR spectrum of gold nanoparticles changes in the presence of anions The LSPR wavelength of gold nanoshells changes when functionalized by alkanethiols The LSPR spectrum of immobilized gold and silver nanoparticles changes when the anti-HSA interacts with HSA

Ligand recognized by the monoclonal antibody [176,177] Lectin [179]

Gold and silver nanoparticles attached to quartz substrates and functionalized by anti-HSA (Human Serum Albumin) Surface attached single gold nanoparticle functionalized by biotin Gold nanoislands on a fiber (fiber optic sensor) Gold nanoislands, prepared by vacuum evaporation onto transparent substrates (T-LSPR) Gold nanoislands, vacuum evaporated onto transparent substrates, modified by cystamine, and then reacted with biotin Silver PPAs (T-LSPR) Biotin functionalized silver PPAs (T-LSPR) Cap-shaped gold nanostructures evaporated onto a dense monolayer of polystyrene spheres Thin film of dodecylamine-capped Au nanocrystals

The LSPR spectrum of single gold nanoparticle shifts upon the interaction between biotin and streptavidin molecules The LSPR spectrum of gold nanoislands shifts upon immersion of the fiber into a medium other that air The LSPR spectrum of gold nanoislands shifts upon attaching the analyte to them

The LSPR spectrum of gold nanoislands shifts upon the interaction between biotin and avidin molecules

The LSPR spectrum of silver PPAs shifts upon attaching the analyte to them The LSPR spectrum of silver PPAs shifts upon the interaction between biotin and anti-biotin molecules The LSPR spectrum of the sensing element shifts upon the binding of analyte molecules The LSPR spectrum of the Au nanocrystals shifts upon the binding of analyte molecules

First, their interparticle distances and positions can be controlled, since they are immobilized on the substrate. Second, they need not contain any capping agents or stabilizers and can, therefore, be readily functionalized or indeed used for sensitive measurements in their pristine state. Third, adsorbed substances can be desorbed and bound molecules can be released with the concomitant recovery of the original localized surface plasmon absorption band; thus sensing can be repeatedly and continuously performed. Development of transmission localized surface plasmon resonance, T-LSPR, spectroscopy merits a special emphasis. Two different approaches for the exploitation of T-LSPR spectroscopy have been reported to date. The first method, developed by the Rubinstein group,[65±67,187] is based on the vacuum deposition of 2±10 nm thick gold nanoislands onto freshly cleaved mica sheets or quartz slides. Depending on individual preparations, the T-LSPR maximum of the gold surface plasmon absorption band was observed between 570 and 630 nm. Self-assembly of cyclic disulfide monolayers onto the gold nanoislands could be monitored quite readily by the position and intensity change of the T-LSPR band (ca. 0.02 absorbance units, Fig. 14).[65] The validity of the technique has been demonstrated by the observed parallel linear increases 1698

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Anions [178] Alkanethiols [62] Human Serum Albumin (HSA) [180]

Streptavidin [181] Any change in the surrounding medium [183,184] Metalloporphyrin and metallophtalocyanine monolayers [67], 4-aminothiophenol, 4-mercaptopyridine, cyclic disulfides, long-chain thiols [65]. Oligonucleotides and their hybridization [68] Avidin molecules [188]

Alkanethiols [70,72], Streptavidin [185] Anti-biotin [73] Octadecanethiol, biotin, avidin [186]

Propanethiol [182]

of i) the advancing contact angle of water deposited onto the self-assembled monolayer (SAM: formed at different times of immersion of the gold nanoislands into a CHCl3 solution of the cyclic disulfide), and ii) the absorbance due to the cyclic disulfide.[65±67] The absorption of other molecules (metalloporphyrin and metallophthalocyanine monolayers,[67] 4-aminothiophenol, 4-mercaptopyridine, aminoethanethiol, longchain thiols,[65] and a submonolayer of thiolated oligonucleotides[62]) onto gold nanoislands was also readily detected by T-LSPR (Table 2). Advantage has been taken in the second approach of the sensitivity of the localized surface plasmon absorption band of 50 nm high and ca. 100 nm wide triangular silver nanoparticles (PPAs).[70,185±188] Exposed to different solvents the PPAs underwent structural changes which manifested themselves in dramatic shifts of the T-LSPR absorption maximum. Self-assembly of hexadecanethiol monolayer shifted the absorption maximum of the triangular silver nanoparticles from 564 nm to 604 nm; this corresponded to a red-shift of 3 nm for every carbon atom in the alkanethiol chain caused by only 60 000 molecules per nanoparticle.[66] SAM-functionalized triangular silver nanoparticles have been shown to function as a biosensor. Electrostatic binding of a cationic polypeptide to mercap-

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Table 3. Assays based on the aggregation of metallic nanoparticles. Sensing Element

Principle

Analyte Detected

Oligonucleotide-functionalized gold nanoparticles in aqueous dispersions Surface immobilized, oligonucleotide-functionalized gold nanoparticles

Hybridization of oligonucleotides links the gold nanoparticles and their proximity changes the LSPR spectrum Hybridization of oligonucleotides links the gold nanoparticles and their proximity changes the LSPR spectrum

Oligonucleotide-functionalized gold and latex nanoparticles in aqueous dispersions

Hybridization of oligonucleotides attach the gold nanoparticles to the latex particles and their complex appears red

Oligonucleotide-functionalized gold nanoparticles and oligonucleotide-functionalized streptavidin in aqueous dispersions A protein recognition element attached to an oligonucleotide is hybridized to gold nanoparticles, that are functionalized by the complementary oligonucleotide

Hybridization of oligonucleotides links the gold nanoparticles and streptavidin molecules to each other which changes the LSPR spectrum The antibodies (with two recognition sites) attach to the proteins and link the gold nanoparticles, changing thus their LSPR spectrum; Dehybridization properties are specific for the DNA sequence, allowing to identify a series of analyte (ªbio-barcodesº) Reversible aggregation of gold nanoparticles in the presence of analyte Aggregation of gold nanoparticles in the presence of analyte Reversible aggregation of gold nanoparticles in the presence of analyte (enhanced hyper-Rayleigh scattering!) The K+ ions link the gold nanoparticles together and the LSPR spectrum changes; The dinitrophenyl units link the pyrenyl conjugated gold nanoparticles together and the LSPR spectrum changes; temperature dependent reversible aggregation Binding of streptavidin to the disulfide biotin analogue induces the aggregation of gold nanocrystals and changes in the LSPR spectra The analyte links the Au and Ag nanoparticles which results in their aggregation and changes in the LSPR spectra

Oligonucleotides complementary to those in the sensing element [189±193] Oligonucleotides complementary to those in the sensing element [211,222,223] Oligonucleotides complementary to those in the sensing elements (latex and gold) [195] Oligonucleotides complementary to those in the sensing element [224]

Gold nanoparticles modified by a-lactosyl-x-mercaptopoly(ethylene glycol) Protein A coated gold nanoparticles, in aqueous dispersions Gold MPCs capped by 11-mercaptoundecanoic acid, in aqueous dispersions Gold nanoparticles modified by 15-crown-5, in aqueous dispersions Gold nanoparticles with pyrenyl units on the surface, in aqueous dispersions

Gold nanocrystals modified by disulfide biotin analogue, in aqueous dispersions A mixture of Au nanoparticles with surface attached IgE molecules (against dinitrophenyl (DNP)) and Ag nanoparticles with surface conjugated anti-biotin IgG antibodies Molecularly imprinted polymer with embedded gold nanoparticles

The polymer swells upon binding of the analyte, altering the proximity of the immobilized Au nanoparticles and causing to shift their LSPR spectrum

toundecanoic acid, self-assembled to triangular silver nanoparticles, caused the absorption maximum to shift from 744.6 nm to 749.8 nm and upon the release of the polypeptide the absorption maximum returned to 744.6 nm.[70]

5.2. Assays Based on the Aggregation of Metallic Nanoparticles The aggregation of metallic nanoparticles results in substantially altered surface plasmon absorption spectra, i.e., in a visible change of color. Based on this principle, a colorimetric DNA hybridization assay was developed, detecting the spectral change of thiolated oligonucleotide-covered gold nanoparticles when they aggregated in the presence of a complementary target oligonucleotide.[189±193] The spectral changes were found to be completely reversible by heating: the DNA strands dehybridized and the particles separated from each other. The melting transition of DNA-linked nanoparticle aggregates was extremely sharp (compared to the melting Adv. Mater. 2004, 16, No. 19, October 4

IgG1 and IgE [196]

Lectin [200] Anti-protein A [197] Heavy-metal ions(Pb2+, Hg2+, Cd2+) [201] K+ ion [202] Dinitrophenyl units [203]

Streptavidin [198]

Hetero-Janus DNP-biotin antigen [199]

Adrenalin [234]

transition of unattached, ªfreeº DNA), the melting temperature was relatively high, and very sensitive to base mismatches.[190] The detection limit of these assays was in the femtomolar range.[189] Silver nanoparticles have also been functionalized by thiolated oligonucleotides and the spectral changes upon their hybridization proved to be different from those of gold nanoparticles, exhibiting a marked decrease in intensity of the SPR peak, as opposed to a shift of its maximum.[194]This specific behavior is sufficiently different to that of the gold nanoparticles to warrant the use of silver nanoparticles as a second marker in DNA hybridization experiments.[194] An interesting and exceptionally simple variant of this colorimetric DNA hybridization assay has been based on the aggregation of oligonucleotide-functionalized large (3.1 lm) white latex and small (13 nm) red gold particles in the presence of a target DNA strand (Fig. 15). After the hybridization, the dispersion was filtered through a 0.45 lm membrane, which allowed the unhybridized small gold particles to pass, but not the large latex or the joined latex±gold complexes. If

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1699

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Table 4. Labeling by metallic nanoparticles manifesting localized surface plasmon resonance. Sensing Element

Principle [a]

Analyte Detected

DNA microarray

Molecular recognition between the sensing element and analyte and labeling of analyte by gold nanoparticles Molecular recognition between the sensing element and analyte and labeling of analyte by anti-biotin conjugated gold nanoparticles Attachment of biotinylated detection oligonucleotides to the template DNA with the subsequent labeling of them by anti-biotin conjugated silver nanoparticles; single particle counting! Toxin A was sandwiched between the sensing element and a biotinylated (detection) antibody against Toxin A; The detection antibody was labeled by anti-biotin conjugated gold nanoparticles The antibodies (and their gold nanoparticle label with them) attach to the lymphocytes surface

Oligonucleotides complementary to those in the sensing element[208,211,224] Biotinylated oligonucleotide, complementary to those in the sensing element[209,225] DNA polymorphism in template oligonucleotides[206]

DNA microarray DNA microarray

Substrate bound antibodies against Toxin A, in a commercial ELISA sandwich assay Gold nanoparticle conjugated antibodies against CD4, in aqueous dispersions Mouse antibody against HER-2/neu tyrosine kinase Human Papilloma Virus specific DNA probe, labeled with FITC (fluorescein isothiocyanate) Biotinylated DNA probe Mouse anti-ryanodine antibody Multisegmented metallic (Au, Ag, Ni) rods (barcodes) derivatized with oligonucleotides or antibodies for human and rabbit IgG

Breast carcinoma tissue was sequentially reacted with the sensing element, biotinylated antimouse antibody and anti-biotin conjugated gold nanoparticles (label) Fixed interface nuclei from HPV infected cells were hybridized with the sensing element and reacted sequentially with mouse antifluorescein antibody, biotinylated antimouse antibody and gold anti-biotin IgG (label) In situ hybridization of sensing element to analyte and labeling of the complex by anti-biotin conjugated silver nanoparticles Chicken skeletal muscle tissue was sequentially reacted with the sensing element and anti-mouse antibody conjugated silver nanoparticles (label) Identification of analyte by the barcode patterns

Toxin A[209]

CD4 lymphocyte[209]

HER-2/neu tyrosine kinase (transmembrane glycoprotein) in breast carcinoma tissue[209] Human Papilloma Virus in infected cells[209]

Targeted DNA site[205] Ryanodine receptors in chicken skeletal muscle[205] Oligonucleotides, antibodies for human and rabbit IgG[100,101]

[a] The labels are detected optically (under microscope, by CCD cameras).

Table 5. Signal amplification by localized surface plasmon resonance. Sensing Element

Method of Detection

Analyte Detected

Antibody (anti human IgG (c)) immobilized on gold film Oligonucleotides immobilized on gold film

Colloidal gold enhanced SPR Spectroscopy

Antigen (human IgG (c)[214]

Colloidal gold enhanced SPR Spectroscopy

Oligonucleotides, complementary to those in the sensing element and labeled by gold nanoparticles [218,225] Oligonucleotides, complementary to those in the sensing element and labeled by gold nanoparticles [73] Oligonucleotides, complementary to those in the sensing element and labeled by gold nanoparticles [223] CdSe quantum dots (on the top of the multilayered structure) [226] Oligonucleotides, complementary to those in the sensing element and labeled with fluorescein [227]

Oligonucleotides immobilized on gold nanoislands Oligonucleotides immobilized on micropatterned chemoresponsive diffraction grating Gold colloidal film, covered by layer-by-layer self-assembled polyelectrolyte film Oligonucleotides immobilized on silver nanoparticle coated substrates Antibody conjugated silver and gold nanoislands on aminodextrane-coated polystyrene beads

Colloidal gold enhanced T-SPR Spectroscopy Colloidal gold enhanced diffraction measurements Colloidal gold enhanced luminescence Colloidal silver enhanced emission

Colloidal gold and silver enhanced flow cytometry

hybridization has not occurred (in the absence of target oligonucleotide), the membrane held the white latex particles only; in case of hybridization (positive result), the gold±latex complexes were trapped, and the membrane appeared red (see Fig. 15).[195] 1700

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

CD4+ and CD8+ lymphocytes [228]

Aggregation assays for the detection of proteins (based on antibody±antigen, biotin±streptavidin interaction), lectin, dinitrophenyl, K+, and several heavy-metal ions were also demonstrated (Table 3).[197±203] An interesting approach for constructing sensors is the immobilization of the gold nano-

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4

particles in molecularly imprinted polymers.[204] The polymer swells upon selective binding of the analyte, the distance between immobilized gold nanoparticles increases, and hence their T-LSPR spectrum changes.

Figure 14. a) TLSPR spectra of gold nanoislands (2.5 nm nominal thickness, evaporated on quartz) before (bottom curve) and after (top curve) the self-assembly of cyclic disulfide molecules. b) Difference spectrum, obtained by subtraction of the bottom spectrum from the top one in (a); the dashed line is a spectrum of a thick layer of the molecules, obtained by evaporation of a drop on quartz (original spectrum divided by 6). Reproduced with permission [65].

5.3. Labeling by Metallic Nanoparticles Manifesting Localized Surface Plasmon Resonance

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Advantage has been taken of the extremely high scattering flux (an 80 nm diameter metallic nanoparticle is estimated to have a scattering flux which corresponds to those of 106 fluorescein molecules[205]) and stable metallic nanoparticles to replace (or complement) established radioactive, fluorescent, chemiluminescent, or colorimetric labels, used in biochemistry, cell biology, and medical diagnostic applications.[205±210] The ultrabright, non-bleaching nanoparticles can be observed individually and can be prepared with a scattering peak at any color of the visible spectrum, thus enabling the detection of single molecules and/or multiple targets. Indeed, the detection of chicken ryanodine receptor and a targeted DNA site on Drosophila polytene chromosomes was demonstrated by insitu hybridization, cytological assay, and immunoassay, using antibody-coated silver nanoparticles as optical labels (see Table 4).[205] The same type of metallic nanoparticles were applied in a microarray-based DNA hybridization assay to label the biotinylated oligonucleotides, and a sensitivity 60 times greater than that of fluorescent labels was achieved.[206] Metallic-nanoparticle-labeled oligonucleotides had a sharper melting (dehybridization) profile than those labeled by fluorophores. The relatively sudden change of optical signature due to the separation of DNA-linked nanoparticles at a sequence- and length-specific ªmeltingº temperature allows the discrimination of single-base-pair mismatches with a selectivity three times greater than that observed with fluorophore labeled targets (Fig. 16).[211] This observation, combined with a signal amplification method based on the nanoparticle-promoted reduction of silver(I) and the use of a conventional flatbed scanner as a reader is the basic principle of the new ªscanometricº chip-based detection system for DNA, which has single mismatch selectivity and a sensitivity

Figure 15. Schematic representation of gold nanoparticle/latex microsphere-based colorimetric DNA detection method. Reproduced with permission from [195].

Adv. Mater. 2004, 16, No. 19, October 4

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1701

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

Figure 16. Dissociation of fluorescein-labeled (A) and nanoparticle labeled (B) targets from the surfaces of oligonucleotide-functionalized glass beads, monitored by quantifying the dissociated labels at k = 520 nm (in the solution above the beads) as a function of temperature. The curves on the right in A,B describe the dissociation of target and probe from a perfectly complementary capture strand and the curves on the left describe the analogous process for a capture strand with a single mismatch. The experiment clearly shows the advantages of nanoparticle labeling. i) The temperature range of dehybridization of nanoparticle complexes is much narrower than that for the fluorophore probes. The sharp melting curve translates into high selectivity, because a temperature can be selected so that a high proportion of labeled probe remains hybridized to perfectly matched capture strands while most of the probe is dehybridized from mismatched capture strands. (See and compare the vertical dotted lines in A,B.) ii) The midpoint temperature in the melting curves (Tm) for the complexes with nanoparticle probes also is significantly higher compared to that with fluorophore probes, and thus the critical concentration below which the complex spontaneously melts is lower. This feature increases the sensitivity in the assay. Reproduced with permission from [211]; copyright 2000, the American Association for the Advancement of Science.

exceeding that of the analogous fluorophore system by two orders of magnitude.[211] Other examples for nanoparticle labeling are summarized in Table 4. The commercial availability of gold nanoparticles and gold-nanoparticle-labeled molecular-recognition elements merits mentioning in passing.[212]

5.4. Signal Amplification by Localized Surface Plasmon Resonance Metallic nanoparticles have been employed to enhance the signals in several optical techniques. Significant amplification of the propagating surface plasmon resonance (SPR) spectroscopic signal by gold (or silver) nanoparticles will be discussed here. An SPR-based sensor device contains four to six stratified layers of different materials (Fig. 17).[6,7] In the Kretschmann configuration, layer 1 is a right-angled or hemispherical prism (which is sometimes optically coupled with a glass slide 1702

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Figure 17. Schematics showing the five stratified layers of different materials in a gold-nanoparticle-enhanced SPR based sensor device. Layer 1 is a right-angled (or hemispherical) prism optically coupled (by a refractive index matching fluid) with a glass slide made of the prism material, layer 2 is a thin film (~ 1.0±1.5 nm) of a binding material (like Cr, Ti, or W), layer 3 is the 35±60 nm thick gold or silver film which generates the propagating surface plasmons, layer 4 is a SAM which incorporates the molecular recognition element. Layer 5 contains the analyte, attached (covalently or electrostatically) to 15±30 nm diameter gold or silver nanoparticles.

made of the prism material), layer 2 is a thin film (~ 1.0± 1.5 nm) of a binding material (like Cr, Ti, or W), usually vacuum deposited onto the surface of the prism base (or the optically coupled glass slide), layer 3 is the 35±60 nm thick gold or silver film which generates the propagating surface plasmons, layer 4 is a SAM which incorporates the molecular recognition element. Layer 5 contains the analyte, in the absence or in the presence of attached (covalently or electrostatically) 15±30 nm diameter gold or silver nanoparticles (Fig. 17). Sensing involves: i) the experimental determination of the surface plasmon resonance spectra of layers 3±5 (typically by scanning the relative intensity of the reflected light as a function of the incident angle = angle-resolved SPR); ii) construction of an appropriate model that describes the behavior of the experimentally probed sample well; iii) generation of expected theoretical data plots (i.e., reflected light versus incident angle plots) from the model; and iv) comparison of the experimentally and theoretically obtained data plots to obtain the needed parameters from the best fit for each layer.[213] Gold nanoparticles attached to the analyte (layer 5 in the above example) caused ”25 signal amplification.[214] The signal amplification was explained by the interaction of localized and propagating surface plasmon resonances[213,215±217]and have been shown to depend on the type of substrate and nanoparticles (gold or silver) as well as on their sizes, shapes, and distance from the layer generating the propagating surface plasmons (layer 5, in the above example).[213±217] A typical example of metallic-nanoparticle-enhanced SPR is the derivatization of Au films by an antibody (a-humanIgG), followed by either direct binding of an antigen±Au conjugate or by binding of a free antigen and then a secondary antibody±Au conjugate. In both cases, the signal enhancement was large enough to allow the detection of solution protein concentration in picomolar range.[214] Based on a similar principle, DNA hybridization was measured.[218] In this work,

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4

oligonucleotide probes were linked covalently to the gold surface, then hybridized to one half of the target DNA molecule. In the next step another oligonucleotide, with a sequence complementary to the other half of the target, was added, either untagged or conjugated with Au nanoparticles. In addition to the plotted SPR curves, 2D SPR images of the measured spots provided an easy way to visualize the results. The detection limit in these experiments was ~ 10 pM (8 ” 108molecules cm±1).[218] The attachment of gold nanoparticles to the materials adsorbed onto SAMs has enhanced the sensitivity of the T-LSPR,[68] in a manner quite analogous to that observed for SPR spectroscopy.[214] Recently, this approach has been utilized for the detection of DNA hybridization (Fig. 18).[68] The

self-assembly of oligoA (12 bases of homogeneous sequence containing only adenine, thiolated at the 5¢ end) and mercaptohexanol spacer onto gold nanoislands evaporated onto a glass slide (Fig. 18), as well as the DNA hybridization (by oligoT = 12 bases of homogeneous sequence containing only thymine, thiolated at the 5¢ end) could be readily sensed by T-LSPR (Fig. 18). The sensitivity of T-LSPR was found to dramatically increase with the functionalization of the complementary DNA by 11.7 ± 1.9 nm diameter gold nanoparticles (Fig. 18). Proximity of the gold nanoparticles to the gold nanoislands could be readily seen by transmission electron microscopy (TEM).[68] Further examples of LSPR-amplified techniques are collected in Table 5.

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

6. Conclusion Exploitation of localized surface plasmons as nanoscale optical devices and sensors has reached a significant level of recognition, even though most of the references cited in this review have been published in this millennium. Reports on the development of innovative concepts, unprecedented applications, and fabrication of sensors with previously unimagined selectivity and sensitivity are seeing daylight at an exponentially increasing rate. Nevertheless, there is much more to be done, much more to be hoped for. Full realization of the potential of surface plasmons requires a much better understanding of the underlying fundamentals and a greater cooperation between physicists, chemists, and material scientists. It is sincerely hoped that this review will inspire researchers to make their own contributions to this exciting and highly relevant fi

±

Figure 18. Top: Schematic representation of the gold nanoparticle enhanced T-LSPR, utilized for DNA detection. Single-stranded DNA was self-assembled onto gold nanoislands on a glass slide with the subsequent introduction of mercaptohexanol spacer molecules (short thiols in the Fig.) and then hybridized by the complementary DNA, which itself was functionalized by 11.7 ± 1.9 nm diameter gold nanoparticles. The diameter of the gold nanoislands and the gold nanoparticles are drawn to scale, but the length of DNA was increased for clarity. Bottom: T-LSPR spectra of gold nanoislands in TE (Tris-EDTA) buffer prior (1) and subsequent to the self-assembly of single-stranded DNA and mercaptohexanol and hybridization by the complementary DNA, unlabeled (2) or labeled (3) by gold nanoparticles. Reproduced with permission from [68]; copyright 2003, the American Chemical Society.

Adv. Mater. 2004, 16, No. 19, October 4

Received: February 26, 2004 Final version: June 15, 2004

[1] U. Kreibig, M. Vollmer, Optical Properties of Metal Clusters, Springer-Verlag, Berlin, Germany 1995. [2] S. Underwood, P. Mulvaney, Langmuir 1994, 10, 3427. [3] P. Mulvaney, Langmuir 1996, 12, 788. [4] D. G. Richards, D. L. McMillin, E. A. Mein, C. D. Nelson, Int. J. Neurosci. 2002, 112, 31. [5] M. Matsumoto, H. Yoshimura, V. S. Kulkarni, K. Nagayama, Colloid Polym. Sci. 1990, 268, 1174. [6] M.-C. Daniel, D. Astruc, Chem. Rev. 2004, 104, 293. [7] J. M. Brockman, B. P. Nelson, R. M. Corn, Annu. Rev. Phys. Chem. 2000, 51, 41. [8] S. Wang, S. Boussaad, N. J. Tao, in Biomolecular Films: Design, Function and Applications, Vol. 111 (Ed: J. F. Rusling), Marcel Dekker, New York 2003, p. 213. [9] J. Homola, S. S. Yee, G. Gauglitz, Sens. Actuators, B 1999, 54, 3. [10] D. F. Evans, H. Wennerstrom, The Colloidal Domain: Where Physics, Chemistry, and Technology Meet, VCH, New York 1999. [11] D. Goia, J. Mater. Chem. 2004, 14, 451. [12] J. Turkevich, P. C. Stevenson, J. Hillier, Discuss. Faraday Soc. 1951, 55. [13] D. A. Handley, in Colloidal Gold: Principles, Methods, and Applications, Vol. 1 (Ed: M. A. Hayat), Academic Press, Inc., San Diego, CA 1989, p. 13. [14] D. Peters, J. Mater. Chem. 1996, 6, 1605.

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1703

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

1704

[15] T. Fujimoto, S. Terauchi, H. Umehara, I. Kojima, W. Henderson, Chem. Mater. 2001, 13, 1057. [16] K. Okitsu, A. Yue, S. Tanabe, H. Matsumoto, Y. Yobiko, Langmuir 2001, 17, 7717. [17] V. G. Pol, D. N. Srivastava, O. Palchik, V. Palchik, M. A. Slifkin, A. M. Weiss, A. Gedanken, Langmuir 2002, 18, 3352. [18] A. Kulak, S. A. Davis, E. Dujardin, S. Mann, Chem. Mater. 2003, 15, 528. [19] V. G. Pol, A. Gedanken, J. Calderon-Moreno, Chem. Mater. 2003, 15, 1111. [20] G. H. Fu, W. P. Cai, C. X. Kan, C. C. Li, L. Zhang, Appl. Phys. Lett. 2003, 83, 36. [21] W. Chen, J. Y. Zhang, Y. Di, Z. M. Wang, Q. Fang, W. P. Cai, Appl. Surf. Sci. 2003, 211, 280. [22] M. Giersig, P. Mulvaney, Langmuir 1993, 9, 3408. [23] M. Brust, M. Walker, D. Bethell, D. J. Schiffrin, R. Whyman, J. Chem. Soc. Chem. Commun. 1994, 801. [24] M. J. Hostetler, S. J. Green, J. J. Stokes, R. W. Murray, J. Am. Chem. Soc. 1996, 118, 4212. [25] A. C. Templeton, M. P. Wuelfing, R. W. Murray, Acc. Chem. Res. 2000, 33, 27. [26] M. M. Alvarez, J. T. Khoury, T. G. Schaaff, M. N. Shafigullin, I. Vezmar, R. L. Whetten, J. Phys. Chem. B 1997, 101, 3706. [27] Z. L. Wang, Characterization of Nanophase Materials, Vol. 1, WileyVCH, Weinheim, Germany 2000, p. 406. [28] L. Q. Wang, G. J. Exarhos, J. Liu, in Characterization of Nanophase Materials, Vol. 1 (Ed: Z. L. Wang), Wiley-VCH, Weinheim, Germany 2000, p. 243. [29] N. R. Jana, X. G. Peng, J. Am. Chem. Soc. 2003, 125, 14 280. [30] S. Stoeva, K. J. Klabunde, C. M. Sorensen, I. Dragieva, J. Am. Chem. Soc. 2002, 124, 2305. [31] H. Choo, E. Cutler, Y. S. Shon, Langmuir 2003, 19, 8555. [32] P. Pengo, S. Polizzi, M. Battagliarin, L. Pasquato, P. Scrimin, J. Mater. Chem. 2003, 13, 2471. [33] C. A. Waters, A. J. Mills, K. A. Johnson, D. J. Schiffrin, Chem. Commun. 2003, 540. [34] E. Hutter, J. H. Fendler, Chem. Commun. 2002, 378. [35] R. L. Whetten, J. T. Khoury, M. M. Alvarez, S. Murthy, I. Vezmar, Z. L. Wang, P. W. Stephens, C. L. Cleveland, W. D. Luedtke, U. Landman, Adv. Mater. 1996, 8, 428. [36] R. L. Whetten, M. N. Shafigullin, J. T. Khoury, T. G. Schaaff, I. Vezmar, M. M. Alvarez, A. Wilkinson, Acc. Chem. Res. 1999, 32, 397. [37] C. Radloff, N. J. Halas, Appl. Phys. Lett. 2001, 79, 674. [38] F. Caruso, Adv. Mater. 2001, 13, 11. [39] J. H. Fendler, Membrane Mimetic Approach to Advanced Materials, Vol. 113, Springer-Verlag, Berlin, Germany 1992. [40] J. H. Fendler, Membrane Mimetic Chemistry, Characterizations and Applications of Micelles, Microemulsions, Monolayers, Bilayers, Vesicles, Host-Guest Systems and Polyions, John Wiley, New York 1982. [41] J. Tanori, M. P. Pileni, Langmuir 1997, 13, 639. [42] M. P. Pileni, Ber. Bunsen-Ges. 1997, 101, 1578. [43] A. Filankembo, M. P. Pileni, J. Phys. Chem. B 2000, 104, 5865. [44] D. A. Tomalia, A. M. Naylor, W. A. Goddard III, Angew. Chem. Int. Ed. Engl. 1990, 29, 138. [45] J. M. J. Frechet, C. J. Hawker, I. Gitsov, J. W. Leon, J. Macromol. Sci. Pure Appl. Chem. 1996, a33, 1399. [46] F. S. Bates, G. H. Fredrickson, Phys. Today 1999, 52, 32. [47] J. W. Zhai, W. S. Shi, L. Y. Zhang, X. Yao, Ferroelectrics 1999, 230, 541. [48] V. M. Rudoy, I. V. Yaminskii, O. V. Dement'eva, V. A. Ogarev, Colloid J. 1999, 61, 800. [49] H. Colfen, H. Schnablegger, M. Antonietti, J. Colloid Interface Sci. 1999, 212, 197. [50] S. T. Selvan, T. Hayakawa, M. Nogami, M. Moller, J. Phys. Chem. B 1999, 103, 7441.

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

[51] C. Nardin, S. Thoeni, J. Widmer, M. Winterhalter, W. Meier, Chem. Commun. 2000, 1433. [52] C. Sonnichsen, T. Franzl, T. Wilk, G. von Plessen, J. Feldmann, Phys. Rev. Lett. 2002, 88, 077 402. [53] G. Gorodyska, A. Kiriy, S. Minko, C. Tsitsilianis, M. Stamm, Nano Lett. 2003, 3, 365. [54] S. H. Feng, R. R. Xu, Acc. Chem. Res. 2001, 34, 239. [55] M. M. Lencka, M. Oledzka, R. E. Riman, Chem. Mater. 2000, 12, 1323. [56] N. R. Jana, L. Gearheart, C. J. Murphy, Adv. Mater. 2001, 13, 1389. [57] C. J. Johnson, E. Dujardin, S. A. Davis, C. J. Murphy, S. Mann, J. Mater. Chem. 2002, 12, 1765. [58] N. R. Jana, L. A. Gearheart, S. O. Obare, C. J. Johnson, K. J. Edler, S. Mann, C. J. Murphy, J. Mater. Chem. 2002, 12, 2909. [59] J. X. Gao, C. M. Bender, C. J. Murphy, Langmuir 2003, 19, 9065. [60] B. Nikoobakht, M. A. El-Sayed, Chem. Mater. 2003, 15, 1957. [61] N. Taub, O. Krichevski, G. Markovich, J. Phys. Chem. B 2003, 107, 11 579. [62] Y. G. Sun, Y. N. Xia, Anal. Chem. 2002, 74, 5297. [63] Y. Sun, Y. Xia, J. Am. Chem. Soc. 2004, 126, 3892. [64] J. E. Mahan, Physical Vapor Deposition of Thin Films, John Wiley & Sons, New York 1999. [65] G. Kalyuzhny, A. Vaskevich, M. A. Schneeweiss, I. Rubinstein, Chem. Eur. J. 2002, 8, 3850. [66] G. Kalyuzhny, M. A. Schneeweiss, A. Shanzer, A. Vaskevich, I. Rubinstein, J. Am. Chem. Soc. 2001, 123, 3177. [67] G. Kalyuzhny, A. Vaskevich, G. Ashkenasy, A. Shanzer, I. Rubinstein, J. Phys. Chem. B 2000, 104, 8238. [68] E. Hutter, M.-P. Pileni, J. Phys. Chem. B 2003, 107, 6497. [69] A. J. Haes, R. P. Van Duyne, A Highly Sensitive and Selective Surface-enhanced Nanobiosensor, in Materials Research Society Symp. Proc. 2002, 123, pp. O3.1.1. [70] M. D. Malinsky, K. L. Kelly, G. C. Schatz, R. P. Van Duyne, J. Am. Chem. Soc. 2001, 123, 1471. [71] M. D. Malinsky, K. L. Kelly, G. C. Schatz, R. P. Van Duyne, J. Phys. Chem. B 2001, 105, 2343. [72] A. D. McFarland, R. P. Van Duyne, Nano Lett. 2003, 3, 1057. [73] J. C. Riboh, A. J. Haes, A. D. McFarland, C. R. Yonzon, R. P. Van Duyne, J. Phys. Chem. B 2003, 107, 1772. [74] W. Gotschy, K. Vonmetz, A. Leitner, F. R. Aussenegg, Appl. Phys. B 1996, 63, 381. [75] N. Felidj, J. Aubard, G. Levi, J. R. Krenn, M. Salerno, G. Schider, B. Lamprecht, A. Leitner, F. R. Aussenegg, Phys. Rev. B 2002, 65, 075 419. [76] W. Rechberger, A. Hohenau, A. Leitner, J. R. Krenn, B. Lamprecht, F. R. Aussenegg, Opt. Commun. 2003, 220, 137. [77] C. L. Haynes, A. D. McFarland, L. Zhao, R. P. Van Duyne, G. C. Schatz, L. Gunnarsson, J. Prikulis, B. Kasemo, M. Kall, J. Phys. Chem. B 2003, 107, 7337. [78] M. Akahashi, Y. Miyahara, Chem. Lett. 1998, 917. [79] M. S. Yeh, Y. S. Yang, Y. P. Lee, H. F. Lee, Y. H. Yeh, C. S. Yeh, J. Phys. Chem. B 1999, 103, 6851. [80] A. Henglein, J. Phys. Chem. 1993, 97, 5457. [81] B. G. Ershov, E. Janata, A. Henglein, A. Fojtik, J. Phys. Chem. B 1993, 97, 4589. [82] C. H. Bae, S. H. Nam, S. M. Park, Appl. Surf. Sci. 2002, 197, 628. [83] Y. H. Chen, C. S. Yeh, Colloids Surf., A 2002, 197, 133. [84] A. V. Kabashin, M. Meunier, C. Kingston, J. H. T. Luong, J. Phys. Chem. B 2003, 107, 4527. [85] F. Mafune, J. Kohno, Y. Takeda, T. Kondow, J. Phys. Chem. B 2001, 105, 9050. [86] F. Mafune, J. Kohno, Y. Takeda, T. Kondow, J. Phys. Chem. B 2003, 107, 12 589. [87] F. Mafune, J. Kohno, Y. Takeda, T. Kondow, H. Sawabe, J. Phys. Chem. B 2000, 104, 9111.

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4

[88] F. Mafune, J. Y. Kohno, Y. Takeda, T. Kondow, H. Sawabe, J. Phys. Chem. B 2000, 104, 8333. [89] F. Mafune, J. Y. Kohno, Y. Takeda, T. Kondow, J. Phys. Chem. B 2002, 106, 7575. [90] F. Mafune, J. Y. Kohno, Y. Takeda, T. Kondow, H. Sawabe, J. Phys. Chem. B 2001, 105, 5114. [91] I. Lee, S. W. Han, K. Kim, Chem. Commun. 2001, 1782. [92] C. Y. Wang, H. Y. Fang, J. B. He, W. L. Zhou, K. L. Stokes, J. Colloid Interface Sci. 2003, 260, 440. [93] Y. Niidome, K. Nishioka, H. Kawasaki, S. Yamada, Chem. Commun. 2003, 18, 2376. [94] M. Maillard, P. Huang, L. Brus, Nano Lett. 2003, 3, 1611. [95] R. C. Jin, Y. C. Cao, E. C. Hao, G. S. Metraux, G. C. Schatz, C. A. Mirkin, Nature 2003, 425, 487. [96] S. Link, M. A. El-Sayed, Annu. Rev. Phys. Chem. 2003, 54, 331. [97] X. Y. Jiang, R. Ferrigno, M. Mrksich, G. M. Whitesides, J. Am. Chem. Soc. 2003, 125, 2366. [98] P. Zhang, T. K. Sham, Appl. Phys. Lett. 2003, 82, 1778. [99] Y. Sun, Y. Xia, Nano Lett. 2003, 3, 1569. [100] C. D. Keating, M. J. Natan, Adv. Mater. 2003, 15, 451. [101] S. R. Nicewarner-Peæa, R. G. Freeman, B. D. Reiss, L. He, D. J. Peæa, I. D. Walton, R. Cromer, C. D. Keating, M. J. Natan, Science 2001, 294, 137. [102] X. Qu, Z. Peng, X. Jiang, S. Dong, Langmuir 2004, 20, 2519. [103] E. Hutter, Ph.D. Thesis, Clarkson University, Potsdam, NY, 2001. [104] G. Xu, M. Tazawa, P. Jin, S. Nakao, K. Yoshimura, Appl. Phys. Lett. 2003, 82, 3811. [105] G. Mie, Ann. Phys. 1908, 25, 377 [106] K. L. Kelly, E. Coronado, L. L. Zhao, G. C. Schatz, J. Phys. Chem. B 2003, 107, 668. [107] S. Link, M. A. El-Sayed, J. Phys. Chem. B 1999, 103, 4212. [108] A. C. Templeton, J. J. Pietron, R. W. Murray, P. Mulvaney, J. Phys. Chem. B 2000, 104, 564. [109] C. F. Bohren, D. R. Huffman, Absorption and Scattering of Light by Small Particles, Vol. 1, John Wiley & Sons, Inc., New York 1998. [110] S. Link, M. A. El-Sayed, Int. Rev. Phys. Chem. 2000, 19, 409. [111] Y. G. Sun, Y. N. Xia, Proc. SPIE±Int. Soc. Opt. Eng. 2003, 5221, 164. [112] B. M. I. Vanderzande, M. R. Bohmer, L. G. J. Fokkink, C. Schonenberger, J. Phys. Chem. B 1997, 101, 852. [113] J. P. Kottmann, O. J. F. Martin, Opt. Express 2000, 6, 213. [114] E. Hao, R. C. Bailey, G. C. Schatz, J. T. Hupp, S. Li, Nano Lett. 2004, 4, 327. [115] I. O. Sosa, C. Noguez, R. G. Barrera, J. Phys. Chem. B 2003, 107, 6269. [116] E. Hao, S. Li, R. C. Bailey, S. Zou, G. C. Schatz, J. T. Hupp, J. Phys. Chem. B 2004, 108, 1224. [117] J. J. Mock, D. R. Smith, M. Barbic, S. J. Oldenburg, D. A. Schultz, S. Schultz, Proc. SPIE±Int. Soc. Opt. Eng. 2003, 5221, 66. [118] B. Lamprecht, G. Schjider, R. T. Lechner, H. Ditlbacher, J. R. Krenn, A. Leitner, F. R. Aussenegg, Phys. Rev. Lett. 2000, 84, 4721. [119] S. A. Maier, M. L. Brongersma, P. G. Kik, H. A. Atwater, Phys. Rev. B 2002, 65, 193 408. [120] K. H. Su, Q. H. Wei, X. Zhang, J. J. Mock, D. R. Smith, S. Schultz, Proc. SPIE±Int. Soc. Opt. Eng. 2003, 5221, 108. [121] J. R. Krenn, A. Dereux, J. C. Weeber, E. Bourillot, Y. Lacroute, J. P. Goudonnet, G. Schider, W. Gotschy, A. Leitner, F. R. Aussenegg, C. Girard, Phys. Rev. Lett. 1999, 82, 2590. [122] S. A. Maier, M. L. Brongersma, P. G. Kik, S. Meltzer, A. A. G. Requicha, H. A. Atwater, Adv. Mater. 2001, 13, 1501. [123] P. C. Andersen, K. L. Rowlen, Appl. Spectrosc. 2002, 56, 124A. [124] W. L. Barnes, A. Dereux, T. W. Ebbesen, Nature 2003, 424, 824. [125] B. Lamprecht, J. R. Krenn, G. Schider, H. Ditlbacher, M. Salerno, N. Felidj, A. Leitner, F. R. Aussenegg, Appl. Phys. Lett. 2001, 79, 51. [126] E. Devaux, T. W. Ebbesen, J. C. Weeber, A. Dereux, Appl. Phys. Lett. 2003, 83, 4936.

Adv. Mater. 2004, 16, No. 19, October 4

[127] H. Raether, Surface Plasmons on Smooth and Rough Surfaces and on Gratings, Springer-Verlag, Berlin, Germany 1988. [128] R. M. Dickson, L. A. Lyon, J. Phys. Chem. B 2000, 104, 6095. [129] S. A. Maier, P. G. Kik, H. A. Atwater, S. Melyzer, E. Harel, B. Koel, A. A. G. Requicha, Nat. Mater. 2003, 2, 229. [130] P. Dawson, A. F. Puygranier, J. P. Goudonnet, Phys. Rev. B 2001, 63, 205 410. [131] T. W. Ebbesen, H. J. Lezec, H. F. Ghaemi, T. Thio, P. A. Wolff, Nature 1998, 391, 667. [132] H. F. Ghaemi, T. Thio, D. E. Grupp, T. W. Ebbesen, H. J. Lezec, Phys. Rev. B 1998, 58, 6779. [133] U. Schroter, D. Heitmann, Phys. Rev. B 1998, 58, 15 419. [134] J. A. Porto, F. J. Garcia-Vidal, J. B. Pendry, Phys. Rev. Lett. 1999, 83, 2845. [135] W.-C. Tan, T. W. Preist, R. J. Sambles, Phys. Rev. B 2000, 62, 11 134. [136] E. Popov, M. Neviere, S. Enoch, R. Reinisch, Phys. Rev. B 2000, 62, 16 100. [137] W.-C. Liu, D. P. Tsai, Phys. Rev. B 2002, 65, 155 423. [138] F. J. Garciavidal, H. J. Lezec, T. W. Ebbesen, L. Martinmoreno, Phys. Rev. Lett. 2003, 90, 213 901. [139] F. J. Garciavidal, L. Martinmoreno, H. J. Lezec, T. W. Ebbesen, Appl. Phys. Lett. 2003, 83, 4500. [140] J. B. Pendry, Phys. Rev. Lett. 2000, 85, 3966. [141] L. V. Panina, A. N. Grigorenko, D. P. Makhnovskiy, Phys. Rev. B 2002, 66, 155 411. [142] V. A. Podolskiy, Opt. Express 2003, 11, 735. [143] D. R. Smith, W. J. Padilla, D. C. VGier, S. C. Nemat-Nasser, S. Shultz, Phys. Rev. Lett. 2000, 84, 4184. [144] N. Fang, Z. Liu, T.-J. Yen, X. Zhang, Opt. Express 2003, 11, 682. [145] H. Ditlbacher, J. R. Krenn, B. Lamprecht, A. Leitner, F. R. Aussenegg, Opt. Lett. 2000, 25, 563. [146] C. A. Constantine, K. M. Gattas-Asfura, S. V. Mello, G. Crespo, V. Rastogi, T. C. Cheng, J. J. De Frank, R. M. Leblanc, J. Phys. Chem. B 2003, 107, 13 762. [147] B. Hecht, B. Sick, U. P. Wild, V. Deckert, R. Zenobi, O. J. F. Martin, D. W. Pohl, J. Chem. Phys. 2000, 112, 7761. [148] R. C. Dunn, Chem. Rev. 1999, 99, 2891. [149] F. Zenhausern, M. P. Oboyle, H. K. Wickramasinghe, Appl. Phys. Lett. 1994, 65, 1623. [150] Y. Martin, F. Zenhausern, H. K. Wickramasinghe, Appl. Phys. Lett. 1996, 68, 2475. [151] R. E. Larsen, H. Metiu, J. Chem. Phys. 2001, 114, 6851. [152] M. Specht, J. D. Pedarnig, W. M. Heckl, T. W. Hansch, Phys. Rev. Lett. 1992, 68, 476. [153] J. R. Krenn, M. Salerno, N. Felidj, B. Lamprecht, G. Schider, A. Leitner, F. R. Aussenegg, J. C. Weeber, A. Dereux, J. P. Goudonnet, J. Microsc. 2001, 202, 122. [154] S. I. Bozhevolnyi, I. I. Smolyaninov, A. V. Zayats, Phys. Rev. B 1995, 51, 17 916. [155] I. I. Smolyaninov, D. L. Mazzoni, J. Mait, C. C. Davis, Phys. Rev. B 1997, 56, 1601. [156] I. I. Smolyaninov, D. L. Mazzoni, C. C. Davis, Phys. Rev. Lett. 1996, 77, 3877. [157] G. A. Wurtz, J. Hranisavljevic, G. P. Wiederrecht, Nano Lett. 2003, 3, 1511. [158] I. I. Smolyaninov, A. V. Zayats, A. Gungor, C. C. Davis, Phys. Rev. Lett. 2002, 88, 187 402. [159] E. Yablonovitch, J. Phys. Condens. Matter 1993, 5, 2443. [160] H. Benisty, C. Weisbuch, D. Labilloy, M. Rattier, Appl. Surf. Sci. 2000, 164, 205. [161] Y. Z. Shen, C. S. Friend, Y. Jiang, D. Jakubczyk, J. Swiatkiewicz, P. N. Prasad, J. Phys. Chem. B 2000, 104, 7577. [162] C. Lopez, Adv. Mater. 2003, 15, 1679. [163] I. Soten, H. Miguez, S. M. Yang, S. Petrov, N. Coombs, N. Tetreault, N. Matsuura, H. E. Ruda, G. A. Ozin, Adv. Funct. Mater. 2002, 12, 71.

http://www.advmat.de

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

1705

REVIEW

E. Hutter, J. H. Fendler/Exploitation of Localized Surface Plasmon Resonance

1706

[164] D. J. Norris, Y. A. Vlasov, Adv. Mater. 2001, 13, 371. [165] S. C. Kitson, W. L. Barnes, J. R. Sambles, Phys. Rev. B 1995, 52, 11 441. [166] W. L. Barnes, S. C. Kitson, J. R. Sambles, Phys. Rev. Lett. 1996, 77, 2670. [167] W. L. Barnes, T. W. Preist, S. J. Kitson, J. R. Sambles, Phys. Rev. B 1996, 54, 6227. [168] W. L. Barnes, S. C. Kitson, T. W. Preist, J. R. Sambles, J. Opt. Soc. Am. A 1997, 14, 1654. [169] S. G. Penn, L. He, M. J. Natan, Curr. Opin. Chem. Biol. 2003, 7, 609. [170] A. Campion, P. Kambhampati, Chem. Soc. Rev. 1998, 27, 241. [171] Z. Q. Tian, B. Ren, D. Y. Wu, J. Phys. Chem. B 2002, 106, 9463. [172] M. Osawa, Bull. Chem. Soc. Jpn. 1997, 70, 2861. [173] S. A. Levi, A. Mourran, J. P. Spatz, F. C. J. M. van Veggel, D. N. Reinhoudt, M. Moller, Chem. Eur. J. 2002, 8, 3808. [174] N. Calander, Anal. Chem. 2004, 76, 2168. [175] S. Ekgasit, C. Thammacharoen, F. Yu, W. Knoll, Anal. Chem. 2004, 76, 2210. [176] P. Englebienne, Analyst 1998, 123, 1599. [177] P. Englebienne, A. Van Hoonacker, M. Verhas, Analyst 2001, 126, 1645. [178] S. Watanabe, M. Sonobe, M. Arai, Y. Tazume, T. Matsuo, T. Nakamura, K. Yoshida, Chem. Commun. 2002, 2866. [179] M. Schwarzott, H. Engelhardt, T. Kluhspies, D. Baurecht, D. Naumann, U. P. Fringeli, Langmuir 2003, 19, 7451. [180] F. Frederix, J.-M. Friedt, K.-H. Choi, W. Laureyn, A. Campitelli, D. Mondelaers, G. Maes, G. Borghs, Anal. Chem. 2003, 75, 6894. [181] G. Raschke, S. Kowarik, T. Franzl, S. Sonnichsen, T. A. Klar, J. Feldmann, Nano Lett. 2003, 3, 935. [182] S. M. Briglin, T. Gao, N. S. Lewis, Langmuir 2004, 20, 299. [183] F. Meriaudeau, T. Downey, A. Wig, A. Passian, M. Buncick, T. L. Ferrell, Sens. Actuators, B 1999, 54, 106. [184] F. Meriaudeau, A. Wig, A. Passian, T. Downey, M. Buncick, T. L. Ferrell, Sens. Actuators, B 2000, 69, 51. [185] A. J. Haes, R. P. Van Duyne, J. Am. Chem. Soc. 2002, 124, 10 596. [186] M. Himmelhaus, H. Takei, Sens. Actuators, B 2000, 63, 24. [187] M. Lahav, A. Vaskevich, I. Rubinstein, Langmuir 2004, 20, 7365. [188] C. L. Haynes, R. P. Van Duyne, J. Phys. Chem. B 2001, 105, 5599. [189] R. A. Reynolds III, C. A. Mirkin, R. L. Letsinger, J. Am. Chem. Soc. 2000, 122, 3795. [190] J. J. Storhoff, R. Elghanian, R. C. Mucic, C. A. Mirkin, R. L. Letsinger, J. Am. Chem. Soc. 1998, 120, 1959. [191] A. C. Mirkin, L. R. Letsinger, C. R. Mucic, J. J. Storhoff, Nature 1996, 382, 607. [192] J. J. Storhoff, A. A. Lazarides, R. C. Mucic, C. A. Mirkin, R. L. Letsinger, G. C. Schatz, J. Am. Chem. Soc. 2000, 122, 4640. [193] R. C. Mucic, J. J. Storhoff, C. A. Mirkin, R. L. Letsinger, J. Am. Chem. Soc. 1998, 120, 12 674. [194] I. Tokareva, E. Hutter, J. Am. Chem. Soc., in press. [195] R. A. Reynolds III, C. A. Mirkin, R. L. Letsinger, Pure Appl. Chem. 2000, 72, 229. [196] J.-M. Nam, S.-J. Park, C. A. Mirkin, J. Am. Chem. Soc. 2002, 124, 3820. [197] N. T. K. Thanh, Z. Rosenzweig, Anal. Chem. 2002, 74, 1624. [198] S. Connolly, D. Fitzmaurice, Adv. Mater. 1999, 11, 1202. [199] W. Shenton, S. A. Davis, S. Mann, Adv. Mater. 1999, 11, 449. [200] H. Otsuka, Y. Akiyama, Y. Nagasaki, K. Kataoka, J. Am. Chem. Soc. 2001, 123, 8226. [201] Y. Kim, R. C. Johnson, J. T. Hupp, Nano Lett. 2001, 1, 165. [202] S. Y. Lin, S. W. Liu, C. M. Lin, C. H. Chen, Anal. Chem. 2002, 74, 330.

 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

[203] K. Naka, H. Roh, Y. Chujo, Langmuir 2003, 19, 5496. [204] K. Glinel, A. M. Jonas, A. Laschwesky, P. Y. Vuillaume, in Multilayer Thin Films, Sequential Assembly of Nanocomposite Materials, Vol. 1 (Eds: G. Decher, J. B. Schlenoff), Wiley-VCH, Weinheim, Germany 2003, p. 177. [205] S. Schultz, D. R. Smith, J. J. Mock, D. A. Schultz, Proc. Natl. Acad. Sci. USA 2000, 97, 996. [206] S. J. Oldenburg, C. C. Genick, K. A. Clark, D. A. Schultz, Anal. Biochem. 2002, 309, 109. [207] D. A. Schultz, Curr. Opin. Biotechnol. 2003, 14, 13. [208] J. M. Kohler, A. Csaki, J. Reichert, R. Moller, W. Straube, W. Fritzsche, Sens. Actuators, B 2001, 76, 166. [209] J. Yguerabide, E. E. Yguerabide, J. Cell. Biochem. 2001, 37, 71. [210] P. Englebienne, A. Van Hoonacker, M. Verhas, N. G. Khlebtsov, Comb. Chem. High Throughput Screening 2003, 6, 777. [211] T. A. Taton, C. A. Mirkin, R. L. Letsinger, Science 2000, 289, 1757. [212] http://www.aurion.nl; http://www.nanoprobes.com; http://www.2spi.com/catalog/chem/gold.html. [213] D. Roy, J. H. Fendler, Adv. Mater. 2004, 16, 479. [214] L. A. Lyon, M. D. Musick, M. J. Natan, Anal. Chem. 1998, 70, 5177. [215] E. Hutter, J. H. Fendler, D. Roy, J. Phys. Chem. B 2001, 105, 11 159. [216] E. Hutter, S. Cha, J. F. Liu, J. Park, J. Yi, J. H. Fendler, D. Roy, J. Phys. Chem. B 2001, 105, 8. [217] S. Chah, E. Hutter, D. Roy, J. H. Fendler, J. Yi, Chem. Phys. 2001, 272, 127. [218] L. He, M. D. Musick, S. R. Nicewarner, F. G. Salinas, S. J. Benkovic, M. J. Natan, C. D. Keating, J. Am. Chem. Soc. 2000, 122, 9071. [219] H. Yockelllelievre, E. F. Borra, A. M. Ritcey, L. V. Dasilva, Appl. Opt. 2003, 42, 1882. [220] D. E. Grupp, H. J. Lezec, T. W. Ebbesen, K. M. Pellerin, T. Thio, Appl. Phys. Lett. 2000, 77, 1569. [221] S. M. Williams, A. D. Stafford, K. R. Rodriguez, T. M. Rogers, J. V. Coe, J. Phys. Chem. B 2003, 107, 11 871. [222] T. A. Taton, R. C. Mucic, C. A. Mirkin, R. L. Letsinger, J. Am. Chem. Soc. 2000, 122, 6305. [223] R. C. Bailey, J. M. Nam, C. A. Mirkin, J. T. Hupp, J. Am. Chem. Soc. 2003, 125, 13 541. [224] S. J. Park, A. A. Lazarides, C. A. Mirkin, R. L. Letsinger, Angew. Chem. Int. Ed. 2001, 40, 2909. [225] A. Yamaguchi, S. Juodkazis, S. Matsuo, H. Misawa, Chem. Lett. 2002, 190. [226] O. Kulakovich, N. Strekal, A. Yaroshevich, S. Maskevich, S. Gaponenko, I. Nabiev, U. Woggon, M. Artemyev, Nano Lett. 2002, 2, 1449. [227] J. Malicka, I. Gryczynski, J. R. Lakowicz, Anal. Chem. 2003, 75, 4408. [228] O. Siiman, A. Burshteyn, J. Phys. Chem. B 2000, 104, 9795. [229] R. A. Shelby, D. R. Smith, S. Schultz, Science 2001, 292, 77. [230] R. Quidant, G. Badenes, S. Cheylan, R. Alcubilla, J.-C. Weeber, C. Girard, Opt. Express 2004, 12, 282. [231] J. J. Lezec, A. Degiron, E. Devaux, R. A. Linke, L. Martin-Moreno, F. J. Garcia-Vidal, T. W. Ebbesen, Science 2002, 297, 820. [232] J. Hashizume, F. Koyama, Appl. Phys. Lett. 2004, 84, 3226. [233] J. M. Steele, C. E. Moran, A. Lee, N. J. Halas, Proc. SPIE±Int. Soc. Opt. Eng. 2003, 5221, 144. [234] J. Matsui, K. Akamatsu, S. Nishiguchi, D. Miyoshi, H. Nawafune, K. Tamaki, N. Sugimoto, Anal. Chem. 2004, 76, 1310. [235] M. A. El-Sayed, Acc. Chem. Res. 2001, 34, 257.

http://www.advmat.de

Adv. Mater. 2004, 16, No. 19, October 4