Light Conversion in Waveguides

In nonlinear waveguides, laser light can be converted nonlinearly with respect to its spectral and spatial composition. In four-wave mixing (FWM) the spectral or spatial distribution of the laser light is changed by a nonlinear medium. In single-mode waveguides, i.e., those that only support a transverse mode, very broadband spectra - so-called supercontinua - can be generated using FWM with ultrashort laser pulses of around 100 fs pulse duration. In addition, longer pulses of around 1 ps pulse duration can be used to build so-called optical parametric oscillators (OPOs), which can efficiently convert laser light of one frequency into another frequency. In multi-mode waveguides, the energy of the laser light can be converted in a nonlinear way, for example from one transverse mode to another.

In multi-mode waveguides, pulses in the different modes can influence each other and thus generate spectral components that would not occur if the modes were excited separately. In particular, the spectral composition of a second mode can be changed during supercontinuum generation in one mode.

Optical parametric oscillators based on tantalum pentoxide and silicon nitride waveguides

In a recent publication, we demonstrated a proof-of-concept of a hybrid waveguide-fiber optical parametric oscillator (WOPO) exploiting four-wave mixing in tantalum pentoxide waveguides. The results indicate that a full integration of the oscillator as a compact and robust light source on a chip is possible. See the publication by Max.

Compared to Ta₂O₅ waveguides, Si₃N₄ offers very low propagation losses with similarly high nonlinearity, which also makes it suitable as a platform for four-wave mixing. We have recently presented a Si3N4-based, very efficient and widely tunable WOPO, which could potentially be used in microscopy for coherent anti-Stokes Raman scattering, for example. See the publication by Ming.

Intermodal generation of dispersive waves

During supercontinuum generation in waveguides by means of soliton dynamics, so-called dispersive waves (DW) arise when a strong pulse becomes so spectrally broad that it can radiate phase-matched energy into a linear wave - the DW (see Fig. a). The pulse itself generates this spectral bandwidth through self-phase modulation, whereby such a pulse can also broaden the spectrum of another (temporally overlapping) pulse via so-called intermodal cross-phase modulation, although the latter propagates in an orthogonal mode. In this case, a weak pulse, which itself does not have enough energy to generate a supercontinuum (see Fig. b), can also emit a dispersive wave (see iDW in Fig. d). This effect, called intermodal dispersive wave generation (iDWG), could be used to broaden the spectral bandwidth or efficiently convert frequencies.

Intermodal cross-phase modulation is central to the iDWG. This means that the spectral broadening of the weak laser pulse by the strong pulse alone ensures that the weak pulse emits a dispersive wave. There is no transfer of photons from one pulse to the other, but energy can be transferred by the frequency conversion itself. In the time image, the iDWG can be understood as the inelastic scattering of two pulses, whereby the strong pulse represents a temporal refractive index barrier, generated by the Kerr effect, for the weak pulse, on which the latter scatters inelastically. See the publication by Niklas and Max.