Ultrafast Spatiotemporal Mode-Locking

• Ultrafast spatiotemporal mode-locking is a nonlinear optical process that synchronizes both spatial and temporal modes to generate ultrashort, high-power, three-dimensional coherent pulses.
• It leverages nonlinear intermodal coupling, advanced spectral and spatial filtering, and saturable absorption to stabilize dissipative solitons in multimode laser systems.
• Experimental platforms such as fiber lasers, microresonators, and photonic lattices demonstrate enhanced pulse energies, metrological precision, and customizable beam structures.

Ultrafast spatiotemporal mode-locking (STML) is the nonlinear optical process in which both longitudinal (temporal/frequency) and transverse (spatial/mode) degrees of freedom in a multimode laser cavity become mutually coherent, thereby producing single ultrashort pulses with three-dimensional field structures that repeat stably every cavity round trip. In contrast to conventional temporal mode-locking, which synchronizes only longitudinal modes to generate transform-limited pulses, STML establishes a fixed phase relationship among a high-dimensional set of spatial and temporal modes, enabling the generation of dissipative spatiotemporal solitons or eigenpulses. STML has been experimentally realized in fiber lasers, microresonator frequency combs, chip-scale photonic platforms, and via spatial structuring of vortex pulses. This regime offers greatly enhanced pulse energies, peak powers, and the ability to sculpt the three-dimensional coherence of ultrafast light, with implications for nonlinear optics, metrology, quantum information, and high-power photonic applications (Wright et al., 2017, Nie et al., 2022, Gao et al., 2024, Xu et al., 21 Feb 2025, Liu et al., 2024).

1. Theoretical Foundations and Modeling Approaches

The spatiotemporal dynamics of STML are governed by high-dimensional generalizations of the nonlinear Schrödinger equation (NLSE) or Lugiato-Lefever equation (LLE). For graded-index multimode fibers (GRIN-MMF), the multimode NLSE (MM-NLSE) or generalized multimode NLSE (GMMNLSE) governs the slow-varying envelopes $A_m(z,t)$ of the M guided modes: $$\frac{\partial A_m}{\partial z} = i [\beta_{0m} - \beta_{0R}] A_m - [\beta_{1m} - v_R^{-1}] \frac{\partial A_m}{\partial t} - \frac{i\beta_{2m}}{2} \frac{\partial^2 A_m}{\partial t^2} + i\!\sum_{n,p,q} \gamma_{mnpq} A_n A_p A_q^* + G_m(A) + \hat{F}(A)$$ Key terms include modal and chromatic dispersions ($\beta_{1m}, \beta_{2m}$), nonlinear intermodal mixing ($\gamma_{mnpq}$), saturable gain $G_m$, and spectral/spatial filtering $\hat{F}$. The transverse degrees are coupled via the spatial overlap of modal intensity profiles and by nonlocal nonlinearities (e.g., cross-phase modulation, modal four-wave mixing) (Wright et al., 2017, Wright et al., 2019).

In microresonators supporting multiple transverse families, the coupled LLE system for modal amplitudes $A_m(\theta, t)$ incorporates linear losses, detuning, modal dispersion, Kerr nonlinearity ($\gamma_m$), and intermodal couplings ($C_{mn}$), as well as external driving (Nie et al., 2022).

Kerr-lens and dissipative mechanisms in multimode fibers are captured by a generalized dissipative Gross–Pitaevskii equation (G-DGPE), balancing diffraction, graded-index spatial potential, spatially dependent loss, gain saturation, Kerr self-action, and spectral filtering (Kalashnikov et al., 2020). Topological photonic lattices supporting multi-timescale STML are modeled by driven–dissipative networks of coupled-mode equations with synthetic gauge phases, supporting nested mode-locked states with independently controlled repetition rates (Xu et al., 21 Feb 2025).

2. Experimental Architectures for STML

Multimode Fiber Lasers

Canonical STML fiber lasers employ a combination of:

• A gain section (Yb-doped step-index or few-mode fiber, supporting several modes);
• A multimode passive section (GRIN-MMF, supporting tens to hundreds of spatial modes/LP families);
• Filtering elements (narrowband spectral, spatial filter via pinhole or mode recoupling);
• Intracavity saturable absorber (nonlinear polarization rotation, Mamyshev configuration via offset spectral filtering);
• Splice offsets or lens coupling to tailor mode excitation and loss.

Cavity lengths typically correspond to tens of MHz repetition rates; pulse energies span 5–150 nJ; dechirped pulse durations can reach 100–200 fs; and >10 spatial modes can be phase-locked in a single pulse (Wright et al., 2017, Haig et al., 2021, Teğin et al., 2019, Gao et al., 2024).

Microresonator STML and Photonic Lattices

Ultrahigh-Q GRIN-MMF Fabry–Pérot microresonators enable STML of dissipative Kerr solitons (DKS), wherein co-propagating eigenmode combs are phase-locked via intermodal stimulated Brillouin scattering (SBS). By controlling the frequency offset and stress, one can select between eigenmode DKS or multi-mode STML DKS, with achieved fundamental linewidths as low as 400 mHz and sub-fs timing jitter (Nie et al., 2022).

On-chip topological photonic lattices, such as 2D arrays of coupled Si$_3$N$_4$ ring resonators, support edge-confined multi-timescale STML, synchronizing both fast (THz, single-ring FSR) and slow (GHz, edge super-ring) repetition rates. Mode-locking signatures include quadratic pump noise distributions and near-transform-limited RF repetition beat notes (Xu et al., 21 Feb 2025).

Spatiotemporal Vortex Pulse Generation

Spatiotemporal vortex pulse bursts, implementing independently controlled Laguerre–Gaussian (LG) or Hermite–Gaussian (HG) modes per comb tooth, have been generated via programmable spatial light modulators (SLMs) and tailored input-output shaping, producing ultrafast bursts with tunable orbital angular momentum (OAM) states (Liu et al., 2024).

3. Key Physical Mechanisms: Synchronization, Filtering, Stabilization

STML is realized through a dynamic balance among modal and chromatic dispersion, Kerr-induced nonlinear coupling, spatial and spectral filtering, and gain/loss saturation. Critical mechanisms include:

• Nonlinear intermodal coupling: Efficient intermodal four-wave mixing and cross-phase modulation synchronize the phases and group delays of both temporal and spatial modes (Wright et al., 2017, Haig et al., 2021).
• Spatial and spectral filtering: Intracavity spatial filters (fiber pinholes, recoupling, apertures) and spectral filters enforce spatiotemporal boundary conditions. Spatial filtering resets intermodal walk-off, favoring lower-order modes; spectral filtering restricts the pulse bandwidth and temporal chirp, stabilizing a three-dimensional dissipative soliton (Wright et al., 2019, Gao et al., 2024).
• Saturable absorbers: Nonlinear polarization rotation, intensity-dependent two-filter constructs (Mamyshev mechanism), or distributed Kerr-lens (graded dissipation) selectively amplify high-intensity, phase-synchronized pulses, suppressing noise and enforcing temporal AND spatial coherence (Kalashnikov et al., 2020, Gao et al., 2024).
• Attractor structure and stabilization: STML states constitute high-dimensional dissipative-soliton attractors. The combined action of saturable absorption and spatial filtering forms a “spatiotemporal stabilizer,” robust to both spatial and temporal perturbations, with the basin of attraction determined by pulse energy and modal condensation (Wright et al., 2019, Gao et al., 2024).

4. Experimental Metrics, Dynamics, and Validations

STML lasers and combs are characterized by:

• Pulse energy and duration: Achievable energies range from several nJ to >100 nJ per pulse. Typical intracavity pulse durations are in the range of 20 ps; external dechirping enables sub-200 fs pulses (Wright et al., 2017, Teğin et al., 2019).
• Spectral and spatial coherence: Interferometric autocorrelation ratios (up to 8:1), high RF contrast (>80 dB), and spatially resolved spectral measurements confirm phase-locked spatiotemporal operation involving >10, and up to $10^7$, transverse modes (Wright et al., 2019, Wright et al., 2017, Nie et al., 2022).
• Beam quality: Nonlinear multimode cavity dynamics and spatial filtering yield near-Gaussian beams (M² < 1.4) from fibers supporting >200 spatial modes (Teğin et al., 2019).
• Resilience to perturbations: Embedded SLMs in the cavity have experimentally demonstrated that STML states maintain mode composition even under strong spatial phase perturbations, a behavior not observed in unmode-locked states (Gao et al., 2024).
• Unique topological and spatiotemporal states: On-chip and microresonator platforms exhibit multi-timescale synchronization, quadratic noise signatures, and robustness associated with topological edge states (Xu et al., 21 Feb 2025).

5. Analytic Principles and Phase Regimes

Analytic frameworks for STML cavity dynamics leverage:

• Attractor dissection and minimum-loss principle: The round-trip cavity map can be dissected into dominant component operators (spatial filtering, saturable absorption, gain, etc.), predicting distinct phases (spatial-filter-dominated, absorber-driven, gain-absorber, SAGE) with associated stability/bifurcation diagrams (Wright et al., 2019).
• Distributed Kerr-lens mode-locking: Graded dissipation in conjunction with Kerr nonlinearity produces self-stabilized (2+1)-D dissipative solitons without requiring separate saturable absorbers, the spatial loss profile acting as a soft aperture (Kalashnikov et al., 2020).
• Stability and scaling laws: The maximum spatial perturbation tolerated before STML breakdown scales linearly with intracavity pulse energy, filtering strength, and saturable absorber modulation depth, guiding robust design (Gao et al., 2024).

6. Applications and Outlook

STML enables and enhances capabilities across multiple domains:

• High-peak-power ultrafast sources: Orders-of-magnitude scaling of pulse energy and coherent brightness possible in MMF-based and microresonator platforms (Wright et al., 2017, Nie et al., 2022).
• Low-noise frequency combs and photonic flywheels: STML DKS combs provide record sub-hertz linewidths and sub-femtosecond timing jitter for metrology, microwave photonics, and clockwork (Nie et al., 2022).
• Programmable spatiotemporal sculpting: Intracavity SLMs and STML protocols enable custom output beams, vortex bursts with tailored OAM, and high-fidelity transfer of multidimensional light through complex media (Cruz-Delgado et al., 2024, Liu et al., 2024).
• Topological synchronization and multi-timescale operation: Chip-based photonic lattices demonstrate independently tunable, multi-Tb/s parallelized light sources, with robustness and pattern formation not possible in single-resonator systems (Xu et al., 21 Feb 2025).
• New physical regimes: STML pulses probe three-dimensional wave turbulence, condensation, soliton molecules, and “Ising-machine” analog computation (Wright et al., 2019, Wright et al., 2017).

Anticipated research directions include environmental stabilization of multimode cavities, precise 3D field metrology, exploration of nonlinear–topological phase transitions, and ultrafast quantum state engineering with spatiotemporal entanglement.

7. Summary Table: Representative STML Platforms and Performance

Device/Platform	Modal Content	Pulse Energy / Linewidth	Key Features
MMF laser (GRIN, Yb gain) (Wright et al., 2017)	≳10–100 modes	5–150 nJ, 100–200 fs (dechirped)	3D dissipative solitons; nonlinear beam cleanup
MM Mamyshev oscillator (Haig et al., 2021)	∼20 modes	7–20 nJ, ∼300 fs (post-comp.)	Switchable spatiotemporal coupling via offset
Microresonator (GRIN-MMF FP) (Nie et al., 2022)	2–5 modes	Comb linewidth 400 mHz	Attosecond timing jitter; STML DKS
On-chip lattice (Si$_3$N$_4$) (Xu et al., 21 Feb 2025)	100s of rings	Multi-timescale ($>1$ THz/GHz)	Topological edge, nested combs
Vortex burst synthesis (Liu et al., 2024)	Multi-OAM (LG)	N/A	Arbitrary ST vortex design, pulse-by-pulse OAM

This collective body of research demonstrates that spatiotemporal mode-locking is a unifying, high-dimensional generalization of ultrafast mode-locked laser science, enabling robust, programmable, and high-power light sources with complex three-dimensional coherence (Wright et al., 2017, Wright et al., 2019, Gao et al., 2024, Nie et al., 2022, Xu et al., 21 Feb 2025, Liu et al., 2024, Teğin et al., 2019).