1: \documentclass[12pt]{article}
2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3: \usepackage{amsmath}
4: \usepackage{amsfonts}
5: %\usepackage{achemso}
6: \usepackage{graphicx}
7:
8: \setcounter{MaxMatrixCols}{10}
9: %TCIDATA{OutputFilter=LATEX.DLL}
10: %TCIDATA{Version=5.50.0.2953}
11: %TCIDATA{<META NAME="SaveForMode" CONTENT="1">}
12: %TCIDATA{BibliographyScheme=Manual}
13: %TCIDATA{LastRevised=Monday, November 27, 2006 14:44:52}
14: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
15:
16: \setlength\voffset{-0.7in} \setlength\hoffset{-0.3in}
17: \setlength\oddsidemargin{0in} \setlength\headheight{0in}
18: \setlength\headsep{0in} \setlength\textwidth{500pt}
19: \setlength\textheight{700pt}
20: \newenvironment{proof}[1][Proof]{\noindent\textbf{#1.} }
21: {\ \rule{0.5em}{0.5em}}
22: %\input{tcilatex}
23: \begin{document}
24:
25:
26: \begin{center}
27: \textbf{\large Light trapping and guidance in plasmonic nanocrystals}\\*[%
28: 0.1cm]
29: \textbf{Maxim Sukharev and Tamar Seideman}\\*[0.2cm]
30:
31: \textit{Department of Chemistry, Northwestern University,\\[0pt]
32: 2145 Sheridan Road, Evanston, IL. 60208-3113 USA}\\[0pt]
33:
34: \textbf{ABSTRACT}
35: \end{center}
36:
37: \vspace*{-0.3cm}{\small \ We illustrate the possibility of light trapping
38: and funneling in periodic arrays of metallic nanoparticles. A controllable
39: minimum in the transmission spectra of such constructs arises from a
40: collective plasmon resonance phenomenon, where an incident plane wave
41: sharply localizes in the vertical direction, remaining delocalized in the
42: direction parallel to the crystal plane. Using hybrid arrays of different
43: structures or different materials, we apply the trapping effect to structure
44: the eigen-mode spectrum, introduce overlapping resonances, and hence direct
45: the light in space in a wavelength-sensitive fashion.}\\*[0.3cm]
46:
47: \section{Introduction}
48:
49: \label{Introduction}
50:
51: Tremendous progress in fabrication methods of metallodielectric
52: nanostructures,\cite{Hutter2004} on the one hand, and rapid development of
53: laser technology,\cite{SlusherRevModPhys1999} on the other, have recently
54: enabled a wide variety of exciting opportunities for light manipulation in
55: subdiffraction length scales. These range from understanding of fundamental
56: concepts of plasmon-polariton dynamics,\cite{SPP-reviews} through coherent
57: and optimal control of electromagnetic energy propagation via metallic
58: structures,\cite{SPP-control} to different applications in modern optical
59: device technology,\cite{Atwater2001review} single atom-molecule
60: manipulations,\cite{VanDuyneScience2004} and ultrasensitive detection of
61: biological molecules.\cite{Alivisatos2004}
62:
63: The physical basis of these applications is the surface plasmon resonance
64: phenomenon, owing to which nobel metal particles of nanosize strongly
65: enhance electromagnetic (EM) fields tuned to resonance with the frequency of
66: collective oscillation of conductive electrons confined in the nanoparticle.
67: It follows from the Mie theory,\cite{Kreibig1995} (and was seen numerically
68: as well as experimentally) that the resonance wavelength depends crucially
69: on the particle size and shape. This sensitivity is the origin of many of
70: the applications of nobel metal nanoparticles, including their use as
71: sensors,\cite{Rindzevicius2005,Lu2006} medical diagnostics,\cite%
72: {El-Sayed2005} and standards.\cite{Reinhard2005}
73:
74: Like the response to incident light of single nanoparticles (NPs), the
75: collective response of NP chains has been the topic of significant
76: experimental\cite{NP-chains-experiments} and theoretical interest in recent
77: years.\cite{NP-chains-theory} Particularly interesting is the possibility of
78: guiding light in the nanoscale via metallic NP chains and junctions.\cite%
79: {Bowden1989} Here, one excites the construct locally in space at one end of
80: the chain, typically by means of a nanosized tip, and detects the
81: transmitted energy, again with spatial resolution, at the other end, e.g.,
82: by means of a dye molecule. Much progress on this question not withstanding,
83: the adverse effect of losses have been identified as the major hurdle to
84: realizing waveguides based on NP chains.\cite{SPP-reviews} Films\cite%
85: {Films-Refs} and wires\cite{Wires-Refs} suffer much less from losses, but
86: lack the local enhancement and mode specificity of NPs.
87:
88: Similarly fascinating, are the properties and applications of plasmonic
89: nanocrystals. By nanocrystals, we refer to ordered two- or three-dimensional
90: arrays of nanoparticles, periodic in the lateral direction, that are
91: collectively excited by an incident (delocalized) plane wave.
92: Metallodielectric crystals exhibit omnidirectional band gaps\cite%
93: {JoannopoulosPRB1996} along with a broadband absorption.\cite{Fan2005}
94: Experimental measurements of optical properties of \ one-\cite%
95: {GiessenPRL2003} and two-dimensional\cite{FriendNanoLett2006} periodic metal
96: nanostructures have been successfully performed. We also note that, not only
97: periodic structures composed of NPs exhibit interesting phenomena but also
98: periodic arrays of sub-wavelength holes have surprising optical transmission.%
99: \cite{LezecNature1998} The physics underlying such constructs has aspects in
100: common with that underlying finite NP chains and junctions, it relies on the
101: collective response of ordered NPs that are sufficiently closely-spaced to
102: interact via dipole and quadruple forces. The new features arise from the
103: periodicity in one direction, which gives rise to certain solid state
104: features, such as a band structure, and justifies the name \textquotedblleft
105: crystals\textquotedblright . The practical interest in nanoplasmonic
106: crystals steams from the premise of scaling to the nano-domain several of
107: the attractive properties of micron-sized analogs, such as photonic crystals
108: and band-gap materials, which have been extensively studied experimentally
109: and theoretically for their many applications.\cite{Soukoulis-review2002} Of
110: particular interest are properties such as light-induced transparency,\cite%
111: {He2006} self-collimation,\cite{Lu2006} and transmission through sharp bends.%
112: \cite{Mekis1996} While the essential physics responsible for these features
113: in the macro-world is quite different from that underlying the nanoscale
114: variant, the effects, as illustrated below, are similar.
115:
116: We show in the following sections that plasmonic nanocrystals offer also new
117: and fundamentally interesting phenomena, that are unique to the nanosize and
118: do not have a direct analog in their micron-sized counterparts. In
119: particular, we illustrate the possibilities of introducing overlapping
120: resonances in the eigen-mode spectrum as a control tool, and of funneling
121: the incident plane wave in a predetermined direction with wavelength
122: selectivity. In the next section we briefly outline the theory and the
123: numerical approach taken. Section \ref{Results} presents and discusses our
124: results and the final section concludes, with an outlook to future research.
125:
126: \section{Theory}
127:
128: \label{Theory}
129:
130: The interaction between the light and the metal nanoconstructs is simulated
131: using a finite-difference time-domain approach.\cite{Taflove2000} We
132: restrict attention to two-dimensional structures, and consider the TE$_{z}$
133: mode (transverse-electric mode with respect to $z$) of the EM field.
134: Maxwell's equations in this case read,
135: \begin{gather}
136: \varepsilon _{eff}\frac{\partial E_{x}}{\partial t}=\frac{\partial H_{z}}{%
137: \partial y}-J_{x}, \notag \\
138: \varepsilon _{eff}\frac{\partial E_{y}}{\partial t}=-\frac{\partial H_{z}}{%
139: \partial x}-J_{y}, \label{Maxwell equations} \\
140: \mu _{0}\frac{\partial H_{z}}{\partial t}=\frac{\partial E_{x}}{\partial y}-%
141: \frac{\partial E_{y}}{\partial x}, \notag
142: \end{gather}%
143: where $E_{x}$, $E_{y}$, and $H_{z}$ denote the Cartesian components of the
144: EM field, $J_{x}$ and $J_{y}$ are the Cartesian components of the current
145: density, $\vec{J}$, $\varepsilon _{eff}$ is the effective dielectric
146: constant defined below, and $\mu _{0}$ is the magnetic permeability of free
147: space. The metal structures are described within the Drude model with a
148: complex valued, frequency-dependent dielectric constant,\cite{Huffman1983}
149: \begin{equation}
150: \varepsilon \left( \omega \right) =\varepsilon _{0}\left( \varepsilon
151: _{\infty }-\frac{\omega _{p}^{2}}{\omega ^{2}+i\Gamma \omega }\right) ,
152: \label{Drude model}
153: \end{equation}%
154: where $\varepsilon _{0}$ is the electric permittivity of free space, $%
155: \varepsilon _{\infty }$ is the dimensionless infinite frequency limit of the
156: dielectric constant, $\omega _{p}$ is the bulk plasmon frequency, and $%
157: \Gamma $ is the damping rate. In the numerical simulations we use the
158: following two sets of parameters, which describe fairly well silver and gold
159: nanoparticles in the incident wavelength regime of interest. For silver
160: nanoconstructs, $\varepsilon _{\infty }=8.926$, $\omega _{p}=1.7601\times
161: 10^{16}$ rad/sec, and $\Gamma =3.0841\times 10^{14}$ rad/sec. For gold
162: nanoconstructs, $\varepsilon _{\infty }=9.84$, $\omega _{p}=1.3673\times
163: 10^{16}$ rad/sec, and $\Gamma =1.0179\times 10^{14}$ rad/sec.
164:
165: In regions of space occupied by metallic particles, the material dispersion
166: gives rise to time-dependent current and Maxwell's equations (\ref{Maxwell
167: equations}) are supplemented by the additional set of equations for the
168: current density $\vec{J}$,\cite{Ziolkowski1995,GrayPRB2003}
169: \begin{equation}
170: \frac{\partial \vec{J}}{\partial t}=a\vec{J}+b\vec{E}, \label{current}
171: \end{equation}%
172: where $a=-\Gamma $, $b=\varepsilon _{0}\omega _{p}^{2}$ and the effective
173: dielectric constant in Eq. (\ref{Maxwell equations}) is $\varepsilon
174: _{eff}=\varepsilon _{0}\varepsilon _{\infty }$ . In the surrounding free
175: space $\varepsilon _{eff}=\varepsilon _{0}$ and $a=b=0$. In order to avoid
176: nonphysical reflections of outgoing electromagnetic waves from the grid
177: boundaries, we implement perfectly matched layers (PML) of absorbing
178: boundaries\cite{BerengerPML} of depth of $16$ spatial steps each.
179:
180: As an initial condition we employ a plane wave generated along a line, each
181: point of which is driven by a time-dependent function of the form,
182: \begin{equation}
183: \vec{E}_{\text{inc}}\left( t\right) =\vec{e}_{x}E\left( t\right) \cos \omega
184: t, \label{incident field}
185: \end{equation}%
186: where $\vec{e}_{x}$ is a unit vector along the $x$-axis, $E\left( t\right)
187: =E_{0}\sin ^{2}\left( \pi t/\tau \right) $ is the pulse envelope, $\tau $ is
188: the pulse duration ($E\left( t>\tau \right) =0$), and $\omega $ is the
189: optical frequency. In all simulations $\tau =50$ fs, the total propagation
190: time is $100$\ fs, with a spatial step size of $\delta x=\delta y=1.5$
191: \thinspace nm and a temporal step size of $\delta t=\delta x/\left(
192: 1.5c\right) $, where $c$ denotes the speed of light in vacuum. All
193: simulations have been performed on distributed memory parallel computers at
194: the National Energy Research Scientific Computing Center and San Diego
195: Supercomputer Center. The parallel technique used in our simulations is
196: described in detail in Ref. 31.
197:
198: \section{Results and Discussion}
199:
200: \label{Results}
201:
202: We open this section by exploring the properties of the simplest plasmonic
203: crystal that could be envisioned. This construct will serve to introduce the
204: basic phenomena that underlie several of the much more complicated
205: constructs discussed below. The nano-crystal envisioned is schematically
206: depicted in the inset of Fig. 1. It is composed of a periodic array of
207: silver NPs with $55$ nm diameter and $66$ nm center-to-center distance. The
208: NPs are excited by an $x$-polarized plane wave propagating in the negative $%
209: y $-direction, and the EM energy is detected along the horizontal line shown
210: in dashed in the inset. The transmission through the nanostructures is
211: expressed in terms of the ratio of the time-averaged EM energy in the
212: presence of the periodic array, $W_{\text{total}}$, to that in its absence, $%
213: W_{\text{inc}}$. This ratio is shown in the main frame of Fig. 1 for five
214: arrays with different number of layers as a function of the incident
215: wavelength. The transmission is close to perfect throughout the wavelength
216: regime of relevance, safe for a narrow window centered at $\lambda \simeq
217: 360 $ nm, where it drops to a sharp minimum. The transmission minimum is
218: sharper the larger the number of layers, converging to essentially zero
219: transmission ($W_{\text{total}}/W_{\text{inc}}\approx 2\times 10^{-7}$) with
220: $5$ layers.
221:
222: The origin of the transmission minimum is the excitation of longitudinal ($x$%
223: -polarized) plasmons, due to which an incident light propagating along the $%
224: y $-axis bends as it enters the periodic array and is guided along the
225: horizontal axis. The efficiency with which the light takes the sharp corner,
226: and the spatial localization of its subsequent path are investigated below.
227: Our results provide a likely explanation to the finding of Ref. 19, where a
228: strong broadband absorption in metallic photonic crystals is reported. As
229: shown below, the resonance feature is due to an interesting trapping
230: phenomenon, whereby the electromagnetic energy is localized along the
231: surface of the crystal. Such collective waves that are fully delocalized
232: laterally but sharply localized in the vertical direction are familiar in
233: electronic systems, where they arise from image potentials.\cite%
234: {image-potentials} We are not aware of previous studies that have observed
235: such phenomena in plasmonic waves.
236:
237: Both the broad transmission window and the ability of the crystal to
238: re-direct the propagating EM energy by $90$ degrees and guide it in parallel
239: to the crystal plane at the transmission minimum wavelength are of
240: significant practical interest, as they address two long researched goals of
241: nanophotonics.\cite{SPP-reviews} More interestingly from a fundamental
242: perspective, the resonance phenomenon that underlies the effect offers
243: exciting opportunities for optical control. Such resonance features have
244: been intensively studied in atomic and molecular physics, where their
245: manifestation in radiative lineshapes is conveniently described within the
246: Feschbach partitioning of space into a bound and a scattering manifold. The
247: present system can be analyzed using the machinery of Fano and Feschbach by
248: analogy to the well-studied case of matter waves, but offers the advantage
249: over the latter case that the interaction coupling the discrete state to the
250: continuum can be systematically tuned by design of the array.
251:
252: We proceed by exploring the dependence of the resonance feature on the
253: experimentally variable parameters of the array. In Fig. 2 we focus on the
254: single layer array and gradually modify the resonance properties by coating
255: the silver nano-spheres by a non-dispersive insulator. The inset of Fig. 2
256: depicts schematically the array envisioned. We remark that core-shell
257: nanoparticles of the sort shown in the inset of Fig. 2 have been synthesized
258: in several recent studies that have noted the potential applications and
259: interesting properties of such constructs. Several trends in Fig. 2, and in
260: related studies that are not shown here, can be qualitatively understood
261: from the Mie expression for the extinction cross section of a single sphere.
262: Both an increase in the dielectric constant of the core material and an
263: increase in the dielectric constant of the shell material red-shift the
264: resonance position. An increase in the shell dielectric constant has the
265: additional effect of broadening the resonance, see Fig. 2.
266:
267: Of particular interest are devices that support multiple resonances. In
268: molecular systems, the case of overlapping resonances has been studied
269: extensively for its fundamental interest as well as for intriguing
270: opportunities for coherent control.\cite{OverlapRes} In the nanoplasmonics
271: considered here, both the case of isolated and that of overlapping multiple
272: resonances are of fundamental and practical interest, and both can be
273: readily realized by proper design of the nano-crystal. Figure 3 explores
274: both the case of non-overlapping and that of overlapping resonances. To that
275: end we consider hybrid nanocrystals consisting of ordered layers of two
276: types of metallic nanoparticles and include both spheres and ellipsoids. The
277: insets depict the two nano-crystals considered, one of which (inset a)
278: combines silver and gold nano-spheres and the other (inset b) considers
279: alternating layers of spheres and ellipsoids, the top two of which are
280: silver, and the bottom two gold. The combination of Au and Ag NPs provides a
281: model for study of non-overlapping resonances, since the plasmon resonance
282: wavelengths of Ag and Au are separated by ca $120$ nm. The combination of
283: spheres and ellipsoids provides a model for study of overlapping resonances,
284: since the shape-dependence of the resonance wavelength is small, leading to
285: a shift of the line center comparable to the resonance width.
286:
287: The main frame of Fig. 3 shows simulations of the ratio, $W_{\text{total}%
288: }/W_{\text{inc}}$, for the two hybrid constructs depicted in the insets,
289: with the solid curve corresponding to the bimetallic nano-crystal of inset
290: (a) and the dashed curve corresponding to the mixed crystal of inset (b).
291: Both spectra show the expected well-defined double-peak structure. The
292: spectrum generated by construct (b) exhibits also significant broadening and
293: additional structure, arising from the partially overlapping resonances of
294: the spheres and ellipsoids.
295:
296: The results of Fig. 3 can be understood through an eigenmode analysis of the
297: elements constituting the nanocrystals. The eigenmodes of a given construct
298: are computed by excitation with a broad band, short ($0.36$ fs) pulse,
299: followed by long time ($2$ ps) simulation of the $x$-component of the
300: resultant scattered electric field. Fourier transform of the field yields
301: the power spectrum $I(\lambda )$. The solid and dashed curves in Fig. 4
302: correspond to the cases of single (periodic) layers of silver and gold
303: nano-spheres, respectively, whereas the dot-dashed curve corresponds to one
304: Ag and one Au nano-spheres layer and the dotted curve to one Ag spheres and
305: one Ag ellipsoids layer. Several interesting features are noted, including
306: Fano-type interference between direct and resonant transmission resulting in
307: marked asymmetry (dashed curve), non-overlapping resonances (dot-dashed
308: curve) and overlapping resonances (dotted curve).
309:
310: The plasmon resonances of silver and gold linear periodic chains are
311: separated by almost $120$ nm, as illustrated in Fig. 4, resulting not only
312: in selective transmission of light by the hybrid nano-structure but also in
313: separation of different resonant wavelengths in space. For example,
314: wavelengths near the plasmon resonance of silver chains, $\lambda \simeq 380$
315: nm, are trapped in the first two silver layers (see the insets in Fig. 3),
316: whereas wavelengths near the plasmon resonance of gold chains ($\lambda
317: \simeq 500$ nm) are transmitted through the silver layers and excite the
318: bottom two gold chains. The wavelength-dependent transparency of different
319: parts of a nano-structure is due to the fact that the effective thickness of
320: the skin-layer at the wavelengths considered is of the same order of
321: magnitude as the size of the NPs. The skin-depth ($d$) at optical
322: frequencies ($\omega $) much larger than the plasmon frequency ($\omega _{p}$%
323: ) can be estimated\cite{Greiner1998} as $d\simeq c/\omega _{p}$, which leads
324: to $17$ nm and $22$ nm for silver and gold, respectively.
325:
326: The wavelength-sensitive transparency of different parts of nanostructures
327: suggests exciting opportunities for use of non-symmetric (with respect to $y$%
328: ) constructs for light manipulation and directed transport. Two examples of
329: light guidance are presented in Fig. 5, where we simulate hybrid structures
330: composed of silver and gold NPs that are finite in both $x$- and $y$%
331: -directions. At $\lambda =340$ nm, for instance, the left part of the
332: nano-crystal depicted in the inset (a) of Fig. 5 traps the light, while the
333: right part, composed of gold NPs is transparent. Excited at their plasmon
334: resonance wavelength, silver layers guide the EM energy to the right. The
335: structure shown in inset (b) of Fig. 5, supports both horizontal directions
336: for EM energy guiding. This is illustrated in Fig. 6, where we plot the
337: spatial time-averaged distribution of the ratio, $W_{\text{total}}/W_{\text{%
338: inc}}$, at the incident wavelength corresponding to the plasmon resonance of
339: silver NPs. Clearly, an incident plane wave, initially symmetric, is broken
340: by the hybrid structure into two parts, one of which is guided to the left,
341: and the other to the right.
342:
343: In summary, we illustrated selective trapping and guidance of light by
344: periodic metallic nanocrystals. Owing to the excitation of collective
345: longitudinal plasmons in linear chains of closely spaced NPs, an incident
346: wave can be selectively localized in the direction vertical to the light
347: propagation, remaining delocalized laterally. By proper device design, EM
348: excitation can be thus efficiently guided in space. A variety of extensions
349: of the phenomena of selective light trapping and funneling can be
350: envisioned. Three dimensional nanocrystals could be used to guide EM energy
351: in 3D space. The interaction of white light with periodic arrays could serve
352: to make interesting optical sources. Likewise inviting is the possibility of
353: guiding light in circular or helical pathways using properly structured
354: plasmonic nanoconstructs.
355:
356: Acknowledgement.\textbf{\ }This work was supported in part by the NCLT
357: program of the National Science Foundation (ESI-0426328) at the Material
358: Research Institute of Northwestern University. We acknowledge the National
359: Energy Research Scientific Computing Center, supported by the Office of
360: Science of the U.S. Department of Energy under Contract No.
361: DE-AC03-76SF00098, and the San Diego Supercomputer Center under Grant No.
362: PHY050001, for computational resources. We thank Joseph Yelk for
363: computational assistance.
364:
365: \begin{thebibliography}{99}
366: \bibitem{Hutter2004} For a review of modern fabrication methods see: Hutter,
367: E.; Fendler, J. H. \textit{Adv. Mat.} \textbf{2004},\textit{\ 16}, 1685.
368:
369: \bibitem{SlusherRevModPhys1999} Slusher, R. E. \textit{Rev. Mod. Phys.}
370: \textbf{1999}, \textit{71}, 00S471.
371:
372: \bibitem{SPP-reviews} (a) Barnes, W. L.; Dereux, A.; Ebbesen, T. W. \textit{%
373: Nature} \textbf{2003}, \textit{424}, 824. (b) Maier,\ S. A.; Atwater, H. A.
374: \textit{J. Appl. Phys.} \textbf{2005}, \textit{98}, 011101. (c) Ozbay, E.
375: \textit{Science}\textbf{\ 2006}, \textit{311}, 189.
376:
377: \bibitem{SPP-control} (a) Stockman, M. I.; Faleev, S. V.; Bergman, D. J.
378: \textit{Phys. Rev. Lett.} \textbf{2002}, \textit{88}, 067402. (b) Lee,
379: T.-W.; Gray, S. K. \textit{Phys. Rev. B} \textbf{2005}, \textit{71}, 035423.
380: (c) Brixner, T.; de Abajo, F. J. G.; Schneider, J.; Pfeiffer, W. \textit{%
381: Phys. Rev. Lett.} \textbf{2005}, \textit{95}, 093901. (d) Sukharev, M.;
382: Seideman, T. \textit{NanoLett}. \textbf{2006}, \textit{6}, 715.
383:
384: \bibitem{Atwater2001review} Maier, S. A.; Brongersma, M. L.; Kik, P. G.;
385: Meltzer, S.; Requicha, A. A. G.; Atwater, H. A.; \textit{Adv. Mat.}\textbf{\
386: 2001}, \textit{13}, 1501.
387:
388: \bibitem{VanDuyneScience2004} Van Duyne, R. P. \textit{Science} \textbf{2004}%
389: , \textit{306}, 985.
390:
391: \bibitem{Alivisatos2004} Alivisatos, P. \textit{Nat. Biotechn.} \textbf{2004}%
392: , \textit{22}, 47.
393:
394: \bibitem{Kreibig1995} Kreibeg, U.; Vollmer, M. \textit{Optical Properties of
395: Metal Clusters}; Springer: Berlin, 1995.
396:
397: \bibitem{Rindzevicius2005} Rindzevicius, T.; Alaverdyan, Y.; Dahlin, A.; H%
398: \"{o}\"{o}k, F.; Sutherland, D. S.; K\"{a}ll, M. \textit{NanoLett}. \textbf{%
399: 2005}, \textit{5}, 2335.
400:
401: \bibitem{Lu2006} Lu, Z.; Shi, S.; Murakowski, J. A.; Schneider, G. J.;
402: Schuetz, C. A.; Prather, D. W. \textit{Phys. Rev. Lett. }\textbf{2006}
403: \textit{96}, 173902.
404:
405: \bibitem{El-Sayed2005} El-Sayed, I. H.; Huang, X.; El-Sayed, M. A. \textit{%
406: NanoLett}. \textbf{2005}, \textit{5}, 829.
407:
408: \bibitem{Reinhard2005} Reinhard, B. M.; Siu, M.; Agarwal, H.; Alivisatos, P.
409: A.; Liphardt, J. \textit{NanoLett}. \textbf{2005}, \textit{5}, 2246.
410:
411: \bibitem{NP-chains-experiments} (a) Krenn, J. R.; Dereux, A.; Weeber, J. C.;
412: Bourillot, E.; Lacroute, Y.; Goudonnet, J. P; Schider, G.; Gotschy, W.;
413: Leitner, A.; Aussenegg, F. R.; Girard, C. \textit{Phys. Rev. Lett. }\textbf{%
414: 1999}, \textit{82}, 2590 (b) Dickson, R. M.; Lyon, L. A. \textit{J. Phys.
415: Chem. B} \textbf{2000}, \textit{104}, 6095. (c) Haynes, C. L.; Van Duyne, R.
416: P. \textit{J. Phys. Chem. B} \textbf{2001}, \textit{105}, 5599. (d) Haynes,
417: C. L.; Van Duyne, R. P. \textit{Mater. Res. Soc. Symp. Proc.} \textbf{2002},
418: \textit{728}, S10.7.1. (e) Haynes, C. L.; McFarland, A. D.; Zhao, L.; Van
419: Duyne, R. P.; Schatz, G. C.; Gunnarsson, L.; Prikulis, J.; Kasemo, B.; Kall,
420: M. \textit{J. Phys. Chem. B} \textbf{2003}, \textit{107}, 7337 (f) Maier, S.
421: A.; Kik, P. G.; Atwater, H. A.; Meltzer, S.; Harel, E.; Koel, B. A.;
422: Requicha, A. A. G. \textit{Nat. Mat.} \textbf{2003}, \textit{2}, 229. (g)
423: Quidant, R.; Leveque, G.; Weeber, J.-C.; Dereux, A.; Girard, C.; Weiner, J.
424: \textit{Europhys. Lett.} \textbf{2004}, \textit{66}, 785.
425:
426: \bibitem{NP-chains-theory} (a) Quinten, R.; Leitner, A.; Krenn, J. R.;
427: Aussenegg, F. R. \textit{Opt. Lett.}\textbf{\ 1998}, \textit{23}, 1331. (b)
428: Rechberger, W.; Hohenau, A.; Leitner, A.; Krenn, J. R.; Lamprecht, B.;
429: Aussenegg, F. R. \textit{Opt. Comm.}\textbf{\ 2003}, \textit{220}, 137. (c)
430: Quidant, R.; Girard, C.; Weeber, J.-C.; Dereux, A. \textit{Phys. Rev. B}%
431: \textbf{\ 2004}, \textit{69}, 085407.
432:
433: \bibitem{Bowden1989} Haus, J. W.; Kalyaniwalla, N.; Inguva, R.; Bowden, C.
434: M. \textit{J. Appl. Phys.} \textbf{1989}, \textit{65}, 1420.
435:
436: \bibitem{Films-Refs} (a) Zia, R.; Selker, M. D.; Catrysse, P. B.;
437: Brongersma, M. L. \textit{J. Opt. Soc. Am. B}\textbf{\ 2004}, \textit{21},
438: 2442. (b) Rokitski, R.; Tetz, K. A.; Fainman, Y. \textit{Phys. Rev. Lett. }%
439: \textbf{2005}, \textit{95}, 177401.
440:
441: \bibitem{Wires-Refs} (a) Dickson, R. M.; Lyon, L. A. \textit{J. Phys. Chem. B%
442: } \textbf{2000}, \textit{104}, 6095. (b) Kottmann, J. P.; Martin, O. J. F.
443: \textit{Opt. Expr.} \textbf{2001}, \textit{8}, 655. (c) M\"{u}ller, T.;
444: Heinig, K.-H.; Schmidt, B. \textit{Mat. Sci. Eng. C}\textbf{\ 2002}, \textit{%
445: 19}, 209. (d) Wang, K.; Mittleman, D. M. \textit{J. Opt. Soc. Am. B}\textbf{%
446: \ 2005}, \textit{22}, 2001. (e) Imura, K.; Nagahara, T.; Okamoto, H. \textit{%
447: J. Chem. Phys. }\textbf{2005}, \textit{122}, 154701.
448:
449: \bibitem{JoannopoulosPRB1996} Fan, S.; Villeneuve, P. R.; Joannopoulos, J.
450: D. \textit{Phys. Rev. B} \textbf{1996}, \textit{54}, 11245.
451:
452: \bibitem{Fan2005} Veronis, G.; Dutton, R. W.; Fan, S. \textit{J. App. Phys.}
453: \textbf{2005}, \textit{97}, 093104.
454:
455: \bibitem{GiessenPRL2003} Christ, A.; Tikhodeev, S. G.; Gippius, N. A.; Kuhl,
456: J.; Giessen, H. \textit{Phys. Rev. Lett.} \textbf{2003}, \textit{91}, 183901.
457:
458: \bibitem{FriendNanoLett2006} Zhang, X.; Sun, B.; Friend, R. H. \textit{%
459: NanoLett. }\textbf{2006}, \textit{6}, 651.
460:
461: \bibitem{LezecNature1998} Ebbessen, T. W.; Lezec, H. J.; Ghaemi, H. F.;
462: Thio, T.; Wolf, P. A. \textit{Nature }\textbf{1998}, \textit{391}, 667.
463:
464: \bibitem{Soukoulis-review2002} Soukoulis, C. M. \textit{Nanotechnology }%
465: \textbf{2002}, \textit{13}, 420.
466:
467: \bibitem{He2006} He, Q.-Y.; Xue, Y.; Artoni, M.; La Rocca, G. C.; Xu, J.-H.;
468: Gao, J.-Y. \textit{Phys. Rev. B}\textbf{\ 2006}, \textit{73}, 195124.
469:
470: \bibitem{Mekis1996} Mekis, A.; Chen, J. C.; Kurland, I.; Fan, S.;
471: Villeneuve, P. R.; Joannopoulos, J. D. \textit{Phys. Rev. Lett. }\textbf{1996%
472: }, \textit{77}, 3778.
473:
474: \bibitem{Taflove2000} Taflove, A.; Hagness, S. C. \textit{Computational
475: Electrodynamics: The Finite-Difference Time-Domain Method}, 2nd ed.; Artech
476: House: Boston, 2000.
477:
478: \bibitem{Huffman1983} Bohren, C. F.; Huffman, D. R. \textit{Absorption and
479: Scattering of Light by Small Particles}, Wiley: New York, 1983.
480:
481: \bibitem{Ziolkowski1995} Ziolkowski, R. W.; Judkins, J. B. \textit{J. Opt.
482: Soc. Am. A} \textbf{1995}, \textit{12}, 1974.
483:
484: \bibitem{GrayPRB2003} Gray, S. K.; Kupka, T. \textit{Phys. Rev. B} \textbf{%
485: 2003}, \textit{68}, 045415.
486:
487: \bibitem{BerengerPML} Berenger, J.-P. \textit{J. Comput. Phys.} \textbf{1994}%
488: , \textit{114}, 185.
489:
490: \bibitem{SukharevJCP2006} Sukharev, M.; Seideman, T. \textit{J. Phys. Chem. B%
491: } \textbf{2006}, \textit{124}, 144707.
492:
493: \bibitem{image-potentials} (a) Forstmann, F. \textit{Progr. Surf. Science }%
494: \textbf{1993}, \textit{42}, 21. (b) Nekovee, M.; Inglesfield, J. E. \textit{%
495: Progr. Surf. Science }\textbf{1995}, \textit{50}, 149. (c) Machado, M.;
496: Chulkov, E. V.; Silkin, V. M.; Hofer, U.; Echenique, P.M. \textit{Progr.
497: Surf. Science }\textbf{2003}, \textit{74}, 219.
498:
499: \bibitem{OverlapRes} Shapiro, M.; Brumer, P. \textit{Principles of the
500: Quantum Control of Molecular Processes}, Hoboken, N.J.: Wiley-Interscience,
501: 2003.
502:
503: \bibitem{Greiner1998} Greiner, W. \textit{Classical Electrodynamics},
504: Springer-Verlag: New York, 1998.
505: \end{thebibliography}
506:
507: \newpage
508:
509: \begin{center}
510: \bigskip {\Large Figures}
511: \end{center}
512:
513: \begin{figure}[htbp]
514: \centering\includegraphics[width=\linewidth]{my1}
515: \caption{\label{fig1}
516: The ratio of time-averaged EM energy calculated in the presence and
517: in the absence of the nanoconstruct schematically depicted in the inset as a
518: function of the incident wavelength, $\protect\lambda $. The solid curve
519: shows the ratio for a single horizontal layer of NPs, the dashed curve for
520: two layers, the dotted curve for three layers, the dash-dotted curve four
521: layers, and the dash-dot-dotted curve for five layers. The inset shows the
522: geometry envisioned, where the dashed vertical lines represent periodic
523: boundaries in the lateral direction. The nanoconstruct is excited by an $x$%
524: -polarized plane wave propagating in the negative $y$ direction and the EM
525: energy is calculated along the horizontal line shown as a dashed line
526: beneath the structure in the far-field zone (the distance of the detection
527: line from the crystal is $600$ nm).}
528: \end{figure}
529:
530: \newpage
531:
532: \begin{figure}[htbp]
533: \centering\includegraphics[width=\linewidth]{my2}
534: \caption{\label{fig2}
535: As in Fig. 1, for a single layer of silver NPs coated by an
536: insulator (see inset). The solid curve shows the ratio $W_{\text{total}}/W_{%
537: \text{inc}}$ for bare particles (no coating, corresponding to $\protect%
538: \varepsilon =1$), the dashed curve for coating dielectric constant of $%
539: \protect\varepsilon =1.5$, the dotted curve for $\protect\varepsilon =2.5$,
540: and the dash-dotted curve for $\protect\varepsilon =3.5$. The silver NP
541: diameter is $35$\ nm, the diameter of the dielectric coating layer is $55$\
542: nm, and the center-to-center distance is $66$\ nm.}
543: \end{figure}
544:
545: \newpage
546:
547: \begin{figure}[htbp]
548: \centering\includegraphics[width=\linewidth]{my3}
549: \caption{\label{fig3}
550: The ratio $W_{\text{total}}/W_{\text{inc}}$ as a function of the
551: incident wavelength for the two nanoconstructs shown schematically in the
552: insets. (a) A silver-gold hybrid structure composed of spheres of diameter $%
553: 32$\ nm and center-to-center distance $42$\ nm. (b) A silver-gold hybrid
554: structure composed of spheres and ellipsoids with major (minor) axes of $20$
555: ($32$) \ nm and center-to-center distance of $42$ nm. Gold NPs are shown as
556: black, silver particles as grey spheres. The solid curve in the main frame
557: corresponds to structure (a) and the dashed curve to structure (b).}
558: \end{figure}
559:
560: \newpage
561:
562: \begin{figure}[htbp]
563: \centering\includegraphics[width=\linewidth]{my4}
564: \caption{\label{fig4}
565: Power spectrum, $\left\vert I\left( \protect\lambda \right)
566: \right\vert $,\ of the $x$-component, $E_{x}$, of the scattered electric
567: field in Fig. 3 as a function of the incident wavelength, $\protect\lambda $%
568: . The solid curve corresponds to a single layer of silver NPs, the dashed
569: curve to a layer of gold NPs, the dash-dotted curve to interacting layers of
570: silver and gold NPs, and the dotted curve to silver layers of spheres and
571: ellipsoids.}
572: \end{figure}
573:
574: \newpage
575:
576: \begin{figure}[htbp]
577: \centering\includegraphics[width=\linewidth]{my5}
578: \caption{\label{fig5}
579: The ratio $W_{\text{total}}/W_{\text{inc}}$ as a function of the
580: incident wavelength for the two finite nanoconstructs shown schematically in
581: the insets. Silver NPs are shown as grey spheres and gold NPs as black
582: spheres. The solid curve in the main frame presents simulations for
583: structure (a), and the dashed curve for structure (b). Both silver and gold
584: particles are $40$ nm in diameter and the center-to-center distance is $50$
585: nm.}
586: \end{figure}
587:
588: \newpage
589:
590: \begin{figure}[htbp]
591: \centering\includegraphics[width=\linewidth]{my6}
592: \caption{\label{fig6}
593: Time-averaged ratio, $W_{\text{total}}/W_{\text{inc}}$, as a
594: function of $x$ and $y$ for the structure shown in inset (b) of Fig. 5 at an
595: incident wavelength of $\protect\lambda =340$ nm, corresponding to the
596: excitation of longitudinal plasmons in silver NPs.}
597: \end{figure}
598:
599: \end{document}
600: