Kerr-induced Synchronization (KIS)

• Kerr-induced synchronization (KIS) is a phenomenon where Kerr nonlinearity mediates phase locking in microcomb systems, allowing dual-pinned control of the repetition rate.
• It employs an injected reference laser and Adler-type phase dynamics to establish finite locking bandwidths and significant noise reduction.
• KIS spans diverse regimes—including parametric, subharmonic, chaotic, and quantum states—impacting optical frequency division, system stability, and coherent state control.

Searching arXiv for recent and foundational papers on Kerr-induced synchronization to ground the article. Kerr-induced synchronization (KIS) denotes synchronization phenomena in which Kerr nonlinearity is the operative mechanism that establishes, modifies, or suppresses dynamical locking. The term is used most explicitly in the dissipative Kerr soliton (DKS) microcomb literature, where an injected reference laser is captured by a comb tooth and the repetition rate becomes disciplined by all-optical phase locking inside the same resonator (Moille et al., 2023). More broadly, the literature also applies or relates the concept to a driven quantum van der Pol oscillator with Kerr anharmonicity, coupled Kerr parametric oscillators, spatial evolution of optical power in Kerr waveguide arrays, and other Kerr-mediated correlated states, although the degree to which these works adopt synchronization theory as their primary framing is not uniform (Lörch et al., 2016, Yamaji et al., 2022, Deka, 2024, Woodley et al., 2020).

1. Terminology and domain of use

In its narrowest and now standard microcomb usage, KIS is the phase-locking of a DKS comb tooth to an externally injected reference laser. In that setting, the reference is not merely an auxiliary tone: once synchronization is achieved, it is effectively a comb tooth, and the comb repetition rate becomes directly controllable through optical frequency division (Moille et al., 2023).

A broader usage treats KIS as synchronization or correlated motion produced by Kerr-mediated nonlinear coupling even when the locked object is not a soliton comb tooth. The driven quantum van der Pol oscillator with Kerr anharmonicity uses the Kerr term to produce a discrete anharmonic spectrum and thereby multiple synchronization resonances absent in the classical system (Lörch et al., 2016). Two Josephson parametric oscillators in the single-photon Kerr regime exhibit correlated binary-phase oscillations with a pump-phase-controlled effective coupling (Yamaji et al., 2022). In a six-waveguide closed coupler, Kerr nonlinearity changes which waveguides exhibit identical synchronization and, at higher strength, destroys periodic synchronized dynamics (Deka, 2024).

The literature also contains cases where the link to synchronization is explicitly qualified. For counter-propagating light in a Kerr ring microresonator, the observed oscillatory antiphase switching is described as symmetry breaking followed by symmetry-restoring crises and attractor merging; it resembles synchronization only in the sense of nonlinear, antiphase-correlated dynamical locking, rather than in the standard picture of two independent oscillators adjusting to a common frequency (Woodley et al., 2020). This terminological spread is central to the subject: KIS is a coherent family of Kerr-mediated locking phenomena, but not a single universally standardized formalism.

2. Standard dissipative-Kerr-soliton mechanism

The canonical KIS configuration is a DKS microcomb in a silicon nitride microring resonator driven by a main pump, with a second laser injected into another cavity mode, typically near a dispersive wave. The Kerr nonlinearity enables cross-phase modulation and multi-color DKS formation when the injected field is nearly phase matched to the soliton. The essential distinction is between unsynchronized coexistence and true synchronization: if the relative phase

$\Phi=\varphi_{\mathrm{ref}}-\varphi_{\mathrm{dks}}$

has nonzero phase slip, the reference-driven component and the DKS retain distinct phase velocities; in the synchronized state,

$\frac{\partial \Phi}{\partial t}=0,$

so the reference and the corresponding comb tooth are phase locked (Moille et al., 2023).

This dynamics is described by a multi-driven Lugiato–Lefever equation together with an extended Adler equation for the phase difference. In the synchronized regime, the reference behaves exactly like a tooth of the frequency comb. The practical consequence is the dual-pinning of the comb: one pin is the main pump and the second is the synchronized reference mode. Once dual-pinned, tuning the reference no longer changes an internal CEO mismatch; instead, it forces a repetition-rate shift obeying the optical-frequency-division relation

$$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$

with $\mu_s$ the synchronized mode index separation from the main pump (Moille et al., 2023).

The same framework makes clear why synchronization has a finite capture window. The paper gives the locking bandwidth as

$\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$

with representative parameters $\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$ and $$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$, while the measured synchronization window is reported to be of order gigahertz, with about $1.75~\mathrm{GHz}$ of reference frequency tuning over which the CEO mismatch vanishes (Moille et al., 2023). In this narrow sense, KIS is a passive, electronics-free optical clockwork: the Kerr nonlinearity supplies the locking mechanism that would otherwise be imposed by an electronic servo.

3. Adler dynamics, parametric locking, and nonstationary synchronized states

Once the DKS-reference system is treated as a driven phase oscillator, the natural reduced model is a second-order Adler equation. In one formulation,

$$\beta\frac{\partial^2\varphi}{\partial \tau^2}+\frac{\partial \varphi}{\partial \tau}=\alpha+\sin(\varphi),$$

where the second-order term is not incidental: it enables underdamped dynamics, subharmonic locking, and chaos (Moille et al., 11 Sep 2025). This structure underlies the AC-KIS regime produced by phase-modulating the reference laser. In that regime, the optical analogue of the AC Josephson effect appears as integer Shapiro steps, and, because the damping parameter is tunable through reference-pump power, fractional Shapiro steps also appear (Moille et al., 2024).

The physical origin of the integer and fractional steps is not identical. Integer Shapiro steps are explained by the sidebands created through reference-laser phase modulation directly triggering synchronization. Fractional steps arise through a nonlinear chain involving optical parametric amplification and four-wave-mixing Bragg scattering, so that the comb tooth is indirectly captured through sub-comb mode-locking rather than by a single direct sideband-locking event (Moille et al., 2024).

A distinct extension is parametric-KIS of a DKS microcomb. Here, two auxiliary lasers do not directly capture a comb tooth; instead, their Kerr four-wave-mixing interaction generates an effective parametric drive acting on the soliton. The synchronized observable is again the repetition rate, but the locking law is now the “triple pinning” relation

$$\omega_\mathrm{rep}^{(\mathrm{pkis})}=\frac{\omega_-+\omega_+-2\omega_0}{M}, \qquad M=\mu_-+\mu_+.$$

This is an indirect or higher-order KIS mechanism: synchronization is still Adler-like and Kerr-mediated, but the locking force is produced internally by parametric mixing of three intracavity colors rather than by direct tooth capture (Moille et al., 2024).

The same conceptual machinery extends to oscillatory soliton states. For breather solitons, a weak monochromatic laser tuned close to a breathing sideband can capture that sideband and lock the breathing oscillation frequency itself. Near the Hopf bifurcation, the phase dynamics reduce to an Adler equation,

$\frac{\partial \Phi}{\partial t}=0,$0

yielding a locking range $\frac{\partial \Phi}{\partial t}=0,$1 and the familiar transition from frequency pulling to phase locking (Liao et al., 29 Jun 2026). By contrast, phase-modulated KIS can also be driven into chaotic group-velocity hopping, where coexisting carrier-KIS and sideband-KIS attractors produce random repetition-rate transitions while the underlying state remains a DKS (Moille et al., 11 Sep 2025). These developments establish that KIS is not confined to a static lock; it includes RF-driven, subharmonic, parametric, breathing, and chaotic synchronized regimes.

4. Frequency division, noise quenching, and spectral self-balancing

A central consequence of KIS is that it converts optical control into repetition-rate control by optical frequency division. In a terahertz-voltage-controlled oscillator based on a KIS microcomb, modulating the reference laser transfers the waveform onto the repetition rate through

$\frac{\partial \Phi}{\partial t}=0,$2

with $\frac{\partial \Phi}{\partial t}=0,$3 in the demonstrated device. The reported repetition-rate carrier is $\frac{\partial \Phi}{\partial t}=0,$4, and dynamic modulation produces a terahertz frequency comb with about $\frac{\partial \Phi}{\partial t}=0,$5 total bandwidth and roughly $\frac{\partial \Phi}{\partial t}=0,$6 comb teeth (Javid et al., 2024).

The same dual-pinning that enables optical frequency division also changes the microcomb noise model. In a single-pumped DKS, the translational degree of freedom is neutral; under KIS, the position-shifting eigenvalue becomes negative, $\frac{\partial \Phi}{\partial t}=0,$7, and at the center of the synchronization window the strongest damping is $\frac{\partial \Phi}{\partial t}=0,$8. The resulting thermorefractive-noise suppression is

$\frac{\partial \Phi}{\partial t}=0,$9

Experimentally, free-running KIS reduces the repetition-rate noise from about $$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$0 to about $$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$1 at $$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$2 Fourier frequency, while comb-tooth effective linewidths remain in the $$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$3–$$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$4 range over $$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$5 comb lines and the CEO linewidth is reduced from a $$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$6 single-pump estimate to $$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$7, with an experimental KIS CEO linewidth of about $$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$8 (Moille et al., 2024). The dominant noise floor is thereby shifted from internal cavity noise to the external pump-laser frequency noise.

KIS also alters the internal energy flow of the DKS. In the self-balancing picture, the synchronized reference becomes part of the intracavity DKS state, so the added asymmetric power can no longer be absorbed by a repetition-rate change. Instead, the comb redistributes spectral energy to maintain its center of mass. In a pure quadratic-dispersion resonator this appears as reference-dependent recoil; with higher-order dispersion it enhances dispersive-wave emission. Experimentally, this produces a $$\mathrm{d}\delta\omega_{\mathrm{ref}}=\mu_s\,\mathrm{d}\omega_{\mathrm{rep}},$$9 increase of comb teeth at $\mu_s$0 in an octave-spanning microcomb (Shandilya et al., 6 Feb 2026).

A further generalization is “color-KIS,” where multi-pumping creates a secondary color wavepacket and a tunable synthetic dispersive wave at otherwise inaccessible modes. Because the secondary color and the main DKS share a unique group velocity through cross-phase modulation, KIS applied to the secondary color automatically disciplines the main DKS repetition rate. In the reported comparison, the color-KIS bandwidth is about $\mu_s$1 times larger than direct KIS, with representative values of about $\mu_s$2 versus $\mu_s$3, together with synthetic-dispersive-wave power improvements of about $\mu_s$4 and $\mu_s$5, and more than $\mu_s$6 of KIS-window tuning (Moille et al., 2024). This decouples efficient optical frequency division from the fixed dispersive-wave structure of the main pump color.

5. Other optical synchronization problems governed by Kerr nonlinearity

Outside the DKS literature, Kerr nonlinearity organizes several other forms of optical synchronization, but its role is not always synchronization-enhancing. A six-waveguide closed coupler with nearest-neighbor evanescent coupling and equal Kerr nonlinearity in every waveguide is modeled by the DNLSE

$\mu_s$7

For $\mu_s$8 and $\mu_s$9, the spatial evolution is regular and periodic, and the PSD of $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$0 shows a single sharp peak at $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$1. For $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$2 and $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$3, synchronization persists only up to about $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$4, beyond which the evolution becomes aperiodic; the synchronized pairs also change, with $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$5 versus $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$6 and $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$7 versus $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$8 showing straight lines indicating identical synchronization. Increasing the coupling to $\Delta\omega_{\mathrm{lock}}=2\mathcal{T},$9 at fixed $\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$0 suppresses the aperiodicity and restores periodicity, with the PSD regaining a distinct sharp peak at $\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$1. The paper’s explicit interpretation is that Kerr nonlinearity acts as a disorder parameter and the coupling constant acts as an antagonist that restores synchronization (Deka, 2024).

In balanced counter-propagating Kerr ring microresonators, the central phenomenon is oscillatory antiphase switching between clockwise and counter-clockwise fields. The governing envelope equations include only self- and cross-phase Kerr nonlinearities,

$\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$2

with $\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$3 and $\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$4. The observed dynamics proceed from spontaneous symmetry breaking to symmetry-restoring crises and attractor merging. This supports antiphase-correlated motion and transient synchronization, but the primary conceptual framing is self-switching rather than conventional synchronization (Woodley et al., 2020).

Kerr-comb synchronization is also not restricted to soliton states. In the normal-GVD, non-solitonic regime, two unidirectionally coupled Si$\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$5N$\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$6 coupled-ring combs synchronize their repetition rates when a filtered fraction of the primary comb is injected into the secondary device. With coupling strength $\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$7, synchronization is observed for heater powers from $\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$8 to $\mathcal{T}\approx 2\pi\times 365~\mathrm{MHz}$9, corresponding to a synchronized repetition-rate range from $$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$0 to $$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$1, and the synchronization boundary forms an Arnold tongue (Kim et al., 2021).

A related but distinct locking problem appears in pulsed-driven Kerr microresonators. There, temporal cavity solitons can frequency-lock or trap to a pump pulse train whose repetition period is close to, but not exactly equal to, the cavity round-trip time. The desynchronization

$$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$2

enters as a drift term in the LLE, and synchronization occurs when the mismatch-induced drift is compensated by the force produced by the pump-envelope gradient. The same mechanism can enforce an attractive single-soliton regime, while stimulated Raman scattering competes with desynchronization and explains asymmetric soliton steps (Hendry et al., 2019).

6. Quantum, circuit, and optomechanical realizations

In the driven quantum van der Pol oscillator with Kerr anharmonicity, KIS refers to synchronization arising because the oscillator frequency depends on excitation number through the Kerr term

$$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$3

The anharmonic spectrum produces a family of resolved synchronization resonances at

$$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$4

rather than the single broad resonance of the classical Kerr-van der Pol oscillator. Phase locking is quantified through

$$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$5

and strong driving near these resonances generates nonclassical steady states with negative Wigner density (Lörch et al., 2016). In this setting, KIS is a genuinely quantum synchronization mechanism tied directly to quantized anharmonic transitions.

In superconducting circuit implementations, two Josephson parametric oscillators in the single-photon Kerr regime, coupled via a static capacitance, exhibit correlated binary phases whose parity and strength are controlled by the relative pump phase. The effective Ising coupling is

$$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$6

so $$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$7 gives maximal ferromagnetic correlation, $$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$8 gives vanishing coupling, and $$\Delta\omega_{\mathrm{lock}}\approx 2\pi\times 730~\mathrm{MHz}$$9 gives maximal antiferromagnetic correlation. Experimentally, the same-phase probability is about $1.75~\mathrm{GHz}$0 at $1.75~\mathrm{GHz}$1, correlation disappears at $1.75~\mathrm{GHz}$2, and antiferromagnetic correlation appears at $1.75~\mathrm{GHz}$3; simulations reproduce this only when pure dephasing is included, with $1.75~\mathrm{GHz}$4 (Yamaji et al., 2022).

Kerr-enhanced synchronization also appears in coupled optomechanical systems. Two mechanical oscillators of different frequencies, each embedded in a cavity filled with a Kerr-type nonlinear medium, can be coupled either through phonon tunneling or through optical-fiber-mediated photon hopping. The synchronization measure is

$1.75~\mathrm{GHz}$5

with $1.75~\mathrm{GHz}$6 and $1.75~\mathrm{GHz}$7. The reported conclusion is that increasing Kerr nonlinearity $1.75~\mathrm{GHz}$8 increases the synchronization measure regardless of coupling or periodic modulation choice, so the nonlinear system shows stronger and easier-to-observe quantum synchronization than the corresponding linear system; with sufficiently large $1.75~\mathrm{GHz}$9, $$\beta\frac{\partial^2\varphi}{\partial \tau^2}+\frac{\partial \varphi}{\partial \tau}=\alpha+\sin(\varphi),$$0 can exceed $$\beta\frac{\partial^2\varphi}{\partial \tau^2}+\frac{\partial \varphi}{\partial \tau}=\alpha+\sin(\varphi),$$1 and even approach $$\beta\frac{\partial^2\varphi}{\partial \tau^2}+\frac{\partial \varphi}{\partial \tau}=\alpha+\sin(\varphi),$$2 (Qiao et al., 2018).

Taken together, these realizations show that KIS is not uniformly synchronization-enhancing or synchronization-disrupting. In the six-waveguide coupler, higher Kerr strength acts as a disorder parameter; in the quantum van der Pol oscillator it creates multiple resolved synchronization resonances; in coupled optomechanical systems it greatly enhances synchronization; and in DKS microcombs it supplies a passive all-optical route to repetition-rate disciplining, optical frequency division, noise quenching, and controlled access to nonstationary synchronized states (Deka, 2024, Lörch et al., 2016, Qiao et al., 2018, Moille et al., 2023).