In nanostructures where confinement of a light wave is manifested, the conditions of strong light–matter interaction become feasible. In this case, the approach based on the environment-sensitive probability of quantum transitions is no longer applicable. Instead, joint states of light and matter should be considered and their time evolution explored. The content of the previous Chapters 13 and 14 can then be treated as a perturbative description of light–matter interactions with a limited range of applicability. In this chapter a few representative examples are given where joint light–matter states bring an unprecedented flavor of nanophotonic engineering to frozen excited states for quantum memory devices and single photon sources for quantum computing.

Cavity quantum electrodynamics in the strong coupling regime

The regime of strong coupling of a quantum system with an electromagnetic field can be performed in a cavity with very high Q-factor, where light can survive for a reasonable time until it either leaves the cavity or dies through being absorbed by cavity imperfections. In a simplified picture, a photon once emitted stays within the cavity so long that it is absorbed again by the emitter (an imaginary two-level system, an atom or a quantum dot). Thus spontaneous emission appears to become reversible. In a more accurate presentation, an “atom+field” state develops which evolves via oscillations between the state |E, 0〉 (“excited atom, depleted cavity”) to the state |G, 1〉 (“de-excited atom, a photon in the cavity”).

Nanostructures with characteristic surface relief of the order of 10–100 nm are known to modify the spatial distribution of an incident electromagnetic field. Local field enhancement results in enhanced absorption of photons by molecules or nanocrystals adsorbed at the surface. The effect is extremely pronounced in metal–dielectric structures because of surface plasmon resonance. A systematic application of field enhancement in Raman scattering enhancement and in photoluminescence enhancement with respect to molecular probes is followed nowadays by application of the effect with respect to nanocrystals (quantum dots) adsorbed at metal–dielectric nanotextured surfaces. It is the purpose of the present chapter to review mechanisms of photoluminescence enhancement and Raman scattering enhancement and factors in the context of their application to enhanced luminescence of molecules and quantum dots and Raman scattering. We consider not only local field enhancement in terms of the excitation process but also the photon density of states enhancement effect on photon emission processes with Raman scattering as a specific photon emission process. In this consideration, scattering of light experiences enhancement as does spontaneous emission. Therefore field enhancement and density of states effects should manifest themselves in the same manner in photoluminescence and scattering processes. Differences in scattering and luminescence enhancement are due to quenching processes which are crucial for luminescence and less pronounced for scattering.

Light and matter on a nanometer scale

The notion of “photonics” implies the science and technology related to generation, absorption, emission, harvesting, processing of light and their applications in various devices. Light is electromagnetic radiation available for direct human perception, in the wavelength range from approximately 400 to approximately 700 nanometers. Typically, adjacent far ultraviolet and near infrared ranges are also involved to give the approximate range of electromagnetic radiation from 100 nanometers to 1–2 micrometers as the subject of photonics. If the space has certain inhomogeneities on a similar scale to the wavelength of the light, then multiple scattering and interference phenomena arise modifying the propagation of light waves. Light scattering is the necessary prerequisite for vision. Shining colors in soap bubbles and thin films of gasoline on a wet road after rain are primary experiences of light-wave interference everybody gains in early childhood. To modifiy the conditions for light propagation, inhomogeneities in space which are not negligible as compared to the wavelength of the light, i.e. starting from the size range 10–100 nm to a few micrometers, become important. Space inhomogeneity for light waves implies inhomogeneity in dielectric permittivity.

Matter is formed from atoms which in turn can be subdivided into nuclei and electrons. An elementary atom of hydrogen has a radius for the first electron orbital of 0.053 nm. Atoms may form molecules and solids.

Is it possible for light to “jump lightly” through the looking glass as Alice did? Actually there is no absolutely reflecting border for electromagnetic waves provided that the material forming the reflecting border is restricted in space along the light propagation direction. A variety of light-through-the-looking-glass tunneling phenomena will be the subject of this chapter with the intriguing and challenging issue of superluminal light propagation in tunneling events, as well as with parallel analogies to quantum mechanical counterparts in nanoelectronics. Before reading this chapter, it is advisable to read Sections 3.4 and 3.5 in Chapter 3 for an introduction to tunneling effects in quantum mechanics and optics, as well as Section 7.11 of Chapter 7 where the problem of speed of light evaluation in complex structures has been addressed.

Tunneling of light: getting through the looking glass

Probably, every known physical case of very high reflectivity of light at some material border or interface brings about an evanescent field which penetrates forward through the border or interface under consideration.

Generally speaking, fields and matter are the two entities which constitute the Universe. These entities continuously interact. The electromagnetic field is the specific type of field which contains the range of oscillation frequencies which human eyes are able to sense. After twelve chapters in this book, we are now approaching the point where photons enter nanophotonics. Photons are necessary to understand how matter emits light. This happens by means of quantum transitions where matter loses and the electromagnetic field gains a certain portion of energy and momentum. The emission of light, in a broad sense, includes all types of processes where the electromagnetic field gains a portion of energy and momentum from matter. This can be classified as different types of secondary radiation which include emission of photons and scattering of photons. In nanophotonics, these elementary processes of field–matter interaction experience modification because of light-wave confinement, which is typically explained in a rather elegant way as a consequence of the photon density of states modification. The main purpose of the present chapter is to explain the notions of field quantization, photons, emission and scattering rates in terms of quantum transitions and density of electromagnetic modes.

Nanoplasmonics is a modern extensively expanding field of optics. It basically develops from the rather old field of metal optics including properties of dielectrics with subwavelength nanometer-size metal inclusions. For this reason, reading of Chapter 6 is strongly recommended prior to proceeding to the content of this chapter.

This chapter is organized as follows. First, local electromagnetic field enhancement of near metal singularities in a dielectric environment is considered. Further, multiple scattering phenomena are discussed including extraordinary transmission of a hole array in a metal film. Then, spatially organized metal–dielectric structures are discussed and for that section to be well understood, Chapter 7 will be helpful as will Chapter 10, where the tunneling of light in metal structures has been treated. Possible effects of metal nanoparticles on laser action and gain media properties are discussed as well. A separate section deals with general properties of negative refraction materials. This is included in this chapter since experimental realization of negative refraction in optics is based on periodic metal–dielectric metamaterials with specific subwavelength-scale architecture. Negative-refraction optics (and electrodynamics) emerged just a few years ago and the introductory style of this issue in the present chapter is believed will assist the interested reader in further self-learning through browsing in the many original papers on that matter.

As with all previous chapters in this book, the consideration in this chapter is organized rather like a guide into the field with emphasized physical phenomena than instructions for calculations and for fabrication of nanoplasmonic structures.

The complex propagation of light in periodic and inhomogeneous media considered in previous chapters forms a solid basis for purposeful controlling of light wave propagation and light energy accumulation. Coupling with light sources and light modulators gives rise to the well-defined concept of photonic circuitry. This recently emerged field is becoming more and more mature. The principal solutions in photonic circuitry are overviewed in this chapter. The discussion is kept at an introductory level to provide conceptual ideas and principal approaches without going too deeply into detail. The extensive list of references will partly compensate for the somewhat sketchy style in this chapter.

Microcavities and microlasers

In a sense, an optical microcavity, or a microresonator, can be treated as a wavelength-scale topological construction capable of accumulating and storing light. This can be implemented with respect to light impinging from the outside as well as with respect to light generated inside the cavity under consideration. The word “generated” here implies spontaneous emission, spontaneous Raman scattering rather than necessarily lasing. Light energy accumulation and storage becomes possible by the spatial confinement of light waves in a cavity. Primary examples of (micro)cavities and (micro)resonators were treated in Chapter 3 (Section 3.4) when the resonant tunneling of electromagnetic waves was analyzed; namely a one-dimensional problem of an electromagnetic wave impinging onto a pair of parallel metallic thin film layers serving as mirrors with dielectric spacing (Fig. 3.19).

Introduction

Continua generated using high pulse energy laser systems to create broad spectra (Alfano and Shapiro, 1970) have been used for spectroscopy for many years (e.g. Busch et al., 1973). The application of fibre-generated continua to spectroscopy was suggested as early as 1976 in work by Lin and Stolen (1976) where a continuum spanning ∼450–600 nm was generated in a step-index fibre pumped by a nitrogen pumped dye laser. Since the demonstration of supercontinuum generation in microstructured optical fibres (MOF), however, the range of spectroscopic and imaging applications has increased enormously, owing to the high average powers, unprecedented spectral width and relatively low cost and low complexity of such sources. This chapter specifically focuses on the applications of supercontinua generated in MOFs and, in particular, on applications in biophotonics.

MOF supercontinuum sources can be broadly grouped into three categories according to whether the laser pump source emits femtosecond pulses, picosecond–nanosecond pulses or cw radiation. In general terms, sub-ps pulses can produce broad supercontinua spanning from the UV to the NIR but the peak intensity damage threshold at the input end of the microstructured optical fibre limits the maximum average power that can be obtained in the supercontinuum to ∼<0.5 W with typically ∼<0.5 mW/nm available in the visible spectrum. The use of pump lasers with longer ps–ns pulses can significantly increase the maximum available average power before the onset of damage in the MOF such that high power supercontinua with several mW/nm in the visible can be achieved (e.g. Rulkov et al., 2005).

Motivation

Chapter 1 of this book and a number of previous publications (see, e.g., Russell, 2003; Dudley et al., 2006; Smirnov et al., 2006; Knight and Skryabin, 2007; Skryabin and Wadsworth, 2009) give excellent bibliographic and historical accounts of the supercontinuum and other nonlinear effects observed in photonic crystal fibres since, and to some extent before, the seminal results by Ranka et al. (Ranka et al., 2000) appeared in 1999–2000. Here we present a focused account of our understanding of the fibre supercontinuum based on the theory developed by the Bath group over recent years (Skryabin et al., 2003; Yulin et al., 2004; Skryabin and Yulin, 2005; Gorbach et al., 2006; Gorbach and Skryabin, 2007a,b,c) and on the experiments carried out in Los Alamos and Bath (Skryabin et al., 2003; Efimov et al., 2004, 2005, 2006; Gorbach et al., 2006; Stone and Knight, 2008), which have closely followed our theoretical work. Concepts and results systematically described below are centred around the problem of the frequency conversion due to interaction between solitons and dispersive waves (Skryabin and Yulin, 2005). In the end our approach leads to a qualitative understanding and quantitative description of the expansion of the femtosecond supercontinua (Gorbach and Skryabin, 2007b).

We start this chapter by introducing the deterministic model of supercontinuum generation, discuss its limitations and move on to the soliton self-frequency shift problem, presenting it in the way trimmed for our goals.

At the time of writing, it is just over 10 years since a seemingly simple research project at Bell Laboratories led to a new revolution in optical frequency metrology, when in 1999, in a packed CLEO postdeadline session, the first public presentation of supercontinuum generation in photonic crystal (or microstructure) fibers was given (Ranka et al. 1999).

This work had started a year earlier as a curiosity I had had as a post-doc at Bell Laboratories on the possibilities of extreme nonlinear interactions in small-core high-numerical aperture (NA) silica fibers that were fabricated with a transverse microstructure or photonic crystal fiber cladding. In all honesty, at the time I did not consider the possibility that the fiber structure would so substantially alter the dispersive and modal properties of the fiber so far from the zero dispersion wavelength of bulk silica. The success of this work is a tribute to the ability of a commercial research organization to understand and encourage basic research with a long-range outlook. Having a management team step back and allow a post-doc to work independently, with support, free from bureaucratic or technical interference, allowed the project to proceed unhindered.

After a survey of a number of different fibers that our optical fiber research group had fabricated and initial calculations of potential nonlinear effects that would occur with femtosecond duration pulses, a simple fiber design was chosen to start experimenting with.

Introduction

Studies of supercontinuum generation in photonic crystal fibers or dispersion-shifted highly nonlinear fibers have led to renewed interest in SC generation in other fiber types. For example, it was shown that fibers with a steadily-decreasing diameter in which the dispersive and nonlinear characteristics change as a function of propagation distance can lead to significant enhancement of the SC bandwidth and allow for an additional degree of control of the SC spectrum. In fact, it was the use of fibers with dispersion-varying profiles that motivated in the mid-1990s the pioneering work on SC generation intended for telecommunication applications [see e.g. Morioka et al., 1994a, 1994b, 1995, 1996; Mori et al., 1995]. Of course, it was already well known that efficient adiabatic pulse compression could be achieved in dispersion-varying fibers (Dianov et al., 1986), but these mid-1990s studies, anterior to the development of PCFs, were the first to specifically suggest that fibers with longitudinally-varying dispersion could be advantageous in generating broad and/or flat SC spectra (Lou et al., 1997; Mori et al., 1997). With the development of the theoretical understanding of SC generation process (see other chapters) renewed interest in dispersion-engineering has therefore occurred naturally. In this context, a key result that was reported is the demonstration of coherent SC generation in millimeter lengths of tapered fibers where the diameter of the fiber is controlled as a function of length to a great degree of accuracy (Lu and Knox, 2004).

Introduction

Over the past few years, new optical fibers with enhanced nonlinearity and tailored dispersion (Russell 2003; Knight 2003) have been providing a constantly growing platform for the development of advanced fiber-format devices and components for optical metrology (Jones et al. 2000; Udem et al. 2002), ultrashort-pulse laser technologies (Zheltikov 2007a), biomedicine (Hartl et al. 2001), quantum optics (Rarity et al. 2005), spectroscopy (Sidorov-Biryukov et al. 2006), and microscopy (Paulsen et al. 2003). Unique options offered by photonic-crystal fiber (PCF) technology (Russell 2006), such as dispersion management through fiber structure engineering (Reeves et al. 2003) and enhancement of optical nonlinearity due to a strong field confinement in a small-size fiber core (Fedotov et al. 2001), have been pushing the frontiers of fiber optics, allowing the creation of efficient sources of supercontinuum radiation (Ranka et al. 2000; Dudley et al. 2006; Zheltikov 2006), novel compact fiber lasers (Lim et al. 2002; Limpert et al. 2006), as well as frequency converters (Akimov et al. 2003), pulse compressors (Südmeyer et al. 2003), fiber components for biomedical optics (Flusberg et al. 2005a, 2005b), and optical sensors (Monro et al. 2001).

In the rapidly expanding field of nonlinear microscopy and spectroscopy, PCFs have been shown to possess a tremendous potential for making laser microscopes and spectrometers simpler and much more compact through the replacement of wavelength-tunable laser sources, such as optical parametric amplifiers and dye lasers, by a specifically designed segment of fiber.

Introduction

There are many applications for broad bandwidth infrared laser sources, including optical frequency metrology (Udem et al., 2002), precision spectroscopy (Holzwarth et al., 2000) and optical tomography (Hartl et al., 2001), and moving into the mid-infrared (mid-IR), uses for wavelengths beyond 2 μm include LIDAR, molecular spectroscopy and active hyperspectral imaging. Fibre-based supercontinuum sources are attractive for these applications due to their combination of high brightness and broad bandwidth in comparison to alternative thermal or laser sources. Current high brightness mid-IR sources are typically based on optical parametric oscillators (OPO) or quantum cascade lasers (QCL). While OPOs achieve excellent performance they require large pump lasers and can be rather complex and costly to maintain, and QCLs are hard to scale up in power and cannot at present be used to access the important 2–3 μm regime. New fibre-based technology could create an important additional source of robust and lower cost broad bandwidth mid-IR light for the future.

Beyond a wavelength of 2 μm, due to the onset of losses in silica, it is necessary to consider the use of non-silica glasses. The fundamental material properties of these glasses can enhance supercontinuum generation across the mid-IR since these glasses can have intrinsic nonlinearities ∼ 10 × to 100 × that of silica. However, the zero-dispersion wavelengths of these materials are generally longer than for silica, implying the need for longer wavelength pump lasers.

Introduction

Guided-wave nonlinear optics has attracted significant interest because of the unique environment that waveguides provide for nonlinear interactions, including tight confinement (high intensity), long interaction lengths (especially for fibres), control of propagation constants, and the possibility to incorporate them with integrated circuits (mainly for planar waveguides) [see (Lin et al., 2007; Knight and Skryabin, 2007; Foster et al., 2008; Afshar and Monro, 2009) and references therein]. Recent and rapid progress in design and manufacturing of complex structured microstructured optical fibres and planar waveguides with subwavelength features (including both subwavelength inclusions and voids) has further extended the opportunities to develop nonlinear devices by enabling extreme nonlinearity to combine with tailorable chromatic dispersion (Lin et al., 2007; Knight and Skryabin, 2007; Foster et al., 2008; Koos et al., 2007).

The nonlinear optical phenomena that occur in waveguides are determined through two main factors; the linear and nonlinear properties of the constituent bulk materials, and the optical properties of the waveguide. Recent advances in the design and fabrication of complex structured waveguides with high contrast linear refractive indices, inhomogeneous cross-sections, and subwavelength features have provided great potential to accelerate the field of guided-wave nonlinear optics. We define a new class of optical waveguides, “emerging waveguides”, as waveguides with a combination of the following features:

1. (i) High index materials

2. (ii) Inhomogeneous and complex structure

3. (iii) Subwavelength features such as voids or material inclusions.

Introduction

The first experiments with supercontinuum generation in a photonic crystal fibre (PCF) demonstrated impressive spectra spanning from 400 nm to 1500 nm using 100 fs pulses (Ranka et al., 2000). Often, one does not require the use of the entire supercontinuum bandwidth, or the spectrum needs to be concentrated in a specific spectral region where other lasers are not readily available. One method is to use the soliton self-frequency shift to simply red-shift a laser pulse over a desired wavelength range, which can be done over 900 nm (Chan et al., 2008). This provides a basis for tunable lasers with applications including broadband spectroscopy (Walewski et al., 2004), and coherent anti-Stokes Raman scattering (CARS) microspectroscopy (Andresen et al., 2007). ZBLAN fluoride (a mixture of zirconium, barium, lanthanum, aluminium, and sodium fluorides) fibres have been used to extend a supercontinuum spectrum beyond 4.5 μm with potential applications in spectroscopy (Xia et al., 2006). Besides these examples of generating light in the near- or mid-infrared, one also finds examples of generating light in the ultraviolet–blue region of the spectrum. This wavelength region is highly interesting for several reasons. Primarily, many fluorescent molecules are excited in a wavelength range from ∼600 nm down to ∼350 nm (Prasad, 2003). Supercontinuum light sources covering this wavelength range are highly useful for fluorescence microscopy. In particular, a high spectral density over a broad wavelength range removes the need for using several lasers, each corresponding to the excitation wavelength of a specific fluorescent molecule.