Merging the fields of plasmonics and nonlinear optics authorizes a variety of fascinating and original physical phenomena. In this study, we specifically study the combination of the strong light confinement ability of surface plasmon polaritons (SPP) with the beam self-trapping effect that occur in nonlinear optical Kerr medium. Although this idea of plasmon-soliton has been the subject of several theoretical or numerical articles, no experimental evidence has been revealed yet. One reason is that in the proposed configurations the requested nonlinear refractive index change amplitude to generate a plasmon-soliton is too high to be reached in available material. Another limitation is due to the large propagation losses associated with plasmons. In the present study, a proper architecture has been designed and then fabricated allowing the first experimental observation of hybrid coupling between a spatial optical soliton and a SPP in a metal-Kerr dielectric structure.
To be able to trigger the nonlinearity at moderate light power and simultaneously to allow propagation over several millimetres distance, a metal-dielectric structure was designed. It consists of a four-layer planar geometry made of a transparent Kerr dielectric layer placed on a lower refractive index medium, with on its top surface a thin dielectric layer covered by a metallic film deposited on top. The Kerr medium is a 3µm thick chalcogenide film (Ge28.1Sb6.3Se65.6) with a high refractive index deposited by RF magnetron sputtering on an oxidized silicon substrate. The thickness of the thin SiO2 layer is 10 nm while the top gold layer is 30 nm. Samples are about 5-6 mm along propagation direction (z-axis).
As shown by numerical simulations, the designed planar nonlinear waveguide with its top silica and gold layer supports a fundamental TE mode profile at NIR wavelengths whose transverse profile along y (perpendicular to the layers) is not affected by the metal layer while the TM mode is clearly localized near the SiO2-metal-chalcogenide interfaces due to its plasmonic part. The estimated nonlinear parameter γ of the TM mode is nearly three times larger than the TE one. Consequently, in nonlinear regime an enhanced self-focusing effect is expected for this TM wave. Experiments are performed with a tunable optical parametric oscillator emitting 200 fs pulses at 1.55 µm with a repetition rate of 80 MHz. The experimental analysis consists in injecting a typical 4 × 30 μm2 (FWHM in x-y cross section) elliptical laser beam into the waveguide and monitoring the output beam spatial profile evolution versus light power. Different arrangements are tested that unambiguously reveal the plasmon-soliton coupling. For instance, experiments are conducted with and without the metallic layer and for both TE and TM polarizations. In addition, different positions on the sample of the metal part with several lengths chosen between 0.1 to 2mm are tested. Additional experiments are in progress to analyze the beam evolution with near-field scanning microscopy and simulations of the beam propagation in the full structure are developed to reach a better and fully quantitative description of the observed phenomena.
We present a new experimental technique based on the analysis of beam self-action to measure optical nonlinearity in planar waveguides. This technique is applied to analyze the nonlinear properties of slab chalcogenide waveguides that can develop Kerr induced self-focusing or self-defocusing, depending upon the waveguide structure and composition. Optical nonlinearity in chalcogenide waveguide is studied in the 1200 nm to 1550 nm wavelength range in femtosecond regime. Results of the proposed technique compare favorably with n<sub>2</sub> values obtained with the Z-scan technique. In addition, beam self-trapping in the chalcogenide waveguides due to material photosensitivity is also observed.