10.5 Plasmonics

Plasmons in Metals

Plasmons are collective oscillations of free electrons in metals, excited by electromagnetic waves. They sit at the intersection of solid-state physics and optics, and they're central to manipulating light at scales smaller than its wavelength. This matters for technologies ranging from solar cells to biosensors.

Collective Oscillations of Electrons

When an external electromagnetic field hits a metal, the free electrons don't respond individually. Instead, they move together in phase, creating a coordinated oscillation of the entire electron density. Think of it like a crowd doing "the wave" in a stadium: each person moves on their own, but the collective motion produces something much larger.

This collective behavior is what makes plasmons special. Because many electrons oscillate together, the interaction with light is far stronger than you'd get from individual electron transitions. That's the foundation of nearly every plasmonic application.

Plasmon Frequency and Dispersion

Every metal has a characteristic plasma frequency ($\omega_p$) at which these collective oscillations naturally occur. It's set by the electron density ($n_e$) and the effective electron mass ($m^*$):

$\omega_p = \sqrt{\frac{n_e e^2}{\epsilon_0 m^*}}$

where $e$ is the electron charge and $\epsilon_0$ is the permittivity of free space. Metals with higher free-electron densities (like aluminum) have higher plasma frequencies.

The dispersion relation describes how the plasmon frequency varies with wavevector $k$. This relationship governs how plasmons propagate and how tightly they can be confined. Below $\omega_p$, the metal reflects incident light; above it, the metal becomes transparent. That threshold behavior is a direct consequence of the plasmon dispersion.

Bulk vs. Surface Plasmons

Plasmons come in two main varieties:

• Bulk plasmons oscillate throughout the interior of the metal. They sit at or near $\omega_p$ and are difficult to excite with light because of a momentum mismatch: the photon simply doesn't carry enough momentum at that frequency.
• Surface plasmons are bound to the metal's surface. They occur at lower frequencies (for a metal-air interface, near $\omega_p / \sqrt{2}$) and can be excited by light under the right conditions.

Surface plasmons are far more useful in practice. Their fields are tightly confined to the surface, making them extremely sensitive to whatever is sitting on or near that surface. That sensitivity is what drives plasmonic sensing.

Excitation of Plasmons

Getting light to actually generate plasmons requires overcoming a fundamental obstacle: momentum matching. A free-space photon at a given frequency carries less momentum than a surface plasmon at the same frequency, so you can't just shine light on a flat metal surface and expect plasmons to appear.

Electromagnetic Waves and Plasmons

The electric field of incident light drives the free electrons into oscillation. Whether this coupling is efficient depends on the light's polarization, angle of incidence, and wavelength. For surface plasmons, only p-polarized (TM) light can couple, because the electric field must have a component perpendicular to the surface.

Coupling Light to Plasmons

Several techniques bridge the momentum gap between photons and surface plasmons:

1. Prism coupling (Kretschmann configuration): Light passes through a high-index prism and hits a thin metal film deposited on the prism base. The prism increases the photon's in-plane momentum enough to match the plasmon. This is the most common lab method.
2. Prism coupling (Otto configuration): A prism is placed near (but not touching) the metal surface, with a thin air gap. Evanescent waves tunnel across the gap to excite plasmons.
3. Grating coupling: A periodic corrugation on the metal surface adds discrete momentum kicks ($\Delta k = 2\pi / \Lambda$, where $\Lambda$ is the grating period) to the incident photon.
4. Near-field excitation: A sharp tip or nano-aperture placed very close to the surface provides the high-wavevector components needed for coupling.

Plasmon Resonance Conditions

Resonance occurs when the incident light frequency matches the natural plasmon frequency of the structure. At resonance:

• Light absorption increases dramatically
• Scattering cross-sections can exceed the physical size of the structure
• Local electric fields are strongly enhanced near the metal surface

The resonance condition depends on the geometry, size, and dielectric environment of the plasmonic structure. By tuning these parameters, you can design structures that resonate at specific target wavelengths.

Localized Surface Plasmons

Localized surface plasmons (LSPs) are non-propagating oscillations confined to metallic nanostructures like nanoparticles or nanoantennas. Unlike propagating surface plasmons, LSPs don't travel along a surface. Instead, the electrons slosh back and forth within the nanostructure itself.

Plasmons in Metal Nanoparticles

Gold and silver nanoparticles are the workhorses of LSP research. Their small size means the entire electron cloud can be displaced by the incident field, creating a strong dipolar (or higher-order) oscillation.

The vivid colors of LSP resonances have been exploited for centuries, even before anyone understood the physics. The deep reds and purples in medieval stained glass come from gold nanoparticles embedded in the glass. The Lycurgus cup (4th century Roman glass) appears green in reflected light but red in transmitted light, all due to gold-silver alloy nanoparticles roughly 70 nm in diameter.

The LSP resonance frequency is set by the nanoparticle's size, shape, and composition, along with the dielectric function of the surrounding medium.

Size and Shape Dependence

• Size: Smaller nanoparticles have sharper, higher-frequency resonances with stronger field confinement. As particles grow larger (beyond ~100 nm for gold), the resonance broadens and redshifts due to retardation effects, and higher-order multipole modes appear.
• Shape: Spherical particles support a single dipolar LSP mode. Anisotropic shapes unlock additional modes. Gold nanorods, for example, have two distinct resonances: a transverse mode (short axis, around 520 nm) and a longitudinal mode (long axis, tunable from visible to near-infrared depending on the aspect ratio).
• Triangular, cubic, and star-shaped nanoparticles concentrate fields at their sharp tips and edges, producing intense "hot spots."

Near-Field Enhancement Effects

The most striking feature of LSPs is the intense electric field generated right at the nanoparticle surface. This near-field enhancement can amplify the local field intensity by factors of $10^2$ to $10^4$ compared to the incident field.

The enhancement decays rapidly with distance from the surface, typically falling off within 10–30 nm. Applications that exploit this effect include:

• Surface-enhanced Raman scattering (SERS): Raman signals scale as roughly the fourth power of the field enhancement
• Near-field optical microscopy: Imaging below the diffraction limit
• Nonlinear optics: Processes like second-harmonic generation become much more efficient in enhanced fields

The strongest enhancements occur in narrow gaps between two nanoparticles (gap < 5 nm), where the fields from each particle couple to create extremely intense hot spots.

Propagating Surface Plasmons

Propagating surface plasmons (PSPs), also called surface plasmon polaritons (SPPs), are electromagnetic waves that travel along a metal-dielectric interface. Unlike LSPs, these modes propagate and can carry information over micrometer-scale distances.

Plasmons at Metal-Dielectric Interfaces

A PSP exists at the boundary between a metal (gold, silver, aluminum) and a dielectric (air, glass, water). The mode is a hybrid: part electromagnetic wave in the dielectric, part collective electron oscillation in the metal.

The field profile is highly asymmetric. Fields peak right at the interface and decay exponentially into both media. The decay length into the dielectric (typically 100–500 nm) is much longer than into the metal (typically 10–30 nm at visible frequencies). This confinement to the surface is what gives PSPs their subwavelength character.

Dispersion Relation and Propagation Length

The PSP dispersion relation lies to the right of the light line in a frequency-vs-wavevector plot. This means at any given frequency, the PSP has a larger wavevector (more momentum) than a free-space photon. That's exactly why direct excitation by light on a flat surface is impossible without a coupling mechanism.

As the frequency approaches the surface plasmon frequency ($\omega_{sp} \approx \omega_p / \sqrt{1 + \epsilon_d}$, where $\epsilon_d$ is the dielectric constant), the dispersion curve flattens and the group velocity drops toward zero.

Propagation length is limited by ohmic losses in the metal. Typical values:

• Silver at 600 nm: ~10–50 $\mu$m
• Gold at 800 nm: ~10–20 $\mu$m
• At near-infrared wavelengths, propagation lengths increase because metal losses decrease

There's always a trade-off: tighter confinement means shorter propagation.

Plasmonic Waveguides and Circuits

PSPs can be routed along nanoscale waveguide structures:

• Metal stripes on a dielectric substrate
• V-grooves cut into a metal surface (channel plasmon polaritons)
• Metal nanowires that act as one-dimensional waveguides

These waveguides confine light to dimensions far below the diffraction limit, enabling plasmonic circuits with components like splitters, interferometers, and modulators, all at scales compatible with electronic integrated circuits. The goal is to combine the speed of photonics with the miniaturization of electronics, sometimes called "the best of both worlds" for on-chip data processing.

Applications of Plasmonics

The strong field enhancement, subwavelength confinement, and environmental sensitivity of plasmons translate into a wide range of practical applications.

Surface-Enhanced Raman Spectroscopy (SERS)

Raman scattering is normally very weak (roughly 1 in $10^7$ photons), which limits its usefulness. SERS overcomes this by placing molecules near plasmonic nanostructures where the local field is enormously amplified.

The Raman signal scales approximately as $|E/E_0|^4$, so a field enhancement of 100× gives a signal boost of $\sim 10^8$. This is enough for single-molecule detection.

SERS substrates are typically rough silver or gold surfaces, colloidal nanoparticle aggregates, or lithographically patterned nanostructure arrays. Applications span chemical sensing, forensic trace analysis, biological diagnostics, and materials characterization.

Plasmonic Biosensing and Diagnostics

Plasmonic biosensors detect changes in the refractive index near the metal surface caused by molecular binding events.

• SPR sensors monitor the angle or wavelength shift of the plasmon resonance when target biomolecules (proteins, DNA, antibodies) bind to receptor molecules on the metal surface. Commercial SPR instruments can detect refractive index changes as small as $10^{-6}$ RIU (refractive index units).
• LSPR sensors use the resonance shift of nanoparticle LSPs. They're simpler and cheaper than SPR setups, though typically less sensitive.
• Plasmonic interferometers measure phase changes in propagating plasmons for even higher sensitivity.

These sensors are label-free (no fluorescent tags needed) and can monitor binding kinetics in real time, making them valuable for drug discovery and clinical diagnostics.

Plasmonic Metamaterials and Cloaking

Metamaterials are engineered structures with optical properties not found in nature. By arranging plasmonic nanostructures in specific geometries (split-ring resonators, fishnet structures), you can create materials with:

• Negative refractive index: Light bends the "wrong" way at the interface
• Perfect absorption: Nearly all incident light is absorbed at a target wavelength
• Optical magnetism: Magnetic response at optical frequencies, which natural materials lack

These properties enable exotic phenomena like superlensing (imaging below the diffraction limit) and cloaking (guiding light around an object so it casts no shadow). Current cloaking demonstrations work only over narrow frequency bands and for small objects, but the underlying physics is well established.

Plasmonic Solar Cells and Energy Harvesting

Plasmonics can boost solar cell efficiency through several mechanisms:

1. Light trapping: Metal nanoparticles on the cell surface scatter incident light at large angles, increasing the optical path length through the absorber layer.
2. Near-field enhancement: LSPs from nanoparticles embedded near the active layer increase absorption, especially in thin-film cells where the absorber is only a few hundred nanometers thick.
3. Far-field scattering: Larger nanoparticles (>100 nm) act as effective scatterers that redirect light into the solar cell.

These approaches are particularly useful in the near-infrared, where silicon absorbs weakly. Plasmonic enhancements have demonstrated absorption improvements of 20–50% in thin-film architectures.

Advanced Topics in Plasmonics

These topics push plasmonics into regimes where classical descriptions start to break down or where plasmons interact with other quantum excitations.

Quantum Plasmonics and Single-Photon Sources

When plasmonic structures shrink below ~10 nm, quantum effects become important. The classical Drude model fails, and nonlocal and electron-spill-out effects alter the plasmon resonance.

Quantum emitters (quantum dots, nitrogen-vacancy centers, single molecules) placed near plasmonic nanostructures experience dramatically enhanced spontaneous emission rates (the Purcell effect). Plasmonic nanoantennas can:

• Increase the radiative rate of a quantum emitter by factors of $10^2$–$10^3$
• Direct single-photon emission into specific directions and polarizations
• Serve as building blocks for quantum information devices

The challenge is that metal losses also increase non-radiative decay, so careful design is needed to maximize the radiative enhancement.

Nonlinear Plasmonics and Second-Harmonic Generation

Nonlinear optical processes require intense fields, which is exactly what plasmonic nanostructures provide. The local field enhancement at a nanoparticle surface can boost nonlinear signals by many orders of magnitude.

• Second-harmonic generation (SHG): Input light at frequency $\omega$ produces output at $2\omega$. In centrosymmetric metals like gold, SHG is surface-sensitive, making it a probe of symmetry breaking at interfaces.
• Third-harmonic generation (THG) and four-wave mixing (FWM) are also enhanced by plasmonic fields.

These effects enable frequency conversion and all-optical switching at the nanoscale, with potential for integrated photonic circuits.

Ultrafast Plasmonics and Femtosecond Dynamics

Plasmon excitations evolve on extremely fast timescales. After a femtosecond ($10^{-15}$ s) laser pulse excites a plasmon, the sequence of events is roughly:

1. Plasmon dephasing (~10–20 fs): The coherent oscillation loses phase coherence through electron-electron and surface scattering.
2. Hot-electron generation (~100 fs): The plasmon energy redistributes into a non-thermal electron distribution, creating "hot" electrons with energies 1–3 eV above the Fermi level.
3. Electron thermalization (~100 fs–1 ps): Electron-electron scattering establishes a hot Fermi-Dirac distribution.
4. Electron-phonon coupling (~1–10 ps): The hot electrons transfer energy to the lattice.
5. Heat dissipation (~100 ps–ns): The nanostructure cools by heat conduction to the surroundings.

Hot electrons generated in step 2 are of particular interest for photocatalysis and photodetection, since they can be injected into adjacent semiconductors before thermalizing.

Plasmon-Exciton Interactions and Strong Coupling

When a plasmonic mode and an excitonic transition (in a semiconductor quantum dot, J-aggregate, or transition metal dichalcogenide) are brought into resonance and interact strongly enough, they form hybrid states called plexcitons.

The hallmark of strong coupling is an anticrossing in the dispersion: instead of the plasmon and exciton modes crossing, they repel each other, forming upper and lower polariton branches separated by the Rabi splitting ($\Omega_R$). Strong coupling requires that $\Omega_R$ exceeds the linewidths of both the plasmon and the exciton.

Plexcitons combine the strong field confinement of plasmons with the long coherence times of excitons. Potential applications include polariton lasing, modified chemical reactivity, and ultrafast energy transfer in optoelectronic devices.