Chiral Photon-Emission Protocol

• The chiral photon-emission protocol leverages confinement-induced spin–momentum locking to map an emitter's dipole helicity onto distinct photonic paths.
• Nanophotonic structures such as waveguides, photonic-crystal circuits, and resonators achieve high directional β-factors for unidirectional single-photon routing and nonreciprocal transport.
• Experimental realizations demonstrate over 90% directionality and near-deterministic spin–photon interfaces, underpinning scalable quantum networks and photonic logic devices.

Searching arXiv for the cited chiral quantum optics papers to ground the article in the source literature. Searching (Lodahl et al., 2016) Chiral Quantum Optics A chiral photon-emission protocol is a direction-dependent light–matter interface in which a quantum emitter emits a photon preferentially, or in the ideal limit deterministically, into one of two counter-propagating optical modes. In chiral quantum optics, the asymmetry is produced by transverse confinement of light in nanophotonic structures, which locks local polarization to propagation direction and thereby makes the overlap between the emitter dipole and the optical mode depend on direction. The protocol therefore maps dipole helicity, and often emitter spin, onto photonic path. Its canonical consequences are unidirectional single-photon emission, direction-selective absorption and scattering, nonreciprocal single-photon transport, cascaded emitter dynamics, and deterministic spin–photon interfaces (Lodahl et al., 2016).

1. Physical principle and scope

The basic mechanism is confinement-induced spin–momentum locking of light. In waveguides, nanofibers, and resonators, the optical field is not purely transverse; it acquires a longitudinal component, so the local field near the emitter becomes elliptically or circularly polarized. The longitudinal component obeys

$$\mathcal{E}_z=\mp\frac{i}{ k}\left(\frac{\partial \mathcal{E}_x}{\partial x}+\frac{\partial\mathcal{E}_y}{\partial y}\right),$$

which changes sign with propagation direction. As a result, the forward and backward guided modes are time-reversal partners but generally have different local polarizations. The spontaneous-emission rates into the two directions are

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$

with

$$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$

so that, in general,

$\gamma_{+}\neq\gamma_{-}.$

The asymmetry is therefore not caused by an external magnetic field in the basic protocol, but by the intrinsic polarization structure of the confined mode and the polarization of the emitter transition (Lodahl et al., 2016).

The cleanest realization is a circularly polarized dipole placed at a point where the local polarizations for opposite propagation directions are circular and orthogonal. In that ideal limit the dipole “solely emits along this direction,” and opposite circular dipoles $\sigma_\pm$ emit into opposite directions. This state-controlled routing was experimentally demonstrated in a glide-plane photonic-crystal waveguide, where the internal state of a quantum dot determines whether an emitted photon exits to the left or right (Söllner et al., 2014).

The term has also broadened beyond spin–momentum locking in nanophotonic modes. In a triatomic artificial giant molecule, chirality arises from interference between emissions from two spatially separated coupling points, with left- and right-moving amplitudes

$\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$

so that destructive interference can suppress one direction entirely (Wang et al., 28 Mar 2026). In a photonic sawtooth lattice, directionality is instead generated by an asymmetric band structure,

$\omega_{u/l}(k)\neq \omega_{u/l}(-k),$

and does not rely on polarization (Sánchez-Burillo et al., 2019). These variants broaden the concept, but the core protocol in chiral quantum optics remains the direction-dependent emission of a quantum emitter into confined optical modes.

2. Formal description and figures of merit

The central efficiency metric is the directional $\beta$-factor,

$$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$

where $\Gamma$ is emission into all other modes, and the total coupling efficiency is $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$0. This quantity determines whether the interface behaves as a weakly perturbative scatterer or as a nearly deterministic photon-emitter interface (Lodahl et al., 2016).

For narrow-band resonant light in the weakly saturated regime, chirality directly modifies the transmission and reflection coefficients:

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$1

The absorption fraction is

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$2

These relations make the nonreciprocal content explicit. For symmetric coupling, absorption is limited to at most $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$3, whereas for ideal chiral coupling and $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$4, one obtains perfect absorption for one direction and transmission for the opposite direction:

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$5

In the ideal unidirectional limit $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$6, $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$7, the emitter is transparent to one propagation direction and does not reflect (Lodahl et al., 2016).

In resonant chiral structures, directional Purcell enhancement is also fundamental. For counterpropagating quasi-normal modes in a whispering-gallery resonator, the directional probabilities are

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$8

and the directional Purcell factors are

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$9

with a direction-dependent chiral mode volume $$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$0. This extends the ordinary cavity-QED mode-volume concept to a position-dependent and direction-dependent chiral quantity (Martín-Cano et al., 2018).

Quantity	Definition	Role
$$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$1	$$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$2	Directional spontaneous-emission rates
$$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$3	$$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$4	Directional coupling efficiencies
$$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$5	$$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$6, $$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$7	Nonreciprocal scattering amplitudes
$$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$8	$$\boldsymbol{\mathcal{E}_+ = \boldsymbol{\mathcal{E}_-^*}}, \qquad \boldsymbol{\mathcal{E}_+ \neq \boldsymbol{\mathcal{E}_-}},$$9	Directional Purcell enhancement

Different platforms also use platform-specific asymmetry metrics. Nanofiber and photonic-crystal works use directional fractions and chiral contrasts such as

$\gamma_{+}\neq\gamma_{-}.$0

or power-flow directionality measures such as

$\gamma_{+}\neq\gamma_{-}.$1

while giant-molecule readout protocols define

$\gamma_{+}\neq\gamma_{-}.$2

These quantities differ operationally, but all encode left-right asymmetry of emission (Xiao et al., 2021, Zhang et al., 2018, Wang et al., 28 Mar 2026).

3. Nanophotonic implementations and experimental realizations

Photonic nanostructures are essential because they supply the confinement needed to generate local circular polarization and suppress or reshape unwanted radiation channels. The principal platforms emphasized in early chiral quantum-optics work are optical nanofibers, glide-plane photonic-crystal waveguides, and whispering-gallery-mode microresonators (Lodahl et al., 2016).

The glide-plane photonic-crystal waveguide is the canonical on-chip implementation. By breaking mirror symmetry through the glide-plane shift, the structure supports counterpropagating modes with opposite local circular polarizations. In a GaAs membrane with embedded In(Ga)As quantum dots, this enabled single-photon emission into the waveguide with a directionality of more than $\gamma_{+}\neq\gamma_{-}.$3 under conditions where practically all emitted photons are coupled to the waveguide. For one quantum dot, the measured average directionality was

$\gamma_{+}\neq\gamma_{-}.$4

with $\gamma_{+}\neq\gamma_{-}.$5, while theory indicated $\gamma_{+}\neq\gamma_{-}.$6 at favorable positions (Söllner et al., 2014).

A related circuit-level implementation uses laterally adjacent GaAs waveguides with InGaAs quantum dots. There, chiral routing originates from waveguide-field chirality and beamsplitting from evanescent coupling between the two waveguides. The local field becomes circular at “chiral points” satisfying

$\gamma_{+}\neq\gamma_{-}.$7

and the directional $\gamma_{+}\neq\gamma_{-}.$8-factors are

$\gamma_{+}\neq\gamma_{-}.$9

The experimentally reported chiral contrast reached $\sigma_\pm$0 and $\sigma_\pm$1 in one-waveguide measurements, and up to $\sigma_\pm$2 in full chiral photonic circuits (Xiao et al., 2021).

Whispering-gallery resonators furnish cavity-enhanced chirality. For a GaP ring resonator with geometric footprint $\sigma_\pm$3, radius $\sigma_\pm$4, width $\sigma_\pm$5, height $\sigma_\pm$6, and $\sigma_\pm$7, two-dimensional eigenmode simulations showed near-perfect directionality, with the plotted directionality reaching $\sigma_\pm$8 in specific spatial regions, directional Purcell factors up to $\sigma_\pm$9 inside the waveguide and $\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$0 outside the bar, and coupling efficiencies approaching $\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$1 inside the waveguide (Martín-Cano et al., 2018).

Hybrid photonic-crystal–plasmonic structures were proposed to combine chirality with strong light–matter coupling. In a W1 photonic-crystal waveguide with an embedded Ag nanoparticle, a $\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$2 emitter placed $\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$3 nm from the nanoparticle and tuned to the quadrupolar mode yielded

$\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$4

with $\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$5 of guided photons propagating in one direction. At $\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$6 and $\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$7 nm, the reported values were

$\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$8

with $\alpha_{L(R)}(t)=\sqrt{\frac{\gamma_0}{2}\left[c_1(t)+e^{\pm(-)i\phi}c_2(t)\right],$9 unidirectional propagation. In the strong-coupling regime of the hybrid band-edge mode, the same architecture gave

$\omega_{u/l}(k)\neq \omega_{u/l}(-k),$0

satisfying $\omega_{u/l}(k)\neq \omega_{u/l}(-k),$1, with $\omega_{u/l}(k)\neq \omega_{u/l}(-k),$2 and a fluorescence linewidth about one-tenth of that for the Ag nanoparticle alone (Zhang et al., 2018).

Open-microcavity implementations realize a chiral one-dimensional atom by combining a low-noise neutral quantum dot with spin–momentum locking in a fiber–microcavity interface. In that geometry, forward light is mapped to the helicity orthogonal to the dipole and does not interact, while backward light has the matching handedness and does interact. This produced a single-photon diode with non-reciprocity as high as $\omega_{u/l}(k)\neq \omega_{u/l}(-k),$3 dB, optimized in situ at $\omega_{u/l}(k)\neq \omega_{u/l}(-k),$4, and reverse-direction transmitted light with

$\omega_{u/l}(k)\neq \omega_{u/l}(-k),$5

showing that the single-photon component is largely removed (Antoniadis et al., 2021).

4. Deterministic interfaces, routing, and nonreciprocal devices

The most direct application of the protocol is directional single-photon emission from a circular dipole. In the ideal case, the local polarizations at the emitter position are circular and orthogonal for opposite propagation directions, so a circularly polarized dipole matched to one propagation direction solely emits along that direction. This is the elementary chiral photon-emission protocol from which more elaborate interfaces are constructed (Lodahl et al., 2016).

When $\omega_{u/l}(k)\neq \omega_{u/l}(-k),$6, the same mechanism becomes a deterministic spin–photon interface. In that regime, the emitter couples almost entirely to the waveguide mode, enabling conversion between a stationary emitter qubit and a flying photonic qubit, together with downstream reabsorption. The canonical state-transfer map is

$\omega_{u/l}(k)\neq \omega_{u/l}(-k),$7

The protocol may require pulse shaping of the flying photon, and $\omega_{u/l}(k)\neq \omega_{u/l}(-k),$8-type three-level systems with Raman processes can be used to achieve it. The directional guarantee supplied by chirality is that the emitted photon propagates in the correct direction and is reabsorbed by the downstream node (Lodahl et al., 2016).

This spin–path conversion also underpins deterministic photonic logic proposals. In the glide-plane photonic-crystal system, the demonstrated chiral emitter was used to motivate a fully scalable deterministic photon-photon CNOT gate. For $\omega_{u/l}(k)\neq \omega_{u/l}(-k),$9, the paper states an entanglement fidelity $\beta$0 and worst-case fidelity $\beta$1 (Söllner et al., 2014).

A second major application class is non-reciprocal single-photon transport. Because forward and backward transport differ, a single emitter can act as an optical isolator, circulator, single-photon switch, or optical diode. The formal scattering relations already show the mechanism; in the most extreme deterministic chiral case, the emitter can impart a nonreciprocal $\beta$2 phase shift while remaining transparent in one direction. The protocol therefore breaks Lorentz reciprocity without requiring magnetic materials, time modulation, or nonlinear optics in the usual sense; chirality arises instead from atomic spin and local optical spin (Lodahl et al., 2016).

Microring-based on-chip isolation makes this explicit. In a silicon microring resonator side-coupled to a waveguide and a negatively charged quantum dot, the transmission probabilities are $\beta$3 and $\beta$4, with isolation contrast

$\beta$5

For the ideal chiral case $\beta$6 and $\beta$7, the resonance values were

$\beta$8

corresponding to insertion loss $\beta$9 dB and $$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$0. The isolation bandwidth was reported as

$$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$1

and the same system supports unidirectional single-photon emission conditioned on the spin-selective quantum-dot transition (Tang et al., 2018).

The open-microcavity diode realizes the same physics at critical coupling. For backward incidence,

$$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$2

so that at

$$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$3

the backward transmission vanishes ideally, while the forward direction remains essentially transparent. This is a chiral photon-emission and photon-absorption protocol rather than merely a passive nonreciprocal element (Antoniadis et al., 2021).

5. Collective, feedback-controlled, and nonlinear extensions

Chiral emission naturally generalizes from single emitters to directed many-emitter networks. For two emitters with perfect unidirectional coupling $$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$4, the reduced dynamics is captured by a cascaded master equation,

$$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$5

with effective Hamiltonian

$$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$6

The term $$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$7 is the cascaded contribution that encodes directional emission from emitter 1 to emitter 2. This structure is the formal basis of chiral quantum channels, directional photonic networks, and reservoir-engineered many-body dynamics (Lodahl et al., 2016).

Measurement-based control adds an explicit feedback layer. For a laser-driven one-dimensional chain of $$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$8 two-level atoms coupled to a nanofiber, chirality is set by

$$\beta_\pm=\frac{\gamma_\pm}{\gamma_++\gamma_-+\Gamma},$$9

and the collective operator is $\Gamma$0. Photon-counting feedback applies a kick

$\Gamma$1

so that the right-moving jump operator becomes

$\Gamma$2

The strongest enhancement occurs at

$\Gamma$3

where each click triggers a $\Gamma$4-pulse and both the photon count rate and its fluctuations increase strongly. Homodyne feedback modifies both the Hamiltonian and dissipator, and possesses a “fully mixed line”

$\Gamma$5

on which the stationary state is fully mixed, independent of $\Gamma$6, $\Gamma$7, and $\Gamma$8 (Buonaiuto et al., 2020).

Collective enhancement can itself make emission nearly perfectly one-way. In an ordered array of atoms near a nanofiber, a single atom emits about $\Gamma$9 of its guided photons to the right, with single-atom chirality $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$00, but a chain of $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$01–$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$02 atoms can form a guided superradiant mode whose phase profile matches the nanofiber mode. With appropriate laser phase matching, the collective chirality reaches

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$03

and the collective $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$04-factor is strongly enhanced (Jones et al., 2019).

The protocol also supports deterministic generation of non-classical light. In a chiral waveguide containing a chain of emitters driven by a coherent field, selecting a single temporal output mode $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$05 via a virtual cavity gives access to the state $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$06 of that mode. The transmitted field obeys

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$07

and in the short-bin limit the selected mode is a displaced vacuum with a weak photon-added contribution,

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$08

The resulting states show Wigner negativity and are well approximated by mixtures of few-photon-added coherent states, with reported fidelities $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$09 in the single-emitter examples (Kleinbeck et al., 2022).

Nonlinear chiral emission appears when the propagating objects are correlated photon pairs rather than single photons. For a pair of giant emitters coupled to a nonlinear waveguide with Kerr interaction,

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$10

the emitted excitations are doublons with dispersion

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$11

The directional decay rates become

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$12

and the chirality

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$13

is controlled by the relative phases at the coupling points. In the optimal geometry, destructive interference cancels one direction entirely, yielding multiphoton chiral emission and a cascaded network of “correlated flying qubits” (Wang et al., 2024).

6. Assumptions, limitations, and related mechanisms

The standard chiral photon-emission protocol is usually derived under the dipole approximation, a Born–Markov treatment of the emitter–reservoir dynamics, and an effectively open one-dimensional geometry with negligible back-reflection. Deterministic behavior requires a high $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$14-factor, ideally close to unity; some state-transfer and gate protocols additionally require pulse shaping or a $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$15-type level structure. Quantum dots may be sensitive to magnetic fields, and multipolar terms can be included when the dipole approximation is insufficient (Lodahl et al., 2016).

Experimentally, chirality is robust but not automatic. The emitter must be positioned at or near a chiral point, the field must be strongly confined on the subwavelength scale, and the local polarization must be controlled accurately. Early isolator devices could operate in a dissipative regime and behave classically; the fully quantum regime requires low loss and strong coupling (Lodahl et al., 2016). This constraint explains the emphasis on photonic band-gap suppression of non-guided modes, resonator enhancement, and low-noise emitters.

A frequent misconception is that chiral emission requires geometrically chiral structures. The central nanophotonic protocol does not. Optical nanofibers, photonic-crystal waveguides, and whispering-gallery resonators are sufficient because confinement alone generates the direction-dependent local polarization responsible for chirality (Lodahl et al., 2016). More recent work has pushed this point further by showing that even achiral particles can emit chiral light under asymmetric excitation in the Purcell regime. In a gold nanorod plus point-emitter system, off-center excitation generates asymmetric surface-plasmon propagation together with spin–momentum locking, and the emitted chirality is converted into nanofiber directionality according to

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$16

At $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$17 nm the measured directionality approached a numerical limit of about $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$18, corresponding to a $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$19 of about $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$20, while simulations at $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$21 nm gave $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$22 (Xuan et al., 20 Oct 2025).

A related but distinct misconception is that all “chiral emission” protocols are waveguide-spin protocols. In fact, some works address chiral light generation rather than directional waveguide-coupled single-photon interfaces. A silicon nanosphere excited by an electron beam can emit circularly polarized light through interference of orthogonal electric dipoles or electric and magnetic dipoles, with Stokes-parameter mapping revealing regions where the ellipticity approaches the circular limit, including values up to

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$23

This is chiral emission in a polarization sense, but it is not the same operational protocol as directional single-photon routing in waveguide QED (Matsukata et al., 2020).

Finally, the broader literature now contains mechanisms in which chirality is implemented through structured baths, synthetic gauge fields, or multi-point interference rather than local optical helicity alone. Photonic quantum Hall systems create bulk chiral channels with Hall velocity

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$24

and at critical coupling

$$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$25

the emitted wavepacket becomes almost time-symmetric and can be reabsorbed by a distant emitter without time-dependent control (Bernardis et al., 2023). This suggests a useful conceptual boundary: the canonical chiral photon-emission protocol is a direction-dependent emitter–photon interface, but the physical origin of the directionality is now platform-dependent.

In that expanded sense, the protocol has evolved from a local statement about $$\gamma_{\pm} \propto \left| {\boldsymbol{d}^*}\cdot \boldsymbol{\mathcal{E}_{\pm}} \right|^2,$$26 into a general design principle for directional quantum light–matter interfaces. Its defining content remains the controlled emission of photons into a preferred direction, together with the reciprocal control of absorption and scattering, but its realizations now span waveguides, resonators, collective atomic arrays, measurement-feedback networks, giant-emitter architectures, and synthetic photonic media (Lodahl et al., 2016).