Self-referencing Nonlinear Frequency Converting Photonic Waveguide and Self-referencing Nonlinear Frequency Conversion

patent description

The unique frequency accuracy of a frequency comb stems from “f-to-2f” self-referencing where an optical octave is generated and the low frequency edge of that octave is doubled and compared to the high frequency end to measure the carrier envelope offset frequency, the key piece that allows a frequency comb to act as a bridge between radio frequencies and optical frequencies.

The principle cost drivers for a commercial frequency comb are the pump diodes (required to make a high-power pulse that will form the octave) and the waveguide doubler (typicall ppLN or ppKTP required for the frequency doubling). Both ppLN and ppKTP have remained stubbornly expensive materials for the last two decades

The attraction of AlGaAs waveguides is that they can generate and octave of light with 1/100th the power required by traditional methods, thus eliminating the cost burden of the pump diodes. While there will be some cost in fabricating and packaging the waveguides, they will add no additional cost to the system since a properly designed waveguide can also act as the waveguide doubler. We expect that this reduction in component cost and overall system complexity will be of significant interest to a number of parties commercializing frequency combs.

Specifically, the invention identifies AlGaAs nanophotonic waveguides as a superior material to use in these waveguides, with specific geometries to enhance the efficiency of the process.  The invention consists of the following features:

(1)         A nanophotonic waveguide made of AlGaAs integrated on a chip/substrate

(2)         On at least some of the total propagation length of the waveguide, the waveguide is dispersion-engineered to enhance supercontinuum generation from a pulsed pump laser (ex. 1560 nm) such that light is nonlinearly broadened out to an octave (ex. 900-1800 nm).  This can be achieved by tailoring the width of the waveguide appropriately, or by utilizing orientation patterning (to periodically invert the crystal domain orientation of AlGaAs) to achieve cascaded second-order nonlinearities.

(3)         For the remainder of the waveguide’s propagation length following the previous section designed for supercontinuum generation, it is engineered to enhance second-harmonic generation, by modulating the structure in such a way as to neutralize the momentum mismatch between the fundamental and second-harmonic light (ex. 1800 nm light and 900 nm light).  This can be achieved by various means, such as form-birefringent phase matching, mode-shape modulation (varying the width of the waveguide periodically), or orientation patterning with the correct period. 

(4)         Alternatively, the waveguide’s geometry can be designed for supercontinuum generation (to achieve the desired dispersion) for the entire length, and additionally is orientation-patterned with a constant periodicity for the entire length, the period being chosen to phase-match the fundamental and second-harmonic waves.  In this way, supercontinuum generation and second-harmonic generation proceed simultaneously. 

(5)         At the facets or ends of the chip containing the AlGaAs nanophotonic waveguide, edge couplers are implemented such that on the input, the coupling efficiency of the pump wavelength is enhanced, and that on the output, the coupling efficiency of the second-harmonic light is enhanced.  These edge couplers may consist of inversely tapered nanophotonic waveguides to expand the mode field diameter, providing better coupling to free-space or fiber-optic modes.

(6)         To minimize absorption losses at short wavelengths in the vicinity of 800-1100 nm wavelengths, the cross-sectional geometry may consist of the following arrangement: a silicon substrate with a thick layer of silicon dioxide on the surface (ex. 3 um), the AlGaAs waveguide having a thickness between 130 and 400 nm, and no material on top (commonly referred to as an air-top-cladding).  This also can enhance the efficiency of the second-harmonic generation process at the same time.

(7)         In another embodiment, the AlGaAs waveguide may be surrounded by air on both the top and bottom, referred to as a suspended or fully air-clad device.  It is tethered to the substrate via a thin slab of AlGaAs.  Refer to the attached manuscript for details on this geometry and experimental progress in this direction.  This geometry has the advantage of very high index contrast, low coupling losses, and optical transparency out into the mid-infrared, if it is desired to implement this invention at longer wavelengths.