The research team, comprising Hamid Nadgaran, Mohammad Amin Izadi, and Rahman Nouroozi, built their design around a hybrid plasmonic waveguide loaded with periodically poled lithium niobate, or PPLN. Lithium niobate is prized in photonics for its strong second-order nonlinearity, which enables a process called difference-frequency generation, or DFG. In DFG, an intense pump pulse transfers energy to a weaker signal pulse, boosting its power while, under the right conditions, squeezing its duration dramatically. The catch has always been efficiency: conventional titanium-diffused PPLN waveguides have cross-sections measured in tens of micrometers, which dilutes the optical fields and demands punishing pump powers.

In their earlier theoretical work, the same group showed that a conventional Ti:PPLN waveguide could regenerate a 1550-nanometer signal pulse that had broadened to 300 picoseconds and lost over 90 percent of its energy after traversing 50 kilometers of standard single-mode fiber. The results were impressive, compressing the pulse to 1.5 picoseconds with 10.9 decibels of amplification, but they came at a steep price: a pump pulse of roughly 15 watts of peak power, a 20-millimeter interaction length, and an exotic fifth-order super-Gaussian pump shape that would require elaborate pulse-shaping optics such as spatial light modulators to synthesize in the laboratory.

To break through these limits, the team turned to surface plasmon polaritons, electromagnetic waves that ride the boundary between a metal and a dielectric and squeeze light into volumes far smaller than any purely dielectric waveguide can achieve. Bare plasmonic structures, however, suffer from brutal ohmic losses: when the researchers modeled a lithium niobate channel sitting directly on silver, the pump mode’s attenuation reached 43.8 per millimeter, meaning the pump energy would vanish within a microscopic distance and render the device useless even with a 20-watt pump. The solution was a hybrid architecture, in which an ultrathin low-index spacer separates the nonlinear core from the metal, pushing the optical field away from the absorptive silver surface while preserving deep-subwavelength confinement.

The choice of spacer material proved decisive. Common photonic dielectrics like silica and magnesium fluoride have refractive indices too high for this geometry, allowing too much field to accumulate near the metal. The team instead selected Teflon AF 2400, an amorphous fluoropolymer whose refractive index approaches that of vacuum and which is compatible with nanoscale thin-film deposition. After a systematic optimization sweep varying spacer thickness, channel width, and channel height, the final design emerged: a PPLN core measuring 450 by 600 nanometers, an 80-nanometer Teflon AF 2400 spacer, 300-nanometer-wide Teflon sidewalls for tight lateral confinement, and a monolithic silica top-cladding for mechanical robustness, all resting on a 100-nanometer silver substrate.

Finite-difference simulations of this optimized geometry yielded a nonlinear coupling coefficient of 841.78 per square-root-watt per meter, a thirteenfold enhancement over the 63 achieved in conventional Ti:PPLN waveguides. The confinement statistics are striking: 96 percent of the 775-nanometer pump power resides inside the lithium niobate core, with a mere 0.0048 percent penetrating the silver, a 99.8 percent reduction compared with the bare plasmonic case. At 1550 nanometers, the signal propagates as a hybrid surface plasmon polariton, with 65 percent of its power confined in the nonlinear core and only 0.023 percent reaching the metal. This delicate balance between confinement and loss is what makes the device viable.

The payoff appears in the pulse-processing simulations. The team fed their waveguide a signal pulse that had degraded over 50 kilometers of ITU-T G.652 fiber, arriving broadened to 300 picoseconds, chirped, and stripped of 90 percent of its energy. Driven by a simple Gaussian pump pulse with a peak power of just 6 watts and a 25-picosecond duration, the 5-millimeter device compressed the signal to 8.7 picoseconds, a 97.1 percent reduction in duration, while regenerating its peak power to 4 milliwatts, four times the power of the original transmitted pulse. The energy conversion efficiency reached a factor of 30, and the process simultaneously generated a compressed idler pulse at 1560.13 nanometers, exploiting the phase-conjugate relationship between signal and idler to achieve compression that is independent of the input chirp.

Crucially, all of this was achieved with an ordinary Gaussian pump, the natural output of standard mode-locked lasers, eliminating the need for complex external pulse shaping. The team also verified that parasitic third-order nonlinearities, a known hazard of plasmonic systems given silver’s large nonlinear susceptibility, remain negligible in their design. Because the low-index spacer confines the fields so effectively, the self-phase modulation on the signal amounts to roughly 23.72 millionths of a radian and the cross-phase modulation from the pump to 0.125 radians, both far below the 1-radian threshold at which pulse distortion becomes problematic. The pump intensity inside the core, about 2.8 gigawatts per square centimeter, also sits safely below lithium niobate’s picosecond damage threshold, which routinely exceeds 10 gigawatts per square centimeter.

The improvements over the group’s previous design are substantial on every axis: peak pump power cut by roughly 60 percent, device length reduced by 75 percent, and pump complexity eliminated altogether. The authors caution that real-world fabrication will bring challenges, including lithographic dimensional drift, scattering from metal surface roughness at the silver-Teflon interface, and the demanding requirement of sub-micron periodic poling in thin-film lithium niobate. The extreme nanoscale mode also mismatches standard single-mode fibers, so integrated adiabatic mode converters such as inverse tapers would be needed for efficient coupling. Even so, the 4-milliwatt output provides a generous power margin over the microwatt sensitivities of standard telecom receivers, leaving ample optical budget to absorb coupling and insertion losses.

If experimentalists can realize the design, the implications extend beyond telecom signal regeneration. A platform that combines the strong second-order nonlinearity of PPLN with plasmonic field enhancement in a five-millimeter footprint could serve as a building block for compact wavelength converters, optical phase conjugation modules, and even quantum frequency conversion on chip. For now, the work stands as a rigorous theoretical demonstration that the decades-old trade-off between nonlinear efficiency and device size can be broken, not by choosing between dielectrics and metals, but by engineering the space between them with nanometer precision.

Subject of Research: Simultaneous pulse compression and amplification via difference-frequency generation in an optimized hybrid PPLN-loaded plasmonic waveguide

Article Title: Theoretical investigation of efficient simultaneous pulse compression and amplification in an optimized hybrid PPLN-loaded plasmonic waveguide

Article References: Nadgaran, H., Izadi, M. A., & Nouroozi, R. (2026). Theoretical investigation of efficient simultaneous pulse compression and amplification in an optimized hybrid PPLN-loaded plasmonic waveguide. Results in Optics, Article 101170. https://doi.org/10.1016/j.rio.2026.101170

Image Credits: AI Generated

DOI: 10.1016/j.rio.2026.101170

Keywords: plasmonics, lithium niobate, PPLN, difference-frequency generation, pulse compression, optical amplification, hybrid waveguide, surface plasmon polariton, optical communications, nonlinear optics, Teflon AF 2400, signal regeneration