Spin-Momentum Locking in Nanophotonics

• Spin–momentum locking is the phenomenon where the electromagnetic field’s local spin is intrinsically tied to its propagation direction.
• This effect facilitates chiral, unidirectional light routing in systems such as plasmonic nanowires, waveguides, and photonic crystals.
• Experimental techniques like polarization-resolved imaging confirm near-unidirectional coupling with efficiencies reaching up to 0.9.

Spin–momentum locking in nanophotonics refers to the intrinsic correspondence between the local polarization state (spin) of electromagnetic fields and their propagation direction (momentum), particularly in systems supporting tightly confined, guided, or evanescent modes. The phenomenon ensures that the transverse spin angular momentum (SAM)—arising from field ellipticity enforced by Maxwell’s equations and boundary conditions—is rigidly determined by the sign of the local wavevector, enabling chiral (unidirectional) coupling, robust routing of light, and novel spin-based device functionalities at the subwavelength scale. This principle underpins a broad class of effects and applications in photonic crystals, waveguides, plasmonic structures, metasurfaces, and quantum-light interfaces.

1. Electromagnetic Foundation and Definition

The electromagnetic fields associated with nanophotonic modes—found in waveguides, surface states, and structured media—exhibit spin–momentum locking as a direct consequence of their complex dispersion relations and boundary-imposed transversality. For monochromatic fields in linear, nonmagnetic media, Maxwell’s equations enforce that the wavevector $\mathbf{k}$ must generally have both real (propagation) and imaginary (decay) components: $$\mathbf{k} = \pmb{\kappa} + i\pmb{\eta},\quad \pmb{\kappa},\pmb{\eta}\,\text{real},\quad \pmb{\kappa}\cdot\pmb{\eta} = 0.$$ This orthogonality, essential in evanescent or guided states, produces a built-in $\pi/2$ phase shift between field components aligned with $\pmb{\kappa}$ and $\pmb{\eta}$. The local time-averaged spin density is

$$\mathbf{S}(\mathbf{r}) = \frac{\epsilon_0}{4\omega}\,\Im[\mathbf{E}^*(\mathbf{r})\times \mathbf{E}(\mathbf{r})] + \frac{\mu_0}{4\omega}\,\Im[\mathbf{H}^*(\mathbf{r})\times\mathbf{H}(\mathbf{r})].$$

The critical feature is that the direction of $\mathbf{S}$ is strictly determined by the propagation direction $\pmb{\kappa}$ of the mode and changes sign with it. This “locking” is universal: it arises from the fundamental properties of Maxwell’s equations together with causality, which demands physically acceptable (i.e., nondivergent) solutions (Mechelen et al., 2015).

2. Canonical Realizations and Physical Consequences

2.1. Evanescent Waves and Total Internal Reflection

For evanescent fields at dielectric interfaces (e.g., total internal reflection, TIR), the right-handed triad $(\pmb{\kappa},\pmb{\eta},\mathbf{\hat{s}})$—propagation, decay, and spin axes, respectively—emerges universally. The spin vector $\mathbf{\hat{s}}$ is perpendicular to both the propagation and decay directions: $$\mathbf{\hat{s}} = \frac{\pmb{\kappa} \times \pmb{\eta}}{|\pmb{\kappa} \times \pmb{\eta}|}.$$ The degree of local circular polarization (Stokes parameter $S_3$) flips with $\pmb{\kappa}$, ensuring intrinsic spin–momentum locking: reversing the propagation direction reverses the spin (Mechelen et al., 2015, Kalhor et al., 2015).

2.2. Surface Plasmon Polaritons and Nanowires

Surface plasmon polariton (SPP) modes at metal/dielectric interfaces and along plasmonic nanowires present near-field electromagnetic distributions exhibiting strong transverse spin angular momentum, with the spin direction determined by momentum. The normalized parameter $S_3 \approx \pm1$ is achieved at the nanowire surface, and the coupling of a local emitter (such as a circularly polarized dipole or valley-polarized exciton) preferentially excites the matching mode, leading to nearly unidirectional propagation and coupling efficiencies up to 0.9 as shown in combined theory and experiment (Gong et al., 2017, Taneja et al., 2021).

2.3. Waveguides, Fibers, and Photonic Crystals

In dielectric waveguides, such as silicon strips or optical fibers, the fundamental guided modes possess transverse SAM (in the direction orthogonal to the group velocity and decay), again locked to the sign of the propagation constant. The $HE_{11}$ mode in fibers is a prototypical example: the transverse spin circulates around the fiber axis, with its sense set by the local momentum (Kalhor et al., 2015, Luo et al., 2017, Mechelen et al., 2015). In photonic crystals, spin–momentum locking emerges at boundaries, edges, and in defect or anti-phase interfaces, where edge states with transverse spin can be harnessed for chiral, unidirectional light guiding (Kong et al., 2019).

3. Microscopic Origins and Theoretical Framework

The microscopic origin of spin–momentum locking lies in the nontrivial vectorial field structure imposed by geometry and material boundaries. Maxwell’s equations together with boundary conditions require a $\pi/2$ phase shift between field components in the decay and propagation directions in evanescent and guided modes. Analytical and numerical analyses confirm that the local spin density is proportional to the imaginary part of the Poynting vector components, and Stokes analysis demonstrates that the probability amplitude for right- or left-handed circular polarization is maximized for fast-decaying modes, reaching $|S_3| \to 1$ (Mechelen et al., 2015, Kalhor et al., 2015).

In structured environments—such as photonic lattices, metasurfaces, and metamaterials—spin–momentum locking is further influenced by material dispersion and symmetry. Anisotropic and bianisotropic media, as in engineered metamaterials, permit the creation of multiple spin-momentum channels and even “anomalous” (out-of-plane) transverse spin components when symmetry is explicitly broken (Li et al., 2023, Kandil et al., 2021).

The general hydrodynamic and optical momentum and spin densities—canonical (Minkowski) and kinetic (Abraham)—are required for quantitative analysis, especially in dispersive systems where the directionality and chirality of the spin–momentum relation can flip, e.g., obeying left-hand rules in metals and right-hand rules in dielectrics (Shi et al., 2022).

4. Experimental Realization and Measurement Protocols

Robust experimental protocols have established the direct measurement and utilization of spin–momentum locking in a range of systems. Key approaches include:

• Polarization-Resolved Fourier-Plane Imaging: Used to measure the spin-dependent emission directionality from local sources (e.g., quantum dots near hyperbolic metamaterials, valley excitons in TMDs, or dipoles near SPP waveguides), extracting parameters such as directionality contrast ($D$), normalized Stokes parameter $S_3$, and chirality ($C$) (Gong et al., 2017, Taneja et al., 2021, Yadav et al., 2020).
• Local Probes and Far-Field Readout: Demonstrated in interfaces combining quantum emitters with metasurfaces or photonic crystals, where input photon polarization state (spin) directly maps to output port or direction, with efficiency and crosstalk quantified at the percent level (Li et al., 2023, Hinamoto et al., 2021, Lorén et al., 2022).
• Nonlinear and Spintronic Readout: In optoelectronic devices, guided photon spin–momentum locking is transduced into spin-polarized photocurrents in materials exhibiting their own electronic spin–momentum locking (e.g., topological insulators), validated by nonreciprocal current-phase reversals in response to light direction (Luo et al., 2017).

5. Device Architectures and Functional Applications

Spin–momentum locking enables novel device concepts and functionalities in nanophotonics, including:

• Chiral Quantum Optics: Deterministic, spin-selective routing and emission of single photons through spin–matched modes, as in quantum dot–metamaterial and WS$_2$–nanowire systems, impacting quantum information processing and valleytronics (Gong et al., 2017, Yadav et al., 2020).
• Nonreciprocal and Reconfigurable Photonic Components: On-chip unidirectional couplers, isolators, and routers employing band inversion or geometric reconfiguration to switch between propagation channels (Kong et al., 2019, Hinamoto et al., 2021).
• Spin Sorting and Multiplexing: Structures such as zigzag nanoparticle chains and twinned hyperbolic metamaterials create multi-channel optical spin routers and multiplexers with subwavelength footprints and high directionality contrast (Hinamoto et al., 2021, Li et al., 2023).
• Metasurface and Plasmonic Logic: Spin–momentum locked metasurfaces based on Pancharatnam–Berry phases enable robust (though approximate) spin-based logic and far-field response engineering, subject to corrections near plasmonic and resonance conditions (Lorén et al., 2022, Kandil et al., 2021).

6. Theoretical Extensions and Universal Principles

Spin–momentum locking is a unifying principle for a broad class of systems beyond specific material or geometric implementations. Its universality extends to:

• Dispersion Tailoring: By engineering material dispersion (e.g., in metals with negative permittivity or plasmonic/magnetic bands), the chirality (handedness) of spin–momentum locking can be selected and tuned, allowing for left-hand or right-hand rule behavior (Shi et al., 2022).
• Topological Lattices and Skyrmion Textures: Structural symmetry enriches the local spin–momentum landscape, enabling the realization of topological spin lattices (skyrmions, merons) with robust local locking even in the presence of disorder or lattice defects (Shi et al., 2022).
• Chiral Optical Forces: The transverse spin density locked to momentum underlies nontrivial lateral optical forces and chiral manipulation of particles, distinct from forces arising due to orbital angular momentum (Kalhor et al., 2015).
• Analogy to Quantum Spin Hall Effect: The strict one-to-one mapping between spin and propagation direction in photonic systems mirrors the non-reciprocal, backscattering-immune edge transport of quantum spin Hall states, albeit realized through Maxwell’s constraints rather than electron spin–orbit coupling (Mechelen et al., 2015).

7. Open Challenges and Research Directions

Despite the maturity of foundational theory and diverse experimental validation, several frontiers remain:

• Disorder Robustness: The stability of spin–momentum locked modes, particularly non-topological edge states in photonic crystals and metasurfaces, against fabrication-induced or environmental disorder is incompletely characterized (Kong et al., 2019, Lorén et al., 2022).
• Active Control and Dynamic Reconfiguration: Real-time tuning of geometric and material parameters (e.g., via MEMS or electro-optic control) to reversibly program spin–momentum channels and directionality is a target for integrable, reconfigurable circuits (Kong et al., 2019, Li et al., 2023).
• Multimode and Multichannel Systems: Extension to systems supporting multiple orthogonal spin–momentum locked channels, the management of crosstalk, and integration with complex on-chip architectures (Li et al., 2023).
• Exploration of Anomalous Spin Components: Engineering and exploiting out-of-plane and higher-order transverse spin densities for sophisticated spin–orbit optical control (Kandil et al., 2021).
• Quantum Photonic Interfaces: Deterministic, lossless spin-to-channel mapping at the single-photon level, coherent coupling to electronic or valley degrees of freedom, and integration with quantum information protocols (Gong et al., 2017, Luo et al., 2017).

Spin–momentum locking thus constitutes a foundational and versatile principle in nanophotonics, enabling a host of chiral optical effects, robust light–matter interfaces, and topologically-inspired device functionalities that are inherently encoded in Maxwell’s equations and the geometry of the electromagnetic field (Mechelen et al., 2015, Kalhor et al., 2015, Gong et al., 2017, Kong et al., 2019, Hinamoto et al., 2021, Li et al., 2023, Shi et al., 2022, Yadav et al., 2020, Lorén et al., 2022, Taneja et al., 2021, Luo et al., 2017, Kandil et al., 2021).