Temporal wave trapping from dynamical pump pulses

Technical Summary: Temporal Wave Trapping from Dynamical Pump Pulses

Problem Statement
Temporal reflection in nonlinear optical fibers offers a framework for manipulating light by utilizing moving refractive-index perturbations as time-dependent boundaries. While previous experimental demonstrations of temporal reflection and optical horizons have relied on ultra-short pump pulses (< 1 ps) to create strong instantaneous intensity gradients, or required complex multi-soliton configurations, there remains a need for mechanisms that can achieve these effects using longer, more practical pulse durations. Specifically, the generation of trapped waves—where probe components are confined within a moving potential—has typically been associated with femtosecond pulses or specific multi-pulse arrangements. The challenge addressed in this work is to demonstrate a mechanism for temporal wave trapping induced by the intrinsic dynamical evolution of a single, few-picosecond high-order soliton pulse, thereby bypassing the strict requirement for ultra-short pulses or multiple solitons.

Methodology
The authors employ a combined theoretical and experimental approach to investigate the interaction between a continuous weak probe field and a dynamical high-intensity pump field in a nonlinear dispersive medium.

• Theoretical Modeling: The propagation is modeled using the generalized nonlinear Schrödinger equation (GNLSE) for the pump field and a linearized equation for the probe field in the presence of the pump's cross-phase modulation (XPM). To gain physical insight, the authors apply the eikonal approximation, deriving a Hamiltonian system where the probe rays follow trajectories determined by the pump's intensity gradient. This allows for the tracing of reflected, transmitted, and trapped rays based on the pump's temporal evolution.
• Numerical Simulation: The GNLSE is solved numerically for a dispersion-shifted fiber (DSF) with specific dispersion parameters ($\beta_2 \approx -1.52$ ps$^2$/km, $\beta_3 \approx 0.12$ ps$^3$/km, $\beta_4 \approx -0.0006$ ps$^4$/km) at a pump wavelength of 1560 nm. Simulations explore input pump powers corresponding to third and fourth-order solitons, tracking the temporal compression and subsequent formation of bipulses over a 5-km fiber length.
• Experimental Setup: Experiments were conducted using a 5-km-long DSF. The pump source was a mode-locked fiber laser delivering nearly sech$^2$ pulses with a full width at half maximum (FWHM) of approximately 6.7 ps at 1560 nm. A continuous wave (CW) probe, tunable between 1520 nm and 1535 nm, was combined with the pump. The pump power was varied to control the soliton order and the resulting dynamical evolution (single compression vs. compression followed by bipulse formation).

Key Contributions

1. Novel Trapping Mechanism: The paper demonstrates a fundamentally different mechanism for wave trapping mediated by a single high-intensity pulse through its intrinsic dynamical evolution, rather than relying on pre-existing multiple solitons or ultra-short pulses.
2. Bipulse-Induced Trapping: The authors identify that the formation of a "bipulse" (a soliton fission event) during the propagation of a higher-order soliton creates a dynamic "soliton cage." This structure traps probe waves via multiple temporal reflections, a phenomenon previously observed only with femtosecond pulses.
3. Validation of Long-Pulse Dynamics: The work validates that pump pulses with durations of a few picoseconds (specifically ~6.7 ps) are sufficient to generate the necessary intensity gradients for temporal reflection and wave trapping, provided the pulse undergoes significant dynamical compression.

Results

• Numerical Findings: Simulations revealed two distinct regimes based on input power. At lower powers ($P_0 \approx 600$ mW), the pump undergoes a single temporal compression stage, generating a temporally reflected wave (red-shifted) but no trapping. At higher powers ($P_0 \approx 850$ mW), the pump compresses and subsequently splits into a bipulse. In this regime, the probe field exhibits three components: a transmitted wave, a temporally reflected wave, and a trapped wave component oscillating between the two parts of the bipulse.
• Experimental Verification: Measurements in the 5-km DSF confirmed the theoretical predictions.
  • Spectral Signatures: As pump power increased, the output spectrum evolved from showing only the input probe to displaying a blue-shifted peak (temporal reflection) and, at higher powers, a third peak at the group-velocity-matched (GVM) wavelength (trapped wave).
  • Ray Tracing Agreement: The experimental spectral features aligned closely with the ray trajectories calculated from the Hamiltonian model, confirming that reflection occurs at maximum compression and trapping occurs within the bipulse structure.
  • Wavelength Dependency: Scanning the probe wavelength revealed a characteristic "cross-like" pattern in the frequency conversion maps. The diagonal branches corresponded to temporal reflection, while the center corresponded to the GVM trapped wave. The experimental maps showed asymmetry and substructures (spectral dips) that were qualitatively reproduced by numerical simulations, attributed to finite propagation distances and internal pump dynamics.

Significance
The authors claim that these results establish a simple and versatile route toward dynamic temporal waveguiding using a single optical pulse. By leveraging the natural compression and reshaping of higher-order solitons, this approach overcomes the limitations of requiring ultra-short pulses (< 1 ps) or complex multi-soliton configurations. The work demonstrates that a single, few-picosecond pump pulse can induce both temporal reflection and group-velocity-matched trapping of a continuous weak probe. The authors suggest this opens new opportunities for the all-optical control and manipulation of ultrafast signals, with potential applications in telecommunications and quantum optics, though they do not specify detailed implementation strategies beyond these general fields.