Patent Grant
9915852
The present invention relates to the field of optics. More particularly, the present invention relates to methods and systems for Raman converters, parametric converters and parametric amplifiers with cavity enhancement and with quasi-phase-matching.
Over the past several decades there has been growing interest in the development of devices based on second-order nonlinear effects such as sum-frequency generation (SFG) and difference frequency generation (DFG) and devices based on third-order nonlinear effects such as Raman-resonant four-wave-mixing (FWM) and Kerr-induced four-wave-mixing (FWM). SFG, DFG and Kerr-induced FWM are parametric light-matter interactions that are not resonant with a material level and that are used in parametric converters and parametric amplifiers. Raman-resonant FWM is a light-matter interaction that is perfectly resonant or almost perfectly resonant with a characteristic energy level of the material such as a vibrational energy level and that is used in Raman converters. SFG, DFG and Kerr-induced FWM involve a pump radiation beam at frequency ωp, a signal radiation beam at frequency ωs, and an idler radiation beam at frequency ωi. Raman-resonant FWM involves a pump radiation beam at frequency ωp, a Stokes radiation beam at frequency ωstrokes that is lower than the pump frequency, and an anti-Stokes radiation beam at frequency ωanti-stokes that is higher than the pump frequency. One also uses the terms signal and idler for the Stokes and anti-Stokes radiation beams, respectively, or vice versa, and uses ωs and ωi to denote their frequencies. Due to the wavelength versatility offered by SFG, DFG, Raman-resonant FWM and Kerr-induced FWM, these processes feature a multitude of application possibilities in different domains such as optical communication and spectroscopy.
Basically, Raman-resonant FWM and Kerr-induced FWM are interactions between two pump photons, one signal photon and one idler photon, and the frequencies of these photons ωp, ωs and ωi satisfy the relation 2ωp−ωs−ωi=0. For Raman-resonant FWM one has in addition that |ωp−ωs|=2π×ΔR, with ΔR being the Raman shift of the considered Raman-active material. In the case of SFG and DFG there is an interaction between 1 pump photon, one signal photon and one idler photon, and the frequencies of these photons ωp, ωs and ωi satisfy the relation ωp+ωs=ωi fpr SFG and ωp−ωs=ωi for DFG. The efficiency of all processes depends on the pump intensity and on the processes' phase mismatch. The linear part Δklinear of the phase mismatch for Raman-resonant FWM and Kerr-induced FWM is given by
Δklinear=2kp−ks−ki
where k{p,s,a}=ω{p,s,a}xn{p,s,a}/c are wave numbers with n{p,s,a} representing the effective indices of the pump, signal and idler waves, respectively. One can also write Δklinear as
where β2=d2k/dω2 is the group velocity dispersion (GVD) at the pump wavelength, β4=d4k/dω4 is the fourth-order dispersion at the pump wavelength, and Δω the frequency difference between the pump and signal waves. For SFG the linear part Δklinear of the phase mismatch is given by
Δklinear=kp+ks−ki
For DFG the linear part Δklinear of the phase mismatch is given by
Δklinear=kp−ks−ki
The total phase mismatch for these processes also contains a nonlinear part that is function of the pump intensity, but since linear phase mismatches are considered here that are mostly much larger than the nonlinear part of the phase mismatch, the latter can be neglected in the remaining part of this text.
Due to their nonlinear nature, all above-mentioned processes perform best at high optical intensities. These can be obtained by tightly confining the light for example in a nanowire waveguide and also by employing ring structures, whispering-gallery-mode disk resonators, or any other resonator structure in which the incoming light waves are resonantly enhanced. Regarding the requirement of having a small effective phase mismatch for the wavelength conversion processes, much progress has been made over the past several years, in particular for converters based on silicon waveguides. For these converters, by engineering the dispersion of a silicon nanowire waveguide one can obtain phase-matched Kerr-induced FWM in the near-infrared for pump-signal frequency shifts with an upper limit of 52 THz (i.e., pump-signal wavelength differences up to 418 nm in the near-infrared region).
Notwithstanding the broad applicability of this phase-matched conversion technique, there are circumstances, applications, and materials where an alternative approach can be useful. First of all, not all materials used for SFG, DFG, Raman-resonant or Kerr-induced FWM are as easily workable as silicon to fabricate waveguide structures, which implies that not all materials can benefit from the waveguide-based phase-matching technique outlined above. Furthermore, even if one considers only a material such as silicon for which the waveguide-based phase-matching technique described above is well developed, it is important to know that, although the phase-matching bandwidth of the silicon nanowire referred to above is more than wide enough to enable phase-matched Raman-resonant FWM in the near-infrared at a pump-signal frequency shift of 15.6 THz, the dispersion-engineered geometry of the waveguide is such that does not comply with the fabrication constraints of multiproject-wafer-oriented silicon photonics foundries, which rely on conventional semiconductor processing technology and employ a standard waveguide thickness of typically 220 nm. Since the use of such foundries can pave the way to the large-volume fabrication of integrated photonic components at low cost, adhering to these foundry standards is a natural strategy to exploit the full potential of silicon photonic devices. Furthermore, the developed silicon-based converters often are not compact and require propagation distances of at least 1 cm to achieve substantial conversion efficiencies. This leads to device footprints that are too large for realizing cost-effective photonic integrated circuits. Also for SFG and DFG in silicon nanowires, there are dispersion-engineering issues. To establish SFG and DFG in silicon, one usually applies strain on a silicon nanowire to induce the second-order nonlinearity that is needed for these processes and hence make the nanowire a quadratically nonlinear optical medium. Because of the very large pump-signal frequency shifts typically used in SFG and DFG, it is practically impossible to engineer the dispersion of the strained silicon nanowires in such a way that phase-matched SFG or DFG is obtained. Hence, it is challenging to achieve efficient SFG or DFG using only dispersion engineering.
One suggestion has been to establish quasi-phase-matching for SFG, DFG, Raman-resonant FWM or Kerr-induced FWM by periodically modifying the material properties within the medium through which the light propagates. This traditional quasi-phase-matching technique for these nonlinear processes can be understood as follows: In case nothing is done about the phase mismatch, the idler intensity for radiation would continuously oscillate along the propagation path between a maximal value and zero, as the phase-mismatch-induced dephasing of the fields—this dephasing evolves periodically with the propagation distance—causes the nonlinear optical processes to either increase or decrease the idler intensity along the propagation path. When using traditional quasi-phase-matching for these processes, one adjusts the propagation regions behind the positions of maximal idler intensity, so that one does not have a total drop down of the idler intensity in these regions but at the same time the fields' dephasing, accumulated up to the positions of maximal idler intensity, can evolve back to zero in these adjusted regions. Hence, after traversing these adjusted areas the idler intensity can start growing again towards a maximum. The type of “adjustment” that needs to be applied to these propagation regions is that the susceptibility should be reversed in sign there for the Raman-resonant or Kerr-induced FWM or for the SFG and DFG processes, so that these nonlinear processes cannot establish a decrease of the idler intensity in these areas whereas the fields' dephasing can still evolve back to zero. This sign reversal is the ideal case; if this is not possible, then quasi-phase-matching can also be obtained by making the susceptibility zero in the “adjustment” regions. Both the latter type of “adjustment” and the ideal “adjustment” of susceptibility sign reversal is traditionally implemented using a conversion medium where the material properties within the medium are periodically manipulated. This is a complex approach and disadvantageous from a practical point of view. Furthermore, like the dispersion-engineered phase matching approach described earlier, this approach typically yields too large device footprints to realize cost-effective photonic integrated circuits.
It is an object of embodiments of the present invention to provide efficient methods and systems for nonlinear optical wave mixing, such as for example SFG, DFG and FWM, e.g. Raman-resonant FWM and/or Kerr-induced FWM. It is an advantage of at least some embodiments of the present invention that efficient nonlinear optical wave mixing, such as for example SFG, DFG and FWM, e.g. Raman-resonant FWM and/or efficient Kerr-induced FWM, can be obtained at wavelengths suitable for use in e.g. telecommunication and spectroscopy.
Nonlinear optical wave mixing may in one example be a third order process, such as four wave mixing, or in another example be a second order process, such as three wave mixing.
It is an advantage of embodiments of the present invention that efficient systems are obtained, which can be made using conventional semiconductor processing technology.
It was surprisingly found that by using a two-dimensional material or quasi-two-dimensional material on top of the three-dimensional medium through which the light propagates, efficient systems for obtaining nonlinear optical wave mixing are obtained.
It is an advantage of embodiments of the present invention that adding a two-dimensional material or quasi-two-dimensional material on top of a uniform three-dimensional medium a better device compactness and smaller device footprints can be obtained.
It is an advantage of at least some embodiments of the present invention, e.g. for embodiments wherein a spiral based converter is used, that quasi-phase-matching can be obtained not just for one single output wavelength but for a band of output wavelengths.
It is an advantage of at least some embodiments of the present invention that efficient closed structure Raman converters such as e.g. ring or disc Raman converters, efficient closed structure parametric converters such as e.g. ring or disc parametric converters, and efficient closed structure parametric amplifiers, e.g. ring or disc parametric amplifiers are provided as well as corresponding methods. It is an advantage of at least some embodiments according to the present invention that methods and systems are provided that allow obtaining relatively high efficiencies using cavity-enhanced quasi-phase matched wave mixing such as cavity-enhanced quasi-phase-matched DFG, cavity-enhanced quasi-phase-matched SFG, cavity-enhanced quasi-phase-matched Raman-resonant FWM and/or using cavity-enhanced quasi-phase-matched Kerr-induced FWM in a resonator. Such cavity-enhanced quasi-phase matched wave mixing may in one example be a third order process, such as four wave mixing, or in another example be a second order process, such as three wave mixing.
It is an advantage of at least some embodiments according to the present invention that cavity-enhanced quasi-phase-matched wave mixing, such as cavity-enhanced quasi-phase-matched SFG, cavity-enhanced quasi-phase-matched DFG, cavity-enhanced quasi-phase-matched Raman-resonant FWM, cavity-enhanced quasi-phase-matched Kerr-induced FWM, in a resonator structure, e.g. ring or disc resonator, can provide both a small effective phase mismatch and a high pump intensity, resulting in a boost of the nonlinear optical wave mixing efficiency, e.g. SFG efficiency, DFG efficiency, Raman-resonant FWM efficiency and /or of the Kerr-induced FWM efficiency. This is especially advantageous for those cases where phase-matched wave mixing, e.g. phase-matched SFG, DFG, Raman-resonant FWM or Kerr-induced FWM performs badly, i.e. in the cases where the group velocity dispersion at the pump wavelength is large and/or the frequency difference between the pump and the signal is large.
It is an advantage of embodiments according to the present invention that efficient nonlinear optical wave matching, e.g. efficient SFG, efficient DFG, efficient Raman-resonant FWM or efficient Kerr-induced FWM can be obtained for any value of the linear phase mismatch Δklinear.
It is an advantage of embodiments according to the present invention that no dispersion engineering of the medium where the light propagates is required to obtain efficient SFG, efficient DFG, efficient Raman-resonant FWM or efficient Kerr-induced FWM.
It is an advantage of embodiments according to the present invention that it is not necessary to use only materials out of which waveguides can be fabricated and that other structures such as whispering-gallery-mode discs and discs with a rim polished according to a polygone shape can also be used.
It is an advantage of at least some embodiments according to the present invention that when working with a waveguide its geometry can be chosen freely without dispersion engineering constraints, and that the waveguide dimensions can be chosen such that they can be fabricated using conventional semiconductor processing technology.
It is an advantage of embodiments according to the present invention that suitable conditions for quasi-phase matched wave mixing, such as for quasi-phase-matched SFG, for quasi-phase-matched DFG, for quasi-phase-matched Raman-resonant FWM and for quasi-phase-matched Kerr-induced FWM are obtained using a uniform three-dimensional material. More particularly, it is an advantage that the obtained system and method is relatively simple and does e.g. not require active periodical adaptation of the nonlinear optical properties within the three-dimensional medium. It is an additional advantage of at least some embodiments according to the present invention that also the two-dimensional or quasi-two dimensional material top layer is uniform.
It is an advantage of the embodiments, where a two-dimensional or quasi-two-dimensional material layer is used which is patterned or which is locally chemically or electrically modified, that the patterning or local chemical modification or local electrical modification can be implemented along pie shapes that have a high tolerance for spatial fabrication deviations.
It is an advantage of at least some embodiments of the present invention that efficient, single-pump Kerr-induced FWM can be obtained for a large pump-signal frequency shift in a spectral domain where the dispersion characteristics of the medium where the light propagates are not optimally engineered for phase-matched Kerr-induced FWM.
It is an advantage of at least some embodiments according to the present invention that a high pump intensity in the Raman converter, in the parametric converter, and in the parametric amplifier does not need to be provided using a high-power pump, but that the pump is resonantly enhanced in the medium for obtaining a sufficiently high pump power. Alternatively or in addition thereto, the signal input power, and the idler power also can be resonantly enhanced regardless the value of the linear phase mismatch Δklinear resulting in high intensities being achieved. In other words also the signal input power initially injected does not need to be a high signal input power pump.
The above objective is accomplished by a method and device according to the present invention.
The present invention relates to a system for conversion or amplification using quasi-phase matched nonlinear optical wave mixing, the system comprising a first radiation source for providing a pump radiation beam, a second radiation source for providing a signal radiation beam, and a bent structure for receiving the pump radiation beam and the signal radiation beam, wherein a radiation propagation portion, e.g. waveguiding portion, of the bent structure comprises a uniform three-dimensional material at least partly covered by a two-dimensional or quasi-two-dimensional material layer and wherein the bent structure comprises a dimension taking into account the spatial variation of the nonlinear susceptibility along the radiation propagation portion as experienced by radiation travelling along the bent structure for obtaining quasi-phase-matched nonlinear optical wave mixing in the radiation propagation portion, the dimension being substantially inverse proportional with the linear phase mismatch for nonlinear optical wave mixing and an outcoupling radiation propagation portion, e.g. waveguiding portion, for coupling out an idler radiation beam generated in the bent structure. It is an advantage of embodiments according to the present invention that quasi-phase matching conditions can be achieved.
The two-dimensional material layer or quasi-two-dimensional material layer may be inducing the quasi-phase matched wave mixing. Where in embodiments of the present invention reference is made to a two-dimensional material or quasi-two-dimensional material, reference is made to a material that consists of a single layer of atoms, or several layers of atoms. In case of several layers of atoms, it is possible that the outer layers have different properties than the inner layers.
The two-dimensional material layer or quasi-two-dimensional material layer may comprise one or a combination of graphene, graphyne, borophene, germanene, silicene, stanine, phosphorene, metals, 2D supracrystals, hexagonal boron nitride, germanane, nickel HITP, transition metal di-chalcogenides (TMDCs), MXenes, black phosphorus, or topological insulators.
The three-dimensional material may be any or a combination of silicon, germanium, GaAs, InGaAs, diamond, cadmium telluride, gallium indium phosphide, indium phosphide, SiN, Ba(NO3)2, CaCO3, NaNO3, tungstate crystals, BaF2, potassium titanyl phosphate (KTP), potassium dihydrogen phosphate (KDP), LiNbO3, deuterated potassium dihydrogen phosphate (DKDP), lithium triborate (LBO), barium borate (BBO), bismuth triborate (BIRO), LiIO3, BaTiO3, yttrium iron garnet (YIG), AlGaAs, CdTe, AgGaS2, KTiOAsO4 (KTA), ZnGeP2 (ZGP), RBTiOAsO4 (RTA).
In some embodiments, the three-dimensional material may be applied as a waveguide structure.
The radiation propagation portion may comprise a uniform three-dimensional medium covered by a two-dimensional or quasi-two-dimensional material layer. In some embodiments the layer of two-dimensional material or quasi-two-dimensional material may be a layer extending over the full surface of the underlying three-dimensional medium.
The two-dimensional or quasi-two-dimensional material layer may be a graphene layer.
The two-dimensional material layer may be adapted for having an electric current flowing through it.
The radiation propagation portion may comprise a uniform three-dimensional material covered by a two-dimensional or quasi-two-dimensional material layer that is patterned.
The two-dimensional or quasi-two-dimensional material layer may be patterned such that periodic variations in the nonlinear optical susceptibility are introduced.
The two-dimensional or quasi-two-dimensional material layer may have a pie-shaped patterning.
The two-dimensional or quasi-two-dimensional material layer may be a full layer, but wherein the layer is locally chemically or electrically modified.
The nonlinear optical wave mixing may be a third order process, such as four wave mixing, or a second order process such as three wave mixing. Three wave mixing may comprise sum-frequency generation or difference frequency generation.
The system for conversion or amplification may be a system for Raman conversion, parametric conversion or parametric amplification. It is an advantage of at least some embodiments according to the present invention that the structure may be closed, so that the structure is a resonator allowing to establish cavity enhancement. In at least some embodiments according to the present invention, the closed structure may be a ring structure or a disc structure, such as for example a circular ring, an elliptical ring, a rectangular ring, an octagonal ring, a circular disc, an elliptical disc, a rectangular disc, an octagonal disc, etc. The structure also may be an open structure such as a snake-like structure, a sickle structure, or a spiral structure.
It is an advantage of at least some embodiments according to the present invention that a relatively simple system can be obtained allowing quasi-phase-matched nonlinear optical wave mixing, e.g. quasi-phase-matched sum-frequency generation (SFG), quasi-phase-matched difference frequency generation (DFG), quasi-phase-matched Raman-resonant FWM or quasi-phase-matched Kerr-induced FWM. It is an advantage of at least some embodiments according to the present invention that no active alteration of the material properties of the three-dimensional medium and the two-dimensional or quasi-two-dimensional top layer are required for accurate operation, so that both the three-dimensional medium and the two-dimensional or quasi-two-dimensional top layer are uniform.
The bent structure may be a ring structure or disc structure.
The ring structure or disc structure may be circular, and the radius R of the ring structure or disc structure may be determined substantially inverse proportional with the linear phase mismatch for nonlinear optical wave mixing. The radius R thereby may be defined as the distance from the center of the circle to the central longitudinal axis in the circular radiation propagation area. The linear phase mismatch thereby may be defined by the following equations:
Δklinear=kp+ks−ki
for SFG,
Δklinear=kp−ks−ki
for DFG, and
Δklinear=2kp−ks−ki
for Raman-resonant FWM and Kerr FWM.
The radius R of the circular ring structure may be determined by the radius R being substantially equal to a factor s, equal to a positive or negative integer , times one divided by the linear phase mismatch for Raman-resonant FWM or divided by the linear phase mismatch for Kerr-induced FWM, i.e. it substantially fulfills relation
with s being a factor equal to a positive or negative integer so that R has a positive value, and Δklinear being linear phase mismatch for Raman-resonant FWM or being the linear phase mismatch for Kerr-induced FWM. The radius R of the circular ring structure may be determined by the radius R being substantially equal to a factor s, equal to a positive or negative integer, times one divided by the linear phase mismatch for SFG or divided by the linear phase mismatch for DFG, i.e. it substantially fulfills relation
with s being a factor equal to a positive or negative integer so that R has a positive value, and Δklinear being linear phase mismatch for SFG or being the linear phase mismatch for DFG . With substantially being equal to or substantially fulfilling the relation there is meant that advantageously the radius is equal or the relation is fulfilled, but that a deviation on the design rule is allowed wherein the quasi-phase-matched SFG efficiency, quasi-phase-matched DFG efficiency or quasi-phase-matched FWM efficiency is still high due to the explored effects. E.g. for a deviation of 5% on the design rule this is a value which certainly lies within the fabrication tolerances that can be achieved nowadays a quasi-phase-matched nonlinear optical wave mixing efficiency of 0.8 times the maximal efficiency at zero deviation may still be guaranteed. For a deviation of 10% on the design rule, a quasi-phase-matched nonlinear optical wave mixing efficiency of 0.5 times the maximal efficiency at zero deviation may still be guaranteed. For a deviation of 20% on the design rule, a quasi-phase-matched nonlinear optical wave mixing efficiency of 0.3 times the maximal efficiency at zero deviation may still be guaranteed. For deviations larger than 25% on the design rule, the quasi-phase-matched efficiency might become smaller than 0.2 times the maximal efficiency at zero deviation, and the quasi-phase-matching approach might not be interesting any longer.
The bent structure may have an inscribed circle and/or circumscribed circle having a radius inversely proportional to the linear phase mismatch for nonlinear optical wave mixing, e.g. SFG, DFG, FWM.
The bent structure may have an average radius inversely proportional to the linear phase mismatch for nonlinear optical wave mixing, e.g. SFG, DFG, FWM.
The system furthermore may be adapted to provide a pump radiation beam with wavenumber kp and a signal radiation beam with wavenumber lc, and result in an idler radiation beam with wavenumber ki, so that at least one of these beams is at resonator resonance, e.g. ring or disc resonance. In ring resonators for example, at least one of the beams' wavenumbers may yield, when multiplying with R, an integer number.
The system may comprise a heating and/or cooling means and a temperature controller for controlling the temperature so that at least one of the pump radiation, the signal radiation and the idler radiation is at resonator resonance.
The materials may be Raman-active materials, and the process may be a quasi-phase-matched Raman-resonant FWM process.
The materials may be Kerr-nonlinear materials and the process may be a quasi-phase-matched Kerr-nonlinear FWM-process.
The materials may be quadratically nonlinear optical materials and the process may be a quasi-phase-matched SFG or DFG process.
It is an advantage of at least some embodiments according to the present invention that quasi-phase-matched nonlinear optical wave mixing, e.g. nonlinear optical four wave mixing or nonlinear optical three wave mixing such as quasi-phase-matched SFG and/or quasi-phase-matched DFG and/or quasi-phase-matched Raman-resonant FWM and/or quasi-phase-matched Kerr-induced FWM in a uniform three-dimensional medium with a uniform two-dimensional or quasi-two-dimensional material layer on top can be obtained so that it does not require special techniques to periodically adapt the material characteristics of the three-dimensional material nor of the two-dimensional or quasi-two-dimensional material layer on top.
The three-dimensional material may be any or a combination of silicon, germanium, GaAs, InGaAs, diamond, cadmium telluride, gallium indium phosphide, indium phosphide, SiN, Ba(NO3)2, CaCO3, NaNO3, tungstate crystals, BaF2, potassium titanyl phosphate (KTP), potassium dihydrogen phosphate (KDP), LiNbO3, deuterated potassium dihydrogen phosphate (DKDP), lithium triborate (LBO), barium borate (BBO), bismuth triborate (BIRO), LiIO3, BaTiO3, yttrium iron garnet (YIG), AlGaAs, CdTe, AgGaS2, KTiOAsO4 (KTA), ZnGeP2 (ZGP), RBTiOAsO4 (RTA). The three-dimensional material may be a crystalline material.
In some embodiments, the three-dimensional material may be applied as a waveguide structure.
It is an advantage of embodiments according to the present invention that nonlinear optical wave mixing processes, such as SFG and/or DFG and/or Raman-resonant FWM and/or Kerr-induced FWM, can be established for wavelengths suitable for e.g. telecommunication and spectroscopy.
It is an advantage of at least some embodiments according to the present invention conventional semiconductor processing technology can be used for obtaining Raman converters, parametric converters or parametric amplifiers.
A controller may be provided for tuning the system with respect to an output wavelength, an output power or an obtained bandwidth.
The system may be adapted for selecting a TE or TM output by selecting a TE or TM input. It is an advantage of at least some embodiments according to the present invention that the polarization of the output is the same as the polarization of the input of the Raman converter, of the parametric converter, and of the parametric amplifier, and thus that no additional polarization filter is required for obtaining a particular polarized output.
The three dimensional material may be a uniform nonlinear optical material.
The present invention also relates to a method for obtaining conversion or amplification, using quasi-phase-matched nonlinear optical wave mixing, the method comprising receiving a pump radiation beam and a signal radiation beam in a bent structure, a radiation propagation portion, e.g. waveguiding portion, of the bent structure comprising a uniform three-dimensional material at least partly covered by a two-dimensional material layer or a quasi-two-dimensional material layer and having a dimension taking into account the spatial variation of the nonlinear optical susceptibility along the radiation propagation portion as experienced by radiation travelling along the bent structure for obtaining quasi-phase-matched nonlinear optical wave mixing in the radiation propagation portion, the dimension being substantially inverse proportional with a linear phase mismatch for nonlinear optical wave mixing, obtaining an idler radiation beam by interaction of the pump radiation beam and the signal radiation beam and coupling out an idler radiation beam from the bent structure. Conversion or amplification may be any of Raman conversion, parametric conversion or parametric amplification. The nonlinear optical wave mixing may be a SFG process, a DFG progress, a Raman-resonant FWM process or a Kerr-induced FWM process.
The method may be performed using a device as described above.
The bent structure may for example be a ring or disc structure, where the pump radiation beam and the signal radiation beam propagate in the ring or disc structure, whereby the ring or disc structure is circular and has a radius R determined substantially inverse proportional with a linear phase mismatch for quasi-phase-matched nonlinear optical wave mixing, e.g. quasi-phase-matched SFG, quasi-phase-matched DFG, quasi-phase-matched Raman-resonant FWM or quasi-phase-matched Kerr-induced FWM.
For three wave mixing processes such as the SFG and DFG processes, the pump radiation beam and the signal radiation beam may be guided in a circular ring structure having a radius substantially fulfilling the relation
with s being a positive or negative integer so that R has a positive value, and Δklinear being the linear phase mismatch for SFG or being the linear phase mismatch for DFG. With substantially fulfilling the relation there is meant that advantageously the relation is fulfilled, but that a deviation on the design rule is allowed wherein the quasi-phase-matched SFG efficiency or quasi-phase-matched DFG efficiency is still high due to the explored effects.
For the FWM processes, the pump radiation beam and the signal radiation beam may be guided in a circular ring structure having a radius substantially fulfilling the relation
with s being a positive or negative integer so that R has a positive value, and Δklinear being the linear phase mismatch for Raman-resonant FWM or being the linear phase mismatch for Kerr-induced FWM.
With substantially fulfilling the relation there is meant that advantageously the relation is fulfilled, but that a deviation on the design rule is allowed wherein the quasi-phase-matched FWM efficiency is still high due to the explored effects.
The method may comprise letting the radiation beams propagate in the ring or disc structure and obtaining ring or disc resonance for at least one of the different radiation beams.
The method may comprise adjusting the in- and/or outcoupling efficiency for adjusting the cavity-enhancement of the radiation beams inside the ring or disc structure.
The method may comprise tuning the system with respect to an output wavelength, an output power or an obtained bandwidth.
The present invention also relates to a method for designing a converter or amplifier using quasi-phase-matched nonlinear optical wave mixing, the converter or amplifier using a pump radiation beam and a signal radiation beam, the method comprising selecting a bent structure suitable for quasi-phase-matched nonlinear optical wave mixing comprising selecting materials for a radiation propagation portion of the bent structure, the radiation propagation portion comprising a uniform three-dimensional material at least partly covered by a two-dimensional material layer or quasi-two-dimensional material layer, and selecting at least one dimension of the radiation propagation portion taking into account the spatial variation of the nonlinear optical susceptibility along the radiation propagation portion as experienced by radiation travelling along the bent structure. The dimension thereby is substantially inverse proportional with the linear phase mismatch for nonlinear optical wave mixing.
The present invention also relates to a computer program product for, when executed on a computer, performing a method and/or controlling a system as described above. The present invention also relates to a data carrier carrying such a computer program product or to the transmission of such a computer program product over a wide or local area network.
In one aspect, the present invention also relates to a system for conversion or amplification using quasi-phase matched nonlinear optical wave mixing, the system comprising a first radiation source for providing a pump radiation beam, a second radiation source for providing a signal radiation beam, and a bent structure for receiving the pump radiation beam and the signal radiation beam, wherein a radiation propagation portion of the bent structure is made of a uniform three-dimensional material and wherein the radiation propagation portion comprises a dimension taking into account the spatial variation of the nonlinear susceptibility along the radiation propagation portion as experienced by radiation travelling along the bent structure for obtaining quasi-phase-matched nonlinear optical wave mixing in the radiation propagation portion, the dimension being substantially inverse proportional with the linear phase mismatch for nonlinear optical wave mixing and an outcoupling radiation propagation portion, e.g. waveguiding portion, for coupling out an idler radiation beam generated in the bent structure. It is an advantage of embodiments according to the present invention that quasi-phase matching conditions can be achieved. The uniform three-dimensional material thereby is a quadratically nonlinear optical material and the process is a quasi-phase-matched sum-frequency generation or quasi-phase-matched difference-frequency generation. It is to be noticed that according to this aspect, there is no need for a layer of two-dimensional or quasi-two-dimensional material to be present on top of the three dimensional material.
In another aspect, the present invention also relates to a method for obtaining conversion or amplification, using quasi-phase-matched nonlinear optical wave mixing, the method comprising receiving a pump radiation beam and a signal radiation beam in a bent structure, a radiation propagation portion of the bent structure being made of a uniform three-dimensional material and comprising a dimension taking into account the spatial variation of the nonlinear optical susceptibility along the radiation propagation portion as experienced by radiation travelling along the bent structure for obtaining quasi-phase-matched nonlinear optical wave mixing in the radiation propagation portion, the dimension being substantially inverse proportional with a linear phase mismatch for nonlinear optical wave mixing, obtaining an idler radiation beam by interaction of the pump radiation beam and the signal radiation beam and coupling out an idler radiation beam from the bent structure.
Said nonlinear optical wave mixing is nonlinear optical three-wave mixing, the uniform three-dimensional material is a quadratically nonlinear optical material and the process is a quasi-phase-matched sum-frequency generation or quasi-phase-matched difference-frequency generation.
Further features of these aspects wherein the three-dimensional material is a quadratically nonlinear optical material may comprise additional features and aspects as described in the different embodiments of the other aspects, but where applicable the covering layer of two-dimensional or quasi-two-dimensional material is omitted.
Particular and preferred aspects of the invention are set out in the accompanying independent and dependent claims. Features from the dependent claims may be combined with features of the independent claims and with features of other dependent claims as appropriate and not merely as explicitly set out in the claims.
These and other aspects of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.
The drawings are only schematic and are non-limiting. In the drawings, the size of some of the elements may be exaggerated and not drawn on scale for illustrative purposes.
Any reference signs in the claims shall not be construed as limiting the scope.
The present invention will be described with respect to particular embodiments and with reference to certain drawings but the invention is not limited thereto but only by the claims. Furthermore, the terms first, second, third and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a sequence, either temporally, spatially, in ranking or in any other manner. It is to be understood that the terms so used are interchangeable under appropriate circumstances and that the embodiments of the invention described herein are capable of operation in other sequences than described or illustrated herein.
It is to be noticed that the term “comprising”, used in the claims, should not be interpreted as being restricted to the means listed thereafter; it does not exclude other elements or steps. It is thus to be interpreted as specifying the presence of the stated features, integers, steps or components as referred to, but does not preclude the presence or addition of one or more other features, integers, steps or components, or groups thereof. Thus, the scope of the expression “a device comprising means A and B” should not be limited to devices consisting only of components A and B. It means that with respect to the present invention, the only relevant components of the device are A and B.
Reference throughout this specification to “one embodiment” or “an embodiment” means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, appearances of the phrases “in one embodiment” or “in an embodiment” in various places throughout this specification are not necessarily all referring to the same embodiment, but may. Furthermore, the particular features, structures or characteristics may be combined in any suitable manner, as would be apparent to one of ordinary skill in the art from this disclosure, in one or more embodiments.
Furthermore, while some embodiments described herein include some but not other features included in other embodiments, combinations of features of different embodiments are meant to be within the scope of the invention, and form different embodiments, as would be understood by those in the art. For example, in the following claims, any of the claimed embodiments can be used in any combination.
In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In other instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Where in embodiments of the present invention reference is made to a Raman-active material, reference is made to a material or medium wherein the Raman susceptibility has a non-zero value.
Where in embodiments of the present invention reference is made to a Kerr-nonlinear material or medium, reference is made to a material or medium wherein the Kerr susceptibility has a non-zero value.
Where in embodiments of the present invention reference is made to a quadratically nonlinear material or medium, reference is made to a material or medium wherein the second-order susceptibility has a non-zero value.
Where in embodiments according to the present invention reference is made to a quasi-phase-matched (QPM) nonlinear optical wave mixing process, such as for example QPM SFG, QPM DFG, QPM Raman-resonant FWM or QPM Kerr-induced FWM, reference is made to a nonlinear optical wave mixing process where quasi-phase-matching in embodiments of the present invention is obtained in a non-traditional way, namely using a uniform three-dimensional material which can be covered with a two-dimensional or quasi-two-dimensional material layer. Nonlinear optical wave mixing may encompass for example four wave mixing or three wave mixing. For example, QPM SFG, QPM DFG, QPM Raman-resonant FWM or QPM Kerr-induced FWM can be obtained for any value of the linear phase mismatch Δklinear. When the value of the linear phase mismatch Δklinear for these nonlinear optical processes is negligibly small (Δklinear≈0) and the processes take place in a device that is not designed for QPM operation, the process is called to be perfectly phase-matched (PPM). When the product of the linear phase mismatch and the propagation distance d has an absolute value smaller than pi (|Δklinear|·d<π) and the nonlinear optical process takes place in a device that is not designed for QPM operation, the process is said to feature coherence-length-dependent operation. A process is not referred to as quasi-phase matched (QPM), as perfectly phase-matched (PPM) or as coherence-length dependent (CLD) in case none of the above approaches apply. To understand how PPM operation can be obtained for example for FWM, one needs to take into account that for FWM Δklinear can be written as
This relation shows that one can establish PPM operation at large |Δω|—values in e.g. a silicon waveguide by engineering the β2 and β4-factors, i.e. the dispersion of the silicon waveguide. As the term function of β4 in the formula above generally is less important than the term function of β2, this dispersion engineering implies that one should establish β2=0 at the preferred pump wavelength, which then corresponds to the so-called zero-dispersion wavelength (ZDW).
Where in embodiments of the present application reference is made to a bent structure, reference is made to a non-straight structure. The latter also may be expressed as a structure wherein the propagation direction of propagating radiation is altered. The latter may for example be a curved structure, such as for example a circular, elliptical, or spiral structure, or a broken structure, such as for example an octagonal shaped structure or a rectangular shaped structure. In addition thereto the bent structure also encompasses the situation whereby the three dimensional material is as such not shaped but has a structure, e.g. a crystallographic structure, allowing to bend radiation when it is passing in the structure.
Where in embodiments of the present invention reference is made to a radiation propagation portion, reference may be made to a medium that allows propagation of radiation, and that for example can be a waveguide or a medium that allows free-space radiation propagation.
In a first aspect, the present invention relates to methods and systems for performing conversion or amplification using QPM nonlinear optical processes, more particularly nonlinear optical wave mixing processes. Such nonlinear optical processes encompass e.g. four wave mixing processes as well as three wave mixing processes such as SFG, DFG, Raman-resonant FWM and Kerr-induced FWM. The methods and systems for performing conversion or amplification may be methods and systems for performing Raman conversion, for performing parametric conversion or for performing parametric amplification. The system according to embodiments of the present invention comprises a first radiation source for providing a pump radiation beam and a second radiation source for providing a signal radiation beam. The system furthermore comprises a bent structure for receiving the pump radiation beam and the signal radiation beam, wherein a radiation propagation portion, e.g. a waveguide portion of the bent structure is made of a uniform three-dimensional material that can be at least partly covered by a two-dimensional material layer or quasi-two-dimensional material layer. Due to the bending or curvature of the radiation propagation structure, radiation travelling through the bent structure will not see a uniform nonlinear optical susceptibility, but will see a variation therein, even if not only the three-dimensional material is uniform but also the two-dimensional or quasi-two-dimensional nonlinear optical material layer on top is uniform. More particularly, whereas the uniform material has a uniform optical nonlinearity in a laboratory reference system fixed to the system, a variation in the nonlinear susceptibility is present felt by the radiation travelling through the bent structure, depending on the polarization of the radiation and the orientation of the principle crystal axes of the material used.
In other words, the two-dimensional material layer may induce a variation in nonlinear optical susceptibility. According to embodiments of the present invention, the variation may be induced by a configuration as indicated above. In the description and examples indicated below, different embodiments will be more explicitly described, the invention not being limited thereto.
According to embodiments of the present invention, the dimensions of the bent structure are selected taking into account the spatial variation of the susceptibility along the bent structure as experienced by the radiation travelling along the bent structure so that non-traditional QPM SFG, DFG, or FWM is obtained in the bent structure made of a uniform three-dimensional material which can be covered with a two-dimensional or quasi-two-dimensional material layer. The bent structure thus may be any structure allowing to change or alter, e.g. curve, the propagation direction of the radiation, such that a variation in susceptibility is felt by the radiation. In advantageous embodiments, the bent structure may be a closed structure, such as for example a ring structure or disc structure. Such ring or disc structure may for example be a circular ring, an elliptical ring, an octagonal ring, a rectangular ring, a circular disc, an elliptical disc, an octagonal disc or a rectangular disc and the properties of the closed structure may be selected such that at least one of the radiation beams is enhanced. Alternatively, the structure may be an open structure wherein a change is induced in the propagation direction of the radiation such that a variation in susceptibility is felt by the radiation. An example thereof could be a sickle-shaped structure, a snake-shaped structure, or a spiral-shaped structure, the contours of which are along a circular ring, an octagon, or another type of polygon. A number of particular examples is shown in
As indicated, a dimension of the bent structure is selected so that QPM FWM is obtained in the bent structure made of a uniform three-dimensional material which can be covered with a two-dimensional or quasi-two-dimensional material layer. The typical dimension of a structure may be an average length of a radiation propagation portion, e.g. waveguide portion, of the bent structure, but also may be for example a radius of the bent structure, an average radius of the bent structure, a radius of an inscribed circle or in-circle of the structure, a radius of a circumscribed circle or circumcircle, etc. In some embodiments, a dimension also may be an average radius of curvature. If for example the average length is used, the average length of the radiation propagation part of the bent structure may be in a range between 1μm and 10 cm.
According to embodiments of the present invention, a dimension of the bent structure or more particularly the radiation propagation portion thereof is such that it is substantially inverse proportional with the linear phase mismatch for SFG, DFG or FWM. The linear phase mismatch for SFG equals the pump wavenumber plus the signal wavenumber minus the idler wavenumber, the linear phase mismatch for DFG equals the pump wavenumber minus the signal wavenumber minus the idler wavenumber, and the linear phase mismatch for FWM equals two times the pump wavenumber minus the signal wavenumber minus the idler wavenumber. These linear phase mismatches indicate how fast the dephasing of the different fields first grows and then returns to zero again in a periodic way along the propagation path.
The typical dimension of the bent structure may be inversely proportional to the linear phase mismatch for SFG, DFG or FWM. In other words
In some embodiments according to the present invention, a closed loop structure is used and the structure is adapted for enhancing at least one and advantageously a plurality or more advantageously all of the radiation beams in the closed loop structure. Nevertheless, also open structures or open loop structures are envisaged. The system furthermore comprises an outcoupling radiation propagation portion, e.g. a waveguide, for coupling out an idler radiation beam generated in the bent structure.
The uniform material used may be a uniform quadratically nonlinear material, a uniform Raman-active and/or uniform Kerr-nonlinear material.
In embodiments of the present invention whereby a bent structure made of a uniform three-dimensional material covered by a two-dimensional or quasi-two-dimensional material is used, different materials can be used. The two-dimensional or quasi-two-dimensional material that may be used may for example be one or a combination of graphene, graphyne, borophene, germanene, silicene, stanine, phosphorene, metals, 2D supracrystals, hexagonal boron nitride, germanane, nickel HITP, transition metal di-chalcogenides (TMDCs), MXenes or black phosphorus, or topological insulator. These materials have a typical thickness ranging from below 1 nm up to a few nm. The three-dimensional material that may be used may for example be silicon, silicon on insulator, SiN, GaAs, InGaAs, diamond, cadmium telluride, gallium indium phosphide, indium phosphide and other crystals such as Ba(NO3)2, CaCO3, NaNO3, tungstate crystals, BaF2, potassium titanyl phosphate (KTP), potassium dihydrogen phosphate (KDP), LiNbO3, deuterated potassium dihydrogen phosphate (DKDP), lithium triborate (LBO), barium borate (BBO), bismuth triborate (BIBO), LiIO3, BaTiO3, yttrium iron garnet (YIG) crystals, AlGaAs, CdTe, AgGaS2, KTiOAsO4 (KTA), ZnGeP2 (ZGP), RBTiOAsO4 (RTA)
The structure may be made in a plurality of ways. It may be processed on a substrate, it may be fabricated using different techniques such as CMOS technology, electron beam lithography, photolithography, chemical vapour deposition (CVD), low-pressure chemical vapour deposition (LPCVD), pulsed laser deposition (PLD), plasma enhanced chemical vapour deposition (PECVD), electrochemical delamination, thermal oxidation, reactive-ion etching, focused ion beam, crystal growth, epitaxial growth, sputtering, flux pulling method from a stoichiometric melt, and polishing.
As indicated above, the system comprises a first and second radiation sources for generating a pump radiation beam and a signal radiation beam. Such radiation sources typically may be lasers, although embodiments of the present invention are not limited thereto. The type of lasers selected may depend on the application. Some examples of lasers that could be used are semiconductor lasers, solid-state lasers, fiber lasers, gas lasers, . . . . The required output power and wavelength of e.g. the pump laser depends on the output that one wants to obtain, e.g. of the output power one expect from the converter or amplifier.
In some embodiments, the system also may comprise a controller for controlling the system, e.g. the first radiation source and the second radiation source, and environmental conditions of the system, so as to be able to slightly tune the system. In one embodiment, a heating and/or cooling means, e.g. heater and/or cooler, may be present for controlling the temperature of the system and in this way also properties of the system. In an advantageous embodiment, the controller may be adapted so that defined conditions for obtaining cavity-enhanced quasi-phase-matched SFG, cavity-enhanced quasi-phase-matched DFG, or cavity-enhanced quasi-phase-matched FWM, such as a well-controlled temperature, are maintained in the system. Such a controller may operate in an automated and/or automatic way. The controller may be implementing predetermined rules or a predetermined algorithm for controlling the system, or it may be adapted for using a neural network for controlling the system. The controller may comprise a memory for storing data and a processor for performing the steps as required for controlling. The controller may be computer implemented. Whereas in the present aspect, the controller is described as a component of the system, in one aspect, the present invention also relates to a controller as such for performing a method of controlling a system for operating in quasi-phase-matched SFG conditions, quasi-phase-matched DFG conditions, or quasi-phase-matched FWM conditions.
In some embodiments, the system also may comprise a feedback system, providing parameters for checking whether the appropriate conditions are fulfilled and for reporting corresponding information. Such information may for example be transferred to the controller and used by the controller for adjusting or correcting the conditions.
In some embodiments, the resonator, converter or amplifier is adapted for providing a given polarization mode. It thereby is an advantage that no filter means is required for obtaining the polarization mode, as the polarization mode is not altered by the structure.
By way of illustration and for the ease of explanation, embodiments of the present invention not being limited thereto, some features and aspects will now further be described with reference to QPM Raman-resonant FWM and to QPM Kerr-induced FWM in a circular ring structure, and with reference to QPM SFG and QPM DFG in a spiral structure. These provide, without embodiments of the present invention being bound by theory, a possible explanation of the features of the obtained structures.
QPM Raman-resonant FWM and QPM Kerr-induced FWM is discussed in a three-dimensional (100) grown ring-shaped silicon-on-insulator (SOI) waveguide, which in embodiments of the present invention can be covered by a layer of a two-dimensional or quasi-two-dimensional material as described above. The shape of the bent structure as used in embodiments according to the present invention is illustrated by way of example in
where s=±4 so that R has a positive value, and R is the ring radius in case of a circular ring. Important to know is that even if this quasi-phase-matching condition is not exactly fulfilled, for example due to small deviations of R , the quasi-phase-matching efficiency will still be high.
It is to be noticed that this approach can also be used for any other Raman-active medium with the same crystal symmetry as silicon, for any other Kerr-nonlinear medium with the same crystal symmetry as silicon, and for some Raman-active media and/or Kerr-nonlinear media with a crystal symmetry similar to that of silicon. Hence, many crystals can be used, some examples of which are SiN, germanium, GaAs, InGaAs, diamond, cadmium telluride, gallium indium phosphide, indium phosphide, Ba(NO3)2, CaCO3, NaNO3, tungstate crystals, BaF2, potassium titanyl phosphate (KTP), potassium dihydrogen phosphate (KDP), LiNbO3, deuterated potassium dihydrogen phosphate (DKDP), lithium triborate (LBO), barium borate (BBO), bismuth triborate (BIBO), LiIO3, BaTiO3, yttrium iron garnet (YIG) crystals.
In addition to achieving QPM FWM, one wants to design the ring so that all waves involved in the Raman-resonant FWM process and in the Kerr-induced FWM process are resonantly enhanced in the ring; this will lead to high intensities in the ring even for low intensity input waves. Complete resonant enhancement occurs when the values of k{p,s,i}R correspond to integer numbers. It is remarked that if kpR and ksR have integer values and if in addition the quasi-phase-matching condition expressed above is fulfilled, then kiR will also correspond to an integer number. As such, whereas for ring converters based on the principle of “coherence-length-dependent (CLD) operation” it is not possible to have the pump, signal and idler waves all at ring resonances in the presence of a non-zero Δklinear, such a triply-resonant condition at a non-zero Δklinear does become possible when using QPM operation. It is also noted that in most cases the free spectral range of the ring will be quite small, so that a small temperature tuning will suffice to guarantee that the pump and signal waves, and automatically also the idler wave, are at ring resonances. Using temperature tuning, one can also compensate for phase-shifting phenomena that might occur in the silicon medium, such as self- and cross-phase modulation.
QPM SFG and DFG are discussed in a spiral-shaped waveguide made of a uniform three-dimensional material which in embodiments of the present invention can be covered by a layer of a two-dimensional or quasi-two-dimensional material as described above. Different bent structures as can be used in embodiments of the present invention are shown in
A first exemplary embodiment of the present invention is shown in
where s=±1 so that R has a positive value, and R is the average radius of the spiral. Important to know is that even if this quasi-phase-matching condition is not exactly fulfilled, for example due to small deviations of R , the quasi-phase-matching efficiency will still be high.
It is to be noticed that this approach can also be used for any other quadratically nonlinear material with the same crystal symmetry as graphene through which an electric current is flowing, and for some quadratically nonlinear media with a crystal symmetry similar to that of graphene through which an electric current is flowing. Hence, many materials can be used, some examples of which are graphyne, borophene, germanene, silicene, stanine, phosphorene, metals, 2D supracrystals, hexagonal boron nitride, germanane, nickel HITP, transition metal di-chalcogenides (TMDCs), MXenes or black phosphorus, topological insulator, SiN, GaAs, InGaAs, Ba(NO3)2, CaCO3, NaNO3, tungstate crystals, BaF2, potassium titanyl phosphate (KTP), potassium dihydrogen phosphate (KDP), LiNbO3, deuterated potassium dihydrogen phosphate (DKDP), lithium triborate (LBO), barium borate (BBO), bismuth triborate (BIRO), LiIO3, BaTiO3, yttrium iron garnet (YIG) crystals, AlGaAs, CdTe, AgGaS2, KTiOAsO4 (KTA), ZnGeP2 (ZGP), RBTiOAsO4 (RTA).
The amplifier or converter according to embodiments of the present invention may also provide the functionality of a resonator, embodiments not being limited thereto.
In a second particular embodiment, similar bent structures as indicated above can be used, but thetwo-dimensional or quasi-two-dimensional cover layer is not fully covering the bent structure but is patterned. The patterning may be adapted such that periodic variations in the nonlinear optical susceptibility are introduced. The patterning may for example be a pie-shaped patterning. The patterning periodicity should be chosen proportional to the radius of the bent structure, with the radius chosen inversely proportional to the linear phase mismatch.
In a third particular embodiment, similar bent structures as indicated above can be used, whereby the two-dimensional or quasi-two-dimensional cover layer is fully covering the bent structure but the covering layer is locally modified chemically or electrically such that the properties of the covering layer show a spatial variation, resulting in periodic variations in the nonlinear optical susceptibility that are introduced for the radiation. This equally results in the possibility of conversion or amplification using a quasi-phase matched nonlinear optical wave mixing. The electrical modification may comprise applying a voltage on the cover layer.
By way of illustration, embodiments of the present invention not being limited thereto, the present invention now will be further illustrated with reference to particular embodiments, illustrating some features and advantages of embodiments according to the present invention. Whereas in the first two embodiments, principles are described for non-covered bent structures made of uniform three-dimensional materials, the principles and features illustrated are equally applicable to the situation of a bent structure made of a uniform three-dimensional material covered with a layer of two-dimensional or quasi-two-dimensional material, e.g. as discussed in the above described embodiments.
Without wishing to be bound by theory, first a mathematical suggestion of how the principles of embodiments of the present invention could be explained also is provided.
In a particular embodiment, reference is made to a QPM Raman-resonant FWM system based on a silicon ring resonator. The system of the example shown thereby is not only adapted for QPM Raman-resonant FWM, but also illustrates that advantageously use can be made of cavity enhancement effects and of the free choice of the waveguide geometry when using quasi-phase matching.
In this embodiment two comparisons will be made between QPM silicon Raman ring converters and PPM Raman converters. To do this, first a modeling formalism for Raman converters is introduced.
Without restricting the general validity of the results, focus is made on quasi-continuous-wave operation and on operation at exact Raman resonance. Assuming that nk/nl≈1 (for k,l=p,s,i) and that Kerr-induced FWM in silicon is negligible at the considered working point of exact Raman resonance, the equations expressing the steady-state spatial variation of the slowly-varying pump, signal and idler field amplitudes Ap(ζ), As(ζ), Ai(ζ) in the SOI ring Raman converter are given by
where ζ=Rθ and A{p,s,i} is normalized such that |A{p,s,i}|2 corresponds to intensity. The function ρ(θ) will be specified further on. The terms containing eiΔk
with j=p,s,i, with the positions of the fields (1)-(4) indicated in
One now can solve numerically equations (2) to (5) to make two comparisons: On one hand, to illustrate the effect of cavity enhancement in the QPM silicon Raman ring converters, a concrete QPM Raman ring converter configuration and a theoretical one-dimensional PPM Raman converter without losses will be compared. On the other hand, to illustrate the effect of having a free choice for the waveguide geometry in QPM silicon Raman ring converters, a concrete QPM Raman ring converter configuration and a concrete PPM Raman ring converter, both with losses also will be compared.
For the first exemplary comparison in the first embodiment, one considers for the QPM Raman ring converter configuration a TE-polarized pump input and a TE-polarized Stokes-shifted signal input with a frequency difference corresponding to the exact Raman resonance: ωp=1.22×1015 rad/s (λp=1.55 μm), ωs=1.12×1015 rad/s (λs=1.686 μm). This leads to a generated idler wave with angular frequency ωi=1.32×1015 rad/s (λi=1.434 μm). The system may have a structure as illustrated in
In conclusion, the idler output intensity of a QPM silicon ring Raman converter can easily become 3×103 times larger than that of a one-dimensional PPM Raman converter of equal length. Taking into account the quadratic dependence of the latter's output on the pump input, this also implies that the QPM ring Raman converter needs a 50 times smaller pump input intensity than the one-dimensional PPM Raman converter to produce the same idler output. Furthermore, signal-to-idler conversion efficiencies larger than unity can be obtained using relatively low pump input intensities. These improvements in conversion performance substantially expand the practical applicability of Raman converters in different application domains.
For the second comparison in the first embodiment, the same QPM Raman ring converter configuration is considered as described above. Instead of comparing it to a theoretical one-dimensional PPM Raman converter without losses as was done above, it is compared to a concrete PPM ring Raman converter with losses included and where Δklinear=0 is obtained by dispersion engineering the nanowire. For the latter device, the same parameter values are adopted as for the QPM converter described above, including L=2.1 mm and Kp2=Ks2=Ki2=0.05, except for the following: Δklinear=0 cm−1, ρR(θ)=1, and τeff=3 ns. Again one assumes all three waves to be at ring resonances. Now one can numerically solve equations (2)-(5) for both Raman converter configurations, while assuming a signal input power of 20×10−6 W, corresponding to Is,in=1×108 W/m2, and while varying the pump input power between 2 mW and 20 mW in steps of 1 mW this corresponds to Ip,in ranging from 1×1010 W/m2 tp 1×1011 W/m2 steps of 0.5×1010 W/m2. The reason for choosing a variable pump input is that for the PPM converter with large τeff the nonlinear losses will become significant already at low pump powers, whereas for the QPM converter with small τeff the onset of nonlinear losses will occur at higher pump powers. For the pump power levels of 20 mW (5 mW), the steady-state distributions along the ring of the pump, signal and idler intensities in the QPM (PPM) converter are represented by the solid (dashed) lines in
It is pointed out that for the QPM device the TPA losses will also undergo a periodic variation proportional to (0.88+0.12 cos2(2θ)) in the ring, but as the varying part of the TPA losses is small compared to the constant part, this variation only has a small influence, as simulations that are not presented here in detail confirm. In the second embodiment it will be shown, however, that an equally small variation of the effective Kerr nonlinearity γ does suffice to effectively establish quasi-phase-matching in a parametric converter, since this variation establishes a phase effect rather than an intensity loss effect.
One might question whether the coupling coefficients Kp2=Ks2=Ki2=0.05 assumed here yield the highest conversion efficiencies for the QPM and PPM Raman devices, and whether the ring circumference L=2.1 mm calculated for the QPM converter is the most optimal ring circumference for the PPM converter as well. When varying the coupling coefficients using intermediate steps of 0.005 for the QPM converter with L=2.1 mm and for the PPM converter with different ring circumferences (see
In conclusion for the present examples, since for a QPM Raman ring converter the nanowire geometry can be chosen such that the FCA losses are minimal, the device should, when considering actual converter operation with losses included, substantially outperform a PPM Raman ring converter based on a dispersion-engineered nanowire of the type presented earlier in the literature. It is remarked that the latter comparison holds provided that both devices are fabricated using the low-cost intrinsic silicon-on-insulator platform without carrier-extracting p-i-n diodes. Furthermore, the QPM Raman ring converter should significantly outperform the best-performing silicon Raman converter demonstrated thus far, as it is able to establish signal-to-idler conversion efficiencies larger than 0 dB at modest pump powers. Such high performance, combined with the fact that no dispersion engineering is required and that the device can be realized in the low-cost intrinsic silicon-on-insulator platform, show the potentialities of QPM Raman wavelength conversion in silicon rings.
In another particular embodiment, reference is made to a QPM Kerr-induced FWM system based on a silicon ring resonator. The system of the example shown thereby is not only adapted for QPM Kerr-induced FWM, but also illustrates that advantageously use can be made of cavity enhancement effects and that efficient conversion can be established for a large pump-signal frequency shift in a spectral domain where the dispersion characteristics of the silicon waveguide are not optimally engineered for PPM Kerr-induced FWM. As mentioned above, the condition for QPM Kerr-induced FWM in the ring is given by
where s=±4 so that R has a positive value, and R is the ring radius in case of a circular ring. Taking into account that Δklinear≈−β2(Δω)2, one finds that this quasi-phase-matching condition can be fulfilled even if the pump-signal frequency shift is large and if one works in a spectral domain where the dispersion characteristics of the silicon waveguide are not optimally engineered for PPM Kerr-induced FWM. Furthermore, the relation Δklinear≈−β2(Δω)2 also indicates that, for a given value of R, the quasi-phase-matching condition (1) can be fulfilled for different combinations of β2 and Δω. Thus, for a ring resonator with a ring radius R and with a properly designed, non-constant dispersion profile, one can convert via QPM Kerr-induced FWM a fixed signal frequency ωs to various idler frequencies ωi spread over the near- and mid-infrared range, by changing only the pump frequency ωp. Finally, if R is chosen to be small to keep the device compact, one finds that Δω can be large also if β2 is large.
As also mentioned above, the quasi-phase-matching condition expressed above complies with the condition for having the pump field, the signal field and the idler field at ring resonances. The fact that efficient non-traditional quasi-phase-matching can be combined with cavity enhancement for all three fields in the ring resonator is an important advantage, since for Kerr-induced FWM with “phase-matched operation” one can obtain cavity enhancement for all three fields only if the pump wavelength is close to the ZDW, i.e. only if one has PPM operation. Otherwise, one has CLD operation in a doubly-resonant condition rather than in a triply-resonant condition. It also can be remarked that, for QPM Kerr-induced FWM, the varying Kerr susceptibility in the ring does not reach zero as minimal value, which is not ideal. However, since it can be combined with cavity enhancement for all three fields also if the GVD at the pump wavelength has a large absolute value and/or the frequency difference between the pump and signal is large, QPM Kerr-induced FWM can in those circumstances establish efficiencies that are relatively high compared to the efficiencies achieved with CLD Kerr-induced FWM.
In this embodiment, two comparisons will be made between QPM ring-based parametric converters and CLD ring-based parametric converters in a spectral domain where the dispersion characteristics of the silicon waveguide are not optimally engineered for PPM Kerr-induced FWM. A one-dimensional CLD parametric converter is not explicitly considered in this comparison to demonstrate the effect of the cavity enhancement of the QPM ring-based parametric converter, as this would yield results along the same lines as those obtained for the first comparison in the previous embodiment on Raman converters. To compare QPM ring-based parametric converters and CLD ring-based parametric converters, a modeling formalism for parametric converters is first introduced. Without restricting the general validity of the results, focus is made on (quasi-)continuous-wave operation. Assuming that nk/nl≈1 (for k,l=p,s,i), the equations expressing the steady-state spatial variation of the slowly-varying pump, signal and idler field amplitudes Ap(ζ), As(ζ), Ai(ζ) in the parametric converter are given by
where ζ=Rθ, γ(θ)=n20ξKρ(θ) (ωp/c) is the effective nonlinearity, n20 is the Kerr-nonlinear refractive index along the [011] direction, ξK=5/4, and A{p,s,i} is normalized such that |A{p,s,i}|2 corresponds to intensity. The function ρ(θ) will be specified further on. The first terms containing the square brackets at the right hand side of Eqs. (6)-(8) correspond to Kerr-induced self- and cross-phase modulation, and the terms containing eiΔk
with j=p,s,i with the positions of the fields (1)-(4) indicated in
One now can numerically solve Eqs. (6)-(9) to make two comparisons: a comparison is made between a concrete near-infrared-pumped QPM ring-based parametric converter and a concrete near-infrared-pumped CLD ring-based parametric converter, both for the case that the dispersion characteristics of the silicon waveguide in the near-infrared domain are not optimally engineered for PPM Kerr-induced FWM. On the other hand, a comparison is made between a concrete mid-infrared-pumped QPM ring-based parametric converter and a concrete mid-infrared-pumped CLD ring-based parametric converter, both for the case that the dispersion characteristics of the silicon waveguide in the mid-infrared domain are not optimally engineered for PPM Kerr-induced FWM.
For the first comparison in the second embodiment, a near-infrared-pumped QPM ring-based parametric converter is initially considered with the following parameter values: ωp=1.18×1015 rad/s (λp=1.6 μm), ωs=1.45×1015 rad/s (λs=1.3 μm), ωi=9.06×1014 rad/s (λi=2.08 μm), Δklinear=1606 cm−1 (corresponding to a dispersion paramater of 1600 ps/(nm*km) at λp), np0=6.5×10−18 m2/W, Ip,in=6×1010 W/m2, Is,in=1×108 W/m2, Ii,in=0 W/m2, waveguide model area A=0.09 μm2, linear loss α=0.9 dB/cm, two-photon absorption coefficient β=0.7×10−11 m/W, free carrier absorption efficiency φ=6×10−10, effective free carrier lifetime τeff=0.1 ns, κp=0.14, κs=0.10, κi=0.17, and ρ(θ)=(0.88+0.12 cos2(2θ)) along the ring. When implementing the value for Δklinear in the quasi-phase-matching condition with s=+1, one obtains that quasi-phase-matching is obtained for a ring radius R=25 Ξm, which corresponds to a ring circumference of 157 μm.
To demonstrate that also efficiencies higher than ×33 dB could be reached while still pumping in the near-infrared region, one now considers a QPM parametric converter that is pumped at another near-infrared pump wavelength where the nonlinear refractive index is larger than in the previous case. More specifically, a QPM parametric converter is considered with the following parameter values: ωp=1.05×1015 rad/s (λp=1.8 μm), ωs=1.32×1015 (λs=1.43 μm), ωi=7.76×1014 (λi=2.43 μm), Δklinear=1606 cm−1 (corresponding to a dispersion parameter of 1600 ps/(nm*km) at λp as in the previous case), n20=12×10−18 m2/W, two-photon absorption coefficient β=0.5×10−11 m/W, and free carrier absorption efficiency φ=(1.8/1.55)2×6×10−10. For all other parameters, the same values are taken as in the previous case.
For the second comparison in this embodiment, a mid-infrared-pumped QPM parametric ring converter with a TE-polarized pump input at ωp=8.20×1015 rad/s (λp=2.3 μm) and a TE-polarized signal input at ωs=9.87×1014 rad/s (λs=1.91 μm) is considered, which feature a large pump-signal angular frequency difference of 2π×26.6 THz. This leads to a generated idler wave at ωi=6.52×1014 rad/s (λi=2.89 μm. It is remarked that one of the most interesting application domains for wavelength conversion towards mid-infrared idler wavelengths around 3 μm is spectroscopy, as many substances are highly absorbing in that spectral range. At a pump wavelength of 2.3 μm the Kerr-nonlinear refractive index n20 of silicon along the [011] direction equals approximately 8×10−14 cm2/W. A nanowire is assumed here which is dispersion-engineered such that its ZDW is situated in the near-infrared, more specifically at 1.5 μm and which features a sufficiently large cross-section to have most of the mode energy at λp=2.3 μm confined in the nanowire core. The nanowire under consideration has a height of 516 nm and a width of 775 nm and has an oxide cladding. For such a waveguide geometry the ZDW is indeed situated at 1.5 μm and that the dispersion D at λp=2.3 μm equals approximately 1200 ps/(nm*km). The latter value yields Δk=2kp−ks−ka=9.40 cm−1. For s=+1, it is found that the condition (5) is met for a ring circumference L of 267 μm. The remaining device parameters are: Aeff=0.4 μm2, α=3 dB/cm, and ρ(θ)=(0.88+0.12 cos2(2θ)) along the ring. In these simulations coupling coefficients are considered ranging from κp2=κs2=κi2=0.06 to a low value of κp2=κs2κi2=0.01, with intermediate steps of 0.005. Since at mid-infrared operation wavelengths the multi-photon absorption and the associated free carrier absorption are negligible in silicon, the free carrier lifetime is not of importance here and the only losses that need to be taken into account in Γ{p,s,i} are the linear losses.
Since the same type of nanowire are considered for the mid-infrared-pumped CLD parametric ring converter configuration, for this converter the parameter values are adopted as described above, except that in this case ρ(θ)=1 along the ring. Furthermore, different values are considered for the ring circumferences. It is also pointed out that for the CLD converter it is considered that the pump and signal waves to be at ring resonances and the idler wave to be detuned from ring resonance with the detuning given by k L=−ΔkL+2πs.
Equations (6) to (9) are numerically solved for the two mid-infrared-pumped parametric converter setups using a signal input power of (Is,in=2.5×109 W/m2) and a fixed pump input power of 40 mW (Ip,in=1×1011 W/m2) . The reason for taking a fixed pump input power is that in this comparison the pump-power-dependent nonlinear losses are negligible for both of the converters. The signal-to-idler conversion efficiencies of the QPM converter and of the CLD converter with different ring circumferences are shown in
When comparing the graphs of
In conclusion, the QPM parametric conversion method offers a feasible and competitive solution when efficient conversion needs to be achieved in the presence of a large-valued Δklineair, i.e. in the presence of a large-valued GVD at the pump wavelength and/or a large frequency difference between pump and signal. The predicted QPM parametric conversion efficiencies of the order of −33 dB, −29 dB, and −26.7 dB in the near- and mid-infrared spectral domains are high enough to generate microwatts of idler output power, which is a sufficiently high power level for the considered application domains such as spectroscopy. Finally, one has to keep in mind that this QPM parametric conversion method only offers efficient conversion for one specific set of pump, signal, and idler wavelengths, as the ring circumference has to be chosen in function of the phase mismatch between these wavelengths. So, the use of the QPM parametric conversion method presented here should be considered in the following context: in case one works with relatively small wavelength spacings yielding moderate |Δklinear| values, one can rely on CLD parametric conversion, but at the specific set of (widely spaced) wavelengths for which the ring circumference allows quasi-phase-matching, one gets due to QPM parametric conversion a much larger conversion efficiency “for free.” Therefore, if this specific set of wavelengths is often used in the application under consideration, the QPM parametric conversion method presented here can be of great value.
In yet another particular embodiment, reference is made to a QPM SFG system based on a spiral-shaped silicon waveguide covered by a graphene sheet through which an electrical current is flowing. In this embodiment the performance is calculated for a QPM SFG system based on a spiral-shaped silicon waveguide covered by a graphene sheet. To do this, first a modeling formalism for SFG converters is introduced. Without restricting the general validity of the results, focus is made on quasi-continuous-wave operation. The equations expressing the steady-state spatial variation of the slowly-varying pump, signal and idler field amplitudes Ap(ζ), As(ζ), Ai(ζ) (in the parametric SFG-based converter are given by:
where ζ represents the propagation distance along the spiral, deff is the effective second-order nonlinearity, and A{p,s,i} normalized such that 2ϵ0n{p,s,i}C|A{p,s,i}|2 corresponds to intensity. The function ρ(θ) defines the variation of the second-order susceptibility along the graphene-covered silicon spiral as experienced by the TE-polarized fields, and is, as specified earlier on, given by ρ(θ)=cos θ with θ defined as the angle between the local field polarization and the direction of the current flow. The terms in Eqs. (10)-(12) containing eiΔk
When sending a current density of 103 A/m through the graphene sheet, deff˜100×10−12 m/V. For the remaining device parameters the following values were taken: waveguide modal area Aeff=0.5 μm, linear loss α=50 dB/cm, effective two-photon absorption coefficient β=25×10−11 m/W, effective free carrier absorption efficiency φ=6×10−12, effective free carrier lifetime τeff=0.5 ns, Ip,in=2×1011 W/m2. It is pointed out that the relatively small effective free carrier absorption efficiency is due to the fact that at the considered pump wavelength only the graphene sheet contributes to free carrier generation, and only a small fraction of these free carriers effectively diffuse to the silicon waveguide. One then can numerically solve equations (10) to (12) for the QPM SFG converter.
In still another particular embodiment, reference is made to a QPM SFG system based on a spiral-shaped silicon waveguide covered by a monolayer of MoS2, which is a type of transition metal di-chalcogenides. In this embodiment the performance is calculated for a QPM SFG system based on a spiral-shaped silicon waveguide covered by a monolayer of MoS2. To do this, we employ the same modeling formalism for SFG converters as introduced in the previous embodiment.
In this formalism, the function ρ(θ) defining the variation of the second-order susceptibility along the spiral covered with MoS2 as experienced by the TE-polarized fields is given by ρ(θ) =cos θ(4 cos2θ−3) with θ defined as the angle between the local field polarization and an armchair direction of the monolayer MoS2 crystal. As such, the QPM condition becomes
with s=±3. Since monolayer MoS2 has, as opposed to graphene, no inversion symmetry, it is not needed here to send a current through the monolayer to induce a second-order nonlinearity deff. In this embodiment one considers for the QPM SFG converter configuration a TE-polarized pump input and a TE-polarized signal input which both have the same wavelength: λp=λs=2.34 μm. This leads to a generated idler wave with wavelength λi=1.17 μm. The system may have a structure as illustrated in
For the considered configuration, deff˜5×10−12 m/V. For the remaining device parameters the same values were taken as in the previous embodiment, except the losses which we take α=5 dB/cm and β=0.5×10−11 m/V. The reason for having lower losses here than for the graphene-based SFG converter of the previous embodiment is that MoS2, as opposed to graphene, features a bandgap, and both the pump and idler photon energies that we consider here are below the bandgap energy of MoS2. After numerically solving equations (10) to (12) for this QPM SFG converter, one obtains an idler intensity distribution as shown in
In still another particular embodiment, reference is made to a QPM DFG system based on a spiral-shaped silicon waveguide covered by a graphene sheet through which an electrical current is flowing. In this embodiment the performance is calculated for a QPM DFG system based on a spiral-shaped silicon waveguide covered by a graphene sheet. To do this, first a modeling formalism for DFG converters is introduced. Without restricting the general validity of the results, focus is made on quasi-continuous-wave operation. The equations expressing the steady-state spatial variation of the slowly-varying pump, signal and idler field amplitudes Ap(ζ), As(ζ), Ai(ζ) in the parametric DFG-based converter are given by:
where ζ represents the propagation distance along the spiral, deff is the effective second-order nonlinearity, and A{p,s,i} is normalized such that 2ϵ0n{p,s,i}c|A{p,s,i}|2 corresponds to intensity. The function ρ(θ) defines the variation of the second-order susceptibility along the graphene-covered silicon spiral as experienced by the TE-polarized fields, and is, as specified earlier on, given by ρ(θ) =cos θ with θ defined as the angle between the local field polarization and the direction of the current flow. The terms in Eqs. (13)-(15) containing eiΔk
When sending a current density of 103 A/m through the graphene sheet, deff˜100×10−12 m/V. For the remaining device parameters the following values were taken: waveguide modal area Aeff=0.5 μm2, linear loss α=50 dB/cm, effective two-photon absorption coefficient β=25×10−11 m/W, effective free carrier absorption efficiency φ=6×10−12, effective free carrier lifetime τeff=0.5 ns, Ip,in=2×1011 W/m2, Is,in=2×108 W/m2. As in embodiment 3, the relatively small effective free carrier absorption efficiency is due to the fact that at the considered pump wavelength only the graphene sheet contributes to free carrier generation, and only a small fraction of these free carriers effectively diffuse to the silicon waveguide. One then can numerically solve equations (13) to (15) for the QPM DFG converter.
i.e. it can also work effectively for pump, signal and idler wavelengths not exactly equal to but around 1.25 μm, 2.1 Ξm, and 3.088 μm, respectively.
In still another particular embodiment, reference is made to a bent structure being adjacent spiral-shaped waveguides made of a uniform three-dimensional material covered with graphene that is locally removed along a pie-shaped pattern with micron-scale resolution. The structure 1300 is shown in
The two-dimensional or quasi-two-dimensional material used in the present embodiment is again graphene. In the present embodiment, only part of the waveguides have a local periodic graphene coverage as a way for establishing quasi-phase-matching. Such a local, periodic coverage is realized by patterning the graphene e.g. with laser ablation without damaging the underlying waveguide. In the present example, a pie-shape-patterned graphene layer is created, whereby the light waves traveling along the bent structure, in the present example a spiral shaped waveguide design, will periodically experience the presence of the graphene top layer.
Similar as in other embodiments, the device results in the possibility for nonlinear optical wave mixing with quasi-phase-matching.
In yet another particular embodiment, a device is described comprising a bent structure with on top a graphene layer that is electrically or chemically modified. Such a structure is shown in
In this further particular embodiment, the wideband wavelength conversion in a foundry-compatible 220 nm-thick SOI waveguide combined with graphene is further explored. The graphene deposition enables for the first time QPM for Kerr FWM along the ‘ideal’ scheme of nonlinearity sign reversal. Making use of the tunability of the graphene properties through control of its chemical potential μ,it is indeed possible to periodically reverse the sign of the effective FWM nonlinearity experienced by the TE waveguide modes along their propagation path, enabling for the first time QPM of FWM processes employing the ‘ideal’ and most effective scenario. In this embodiment, as an example the wavelength conversion performance of a spiral-shaped graphene-covered foundry-compatible SOI waveguide in the QPM regime was simulated, and compared with the performance attainable in the bare SOI waveguide. The structure is shown in
By way of illustration, the theoretical principles applicable for the exemplary embodiment are given below. The spatial evolution of the pump, signal and idler fields at frequencies v(p,s,i) or wavelengths λ(p,s,i) in a graphene covered SOI waveguide converter is governed by nonlinear propagation equations that describe Kerr-nonlinear interactions, including not only the FWM conversion process but also Kerr-induced phase modulation. Raman nonlinear interactions are excluded since one can assume that the FWM transitions are detuned far away from the Raman resonances of graphene. Furthermore, focus is put on (quasi-) continuouswave operation in the strong-pump approximation, and it is assumed that all fields are TE polarized. Finally, although a graphene-covered SOI waveguide is a heterogeneous medium, the fields will be described in the same way as those in a homogeneous waveguide and effective waveguide parameters are introduced to take into account the medium's hybrid structure. As such, one can use the following set of generic equations for describing the steady-state spatial evolution of the slowly varying pump, signal, and idler field amplitudes Ap(ζ), As(ζ), Ai(ζ):
where ζ is the spatial coordinate along the waveguide, and where Ap,s,i(ζ) are normalized such that |Ap,s,i|2 corresponds to power. The FWM terms are those that contain both the linear phase mismatch_
Δk=−2kp+ks+ki
(with kp,s,i being the pump, signal and idler wave numbers), and the effective coefficients
comprising contributions from the SOI waveguide and the graphene sheet. In Eq. (16) the effective coefficient contains γS,Si accounting for self-phase modulation and two-photon absorption at the pump wavelength in the SOI waveguide, and also comprises γS,g capturing the corresponding phenomena in the graphene layer. In Eqs. (17)-(18) the effective coefficients
account for the linear losses in the SOI waveguide and the graphene sheet. All the effective parameters are function of the spatial coordinate ζ, as is required for QPM operation. The last term in Eqs. (16)-(18) represents free-carrier effects with σFCA and σFCI coefficients quantifying the efficiency of free-carrier
absorption and free-carrier index change, respectively, and with ω(p,s,i)=2πν(p,s,i) and ωr2πc/(1550 nm) where c indicates the speed of light. The factor
with h Planck's constant. The first term at the right-hand side of Eq. (19) represents the free-carrier generation induced by two-photon absorption in the SOI waveguide and the corresponding absorption contribution in the graphene top layer, and the second term indicates the free-carrier generation induced by one-photon absorption in the graphene layer only. The graphene contribution to these two terms has been expressed in a rather phenomenological way since not all photons absorbed in the graphene sheet give rise to the creation of free carriers and instead can contribute to, amongst others, intra-band transitions. The factor τeff in Eq. (19) indicates the effective free-carrier lifetime. Because of the short free-carrier lifetime in graphene, only the graphene-generated free carriers that diffuse into the silicon waveguide will effectively contribute to the free-carrier effects. As such, in the numerical simulations it is allowed to employ the values for τeff, σFCA and σFCI of the bare SOI waveguide, and incorporate the contribution from the graphene-generated free carriers by including a graphene-to-SOI diffusion ratio rD in Eq. (19). This ratio quantifies the fraction of the free carriers generated in the graphene that diffuses into the SOI waveguide. Finally, the factor A′ in Eq. (19) represents the waveguide crosssectional area over which the free carriers are distributed. Turning now to the FWM terms in Eqs. (16)-(18), both the linear phase mismatch_
k=−2kp+ks+ki
and the nonlinear phase mismatch contribution, function of Re(
Δktotal=β2Δωpk2+(1/12)β1Δωps4+2Re(
with βi representing the i-th order dispersion at the pump wavelength and with ωps=2π|νs−νp|.
For signal and idler wavelengths far away from the pump wavelength, the total phase mismatch can become very large in absolute value, hence inducing a change in the fields' phase relation so that the term cos(−Δktotal ζ) determining the idler power evolution will periodically evolve along the waveguide from cos(0)=1 to cos(π)=−1 and back. As a result, there will be a reversal of the conversion process with idler photons being annihilated, deteriorating the net idler growth. The latter can be overcome by employing QPM, so that for a discrete set of signal and idler frequencies very far away from the pump frequency one can also attain an efficient idler growth. Generally speaking, QPM aims at periodically compensating the phase-mismatch-induced change in the fields' phase relation while avoiding the detrimental reversal of the conversion process, so that an overall efficient growth of the idler power is achieved. Ideally, in the sections where the sign of cos(−Δktotal ζ) has changed, one should also have a sign reversal in the Kerr nonlinearity, so that both sign changes annihilate each other and the idler continues to grow.
As mentioned, the different effective coefficients in Eqs. (16)-(18) take into account the contributions from both the SOI waveguide and the graphene top layer. The impact of the latter on both the loss parameters and the nonlinear parameters can be very strong. Experimental investigations with isolated graphene at photon energies above the one-photon absorption onset (i.e. hν>2 |μ|) have revealed extremely high values for the nonlinearity of the two-dimensional material. Our recent calculations indicate that, when the chemical potential μ is tuned such that the photon energy is just below the onset of one-photon absorption, the nonlinearities also become very strong due to the presence of a resonance peak. This is not surprising, since conventional semiconductors also exhibit strong nonlinearity just below the threshold for single photon absorption. In addition, when moving from the one-photon to the two-photon absorption threshold a sign change occurs in the graphene nonlinearities, which is also in line with the behavior of the nonlinearities of direct-bandgap semiconductors. For fixed photon energies this sign change can be controlled by changing μ. Hence, by spatially varying the chemical potential μ of a graphene layer on top of an SOI waveguide using e.g. locally deposited electrolyte gates, it should be possible to establish QPM conversion along the ‘ideal’ scheme where the nonlinearity periodically reverses sign along the propagation path.
To numerically investigate the attainable conversion efficiency in QPM regime, we considered as a case study a foundry-compatible 220 nm-thick SOI waveguide converter shaped as a double spiral and covered with graphene. As shown in
Another important asset of the spiral design is that in the QPM regime the periodicity of the positive nonlinear and negative nonlinear waveguide sections is chirped. As such, this converter design allows QPM operation not just for one discrete signal wavelength far away from the pump wavelength, but for a continuous band of signal wavelengths, hence enabling truly wideband conversion.
In what follows, the design parameters of the targeted graphene-covered SOI waveguide converter were determined. To find the optimal values for the graphene chemical potential |μ|, graphs were used generated based on a theory for both the linear and nonlinear conductivities of graphene. This way it was determined at which |μ|—values both low linear absorption and strong nonlinear effects were seen. Our theoretical curves for the linear conductivity are in line with experimental data and with the widely used Kubo-formalism, so that we can directly implement these theoretical data in our numerical simulations for the converter. In contrast, the theoretical curves for the nonlinear conductivity are systematically lower than what has been experimentally observed, so we take them only as a qualitative guide. We use them only to estimate at which values of |μ| with low linear absorption we can expect nonlinearities as strong as those observed experimentally at |μ|—values where the linear absorption was high. We then take these values of |μ| with small linear absorption as our working point, and implement the experimentally determined values of the nonlinearity in our numerical simulations.
As a case study, we consider QPM conversion between λs=2370 nm (i.e. a signal photon energy of 0.523 eV) and λi=1236.4 nm (i.e. an idler photon energy of 1.003 eV), while pumping at λp=1625 nm. For the converter of
We now turn to the properties of the SOI spiral waveguide. In the QPM regime the sections where the sign of cos(Δktotal ζ) has changed should correspond to the sections where the FWM nonlinearity has a reversed sign as well. For a converter configuration as in
with Ravg representing the average radius of the spiral-shaped waveguide. For the pump, signal and idler wavelengths under consideration and for an SOI waveguide width of 670 nm and height of 220 nm, the linear phase mismatch Δk in the graphene covered waveguide equals 4.4×104 m−1, for which Eq. (21) yields Ravg=23 μm.
Using the graphene parameter values specified above, we obtain for the effective parameters the values in Table 1. To calculate the attainable conversion efficiencies in the QPM regime for the considered graphene-covered SOI waveguide converter and for its bare counterpart, we solve Eqs. (16)-(19) for signal and idler wavelengths of 2370 nm and 1236.4 nm, and for pump input powers Pp,in ranging between 10 mW and 900 mW, and we plot the corresponding conversion efficiencies in
Although we have deliberately plotted the best-case scenario conversion efficiency for the bare SOI converter in
Whereas the above aspect has been mainly described with reference to system features, as indicated it also relates to a method for obtaining conversion or amplification, using QPM nonlinear optical wave mixing. Such a method comprises receiving a pump radiation beam and a signal radiation beam in a bent structure, a waveguiding portion of the bent structure being made of a uniform three-dimensional material at least partly covered by a two-dimensional or quasi-two-dimensional material layer, and the dimensions of the bent structure being selected for obtaining QPM nonlinear optical wave mixing. The method also comprises obtaining an idler radiation beam by interaction of the pump radiation beam and the signal radiation beam using at least one QPM nonlinear optical process such as for example a QPM SFG, a QPM DFG, a QPM Raman-resonant FWM or QPM Kerr-induced FWM process. The method furthermore encloses coupling out an idler radiation beam from the bent structure. Other or more detailed method steps may be present, expressing the functionality of components of the system as described above.
In one aspect, the present invention also relates to a method for designing a converter or amplifier using QPM nonlinear optical wave mixing. The converter or amplifier thereby may be using a pump radiation beam and a signal radiation beam. The method for designing comprises selecting a bent structure suitable for QPM nonlinear optical wave mixing, comprising selecting materials for a radiation propagation portion of the bent structure, e.g. a waveguide, and selecting dimensions of the bent structure taking into account the spatial variation of the nonlinear optical susceptibility along the structure as experienced by radiation travelling along the bent structure. At least one dimension of the bent structure are selected such that QPM nonlinear optical wave mixing is obtained. More particularly, at least one dimension of the radiation propagation portion of the bent structure is selected taking into account the spatial variation of the nonlinear optical susceptibility along the radiation propagation structure as experienced by radiation travelling along the bent structure for obtaining quasi-phase matched nonlinear optical wave mixing in the radiation propagation portion. The dimension may be substantially inverse proportional with the linear phase mismatch for the nonlinear optical wave mixing. The method for designing furthermore may be adapted so that the structure provides cavity enhancement for at least one of the radiation beams that will travel in the system, i.e. for which the system is designed, preferably more or all of the radiation beams are cavity enhanced.
In a further aspect, the above described methods for designing or controlling a system for resonating, converting or amplifying using QPM nonlinear optical wave mixing e.g. the controller may be at least partly implemented in a processing system 500 such as shown in
The present invention also includes a computer program product which provides the functionality of any of the methods according to the present invention when executed on a computing device. Such computer program product can be tangibly embodied in a carrier medium carrying machine-readable code for execution by a programmable processor. The present invention thus relates to a carrier medium carrying a computer program product that, when executed on computing means, provides instructions for executing any of the methods as described above. The term “carrier medium” refers to any medium that participates in providing instructions to a processor for execution. Such a medium may take many forms, including but not limited to, non-volatile media, and transmission media. Non-volatile media includes, for example, optical or magnetic disks, such as a storage device which is part of mass storage. Common forms of computer readable media include a DVD, a USB-stick, or any other medium from which a computer can read. Various forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to a processor for execution. The computer program product can also be transmitted via a carrier wave in a network, such as a LAN, a WAN or the Internet. Transmission media can take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications. Transmission media include coaxial cables, copper wire and fibre optics, including the wires that comprise a bus within a computer.
It is to be understood that although preferred embodiments, specific constructions and configurations, as well as materials, have been discussed herein for devices according to the present invention, various changes or modifications in form and detail may be made without departing from the scope and spirit of this invention. For example, any formulas given above are merely representative of procedures that may be used. Functionality may be added or deleted from the block diagrams and operations may be interchanged among functional blocks. Steps may be added or deleted to methods described within the scope of the present invention.
By way of illustration, in
| Number | Name | Date | Kind |
|---|---|---|---|
| 7200308 | Hochberg et al. | Apr 2007 | B2 |
| 7532656 | Yang et al. | May 2009 | B2 |
| 8873133 | Vermeulen | Oct 2014 | B2 |
| 9223187 | Vermeulen | Dec 2015 | B2 |
| 20040233511 | Kurz et al. | Nov 2004 | A1 |
| 20060092500 | Melloni et al. | May 2006 | A1 |
| 20060132901 | Miller | Jun 2006 | A1 |
| 20060159398 | Knox et al. | Jul 2006 | A1 |
| 20130010351 | Vermeulen et al. | Jan 2013 | A1 |
| 20160161675 | Englund | Jun 2016 | A1 |
| 20170146738 | Xu | May 2017 | A1 |
| 20170199443 | Vermeulen | Jul 2017 | A1 |
| Entry |
|---|
| Bonaccorso et al., “Solution Processing of Graphene, Topological Insulators and Other 2d Crystals for Ultrafast Photonics”, vol. 4, No. 1, Jan. 1, 2014. 16 Pages. |
| Donnelly et al., “Ultra-Large Nonlinear Parameter in Graphene-silicon Waveguide Structures”, vol. 22, No. 19, Sep. 22, 2014, 12 Pages. |
| Gu et al., “Coherent Four-Wave Mixing on Hybrid Graphene-Silicon Photonic Crystals”, IEEE Journal of Selected Topics in Quantum Electronics, vol. 20, No. 1, Jan./Feb. 2014, 6 Pages. |
| Kelvin J.A. et al., “Waveguide Engineering of Graphene's Nonlinearity”, AIP Publishing LLC, Applied Physics Letters, vol. 105, 2014, 4 Pages. |
| Moss et al., “New CMOS-Compatible Platforms Based on Silicon Nitride and Hydex for Nonlinear Optics”, Nature Photonics, vol. 7, Aug. 2013, 12 Pages. |
| Wang et al., “Broadband Ultrafast Nonlinear Absorption and Nonlinear Refraction of Layered Molybdenum Dichalcogenide Semiconductors”, Nanoscale, vol. 6, pp. 10530-10535, 2014. |
| International Search Report from PCT Application No. PCT/EP2015/072049, dated Nov. 27, 2015. |
| Turner et al., “Ultra-Low Power Parametric Frequency Conversion in a Silicon Microring Resonator”, Optics Express, vol. 16, No. 7, Mar. 31, 2008, pp. 4881-4887. |
| Yang et al., “Enhanced Second-Harmonic Generation in AlGaAs Microring Resonators”, Optics Letters, vol. 32, No. 7, Apr. 1, 2007, pp. 826-828. |
| Ferrera et al., “Low Power Parametric Wave-Mixing in a Zero Dispersive CMOS Compatible Micro-Ring Resonator”, IEEE, 2009, pp. 481-482. |
| International Search Report for PCT/EP2010/064750, dated Jan. 14, 2011. |
| Written Opinion for PCT/EP2010/064750, dated Sep. 18, 2012. |
| Vermeulen et al., “Applications of Coherent Anti-Strokes Raman Scattering in Silicon Photonics”, The Institution of Electrical Engineers, vol. 7606, Jan. 24, 2010, (Abstract). |
| Vermeulen et al., “Enhancing the Efficiency of Silicon Raman Converters”, Photonic Integrated Circuits II, vol. 7719, Apr. 12, 2010. (Abstract). |
| Vermeulen et al.,“Cavity-Enhanced Quasi-Phase-Matched Wavelength Conversion in Silicon Ring Resonators: Two Approaches”, 2010 IEEE Photonics Society Summer Topical Meeting Series, vol. 7719, pp. 92-93, Jul. 19, 2010. (Abstract). |
| Number | Date | Country | |
|---|---|---|---|
| 20170199443 A1 | Jul 2017 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 14494658 | Sep 2014 | US |
| Child | PCT/EP2015/072049 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | PCT/EP2015/072049 | Sep 2015 | US |
| Child | 15469097 | US |
