CLEO:2015 Kou Yoshiki

 [STu1I.4] Low-power on-chip all-optical Kerr switch with silica microcavity

I gave a presentation in the "Nonlinear optics" session on the morning of the 12th. My presentation was on optical switching using optical Kerr effect in silicatroid resonators. In addition to my presentation, there were many other presentations using micro optical resonators in this session. However, most of them were related to wavelength conversion or optical Kerr-comb, and none of them were about time-domain applications like mine. In the same session, there were presentations by Purdue univ./Weiner group, Conrnell univ./Gaeta group, Caltech/Vahala group, and other distinguished groups. There were also many high quality research presentations such as "Brillouin-scattering-induced transparency and non-reciprocal light storage" which was recently published in Nature Communications. There were also many high quality presentations.

My presentation was the fourth one in the session, starting at 8:45 a.m., which was just when many people were gathering. In fact, the session was so crowded that there was a standing-room-only audience due to many high-profile com-related presentations. The following questions were asked during the Q&A session.

• What is the time width of the controlled light pulse in which the thermal effect becomes apparent?
• Is some method used to stabilize the resonance wavelength?
• What is the relationship between the mode used and the control optical power required for the switch (since different modes require different mode volumes)?
• What limits the switch speed?
• Is the time interval of the input pulses determined by considering the FSR of the resonator?

I was able to answer (1) to (4) without any trouble because the questions (1) to (4) were assumed questions (I think), but I could not answer (5) because I did not understand either the English or the content of the question. Looking back on it now, since the question was about FSR and time intervals, I can guess that the questioner must have confused it with the research on optical carcoms, but I could not think that far back then. I would like to be more diligent in answering the question next time, taking the questioner's intention into consideration.

3. presentations that caught our attention

 [SM1l.4] Integragted on-chip C-band optical spectrum analyzer using dual-ring resonator

This is a study of a spare-like functionality on a silicon chip. The wavelength-decomposing elements are made of coupled silicon microrings with slightly different FSRs; since the FSRs are slightly different, light can basically only be transmitted from a pair of peaks with matching wavelengths. By thermally tuning the resonance wavelength of one of the microrings, the pair of modes with matching resonance wavelengths changes, allowing the wavelength to be swept. The key point is that a small change in the resonance wavelength of one of the micro rings can result in a large wavelength sweep. The device has a built-in Mach-Zehnder modulator at the input port to enable lock-in detection, and a built-in photodetector. Although the use of a microring has the disadvantage that the wavelength resolution cannot be increased much, it was interesting to see the above-mentioned detailed devices.

[STu1I.3] Highly Efficient Four-Wave Mixing in an AlGaAs-On-Insulator (AlGaAsOI) Nano-Waveguide (Technical University of Denmark)

When nonlinear optical phenomena are used in silicon, carrier generation by two-photon absorption is always a problem (unless carrier extraction is used). In recent years, SiN and a-Si:H have been investigated as alternative materials to silicon. However, although SiN has a wide band gap, its nonlinearity is lower than that of silicon. Although a-Si:H has higher nonlinearity than silicon, it cannot completely suppress carrier generation by two-photon absorption. On the other hand, AlGaAs used in this study has higher nonlinearity than silicon, and its band gap can be controlled by the Al concentration. The wavelength conversion efficiency of AlGaAs has been improved by devising the waveguide structure.

[FTu4B.8] Controlling Carbon Nanotube Mechanics with Optical Microcavities (Lipson, Cornell univ.)

Free-standing resonator with silicon nitride (Q=5×10⁶CNTs are thermally vibrating, and the amplitude of the vibration is as large as pm. Therefore, by bringing the CNTs into close proximity to the resonator, the micro-displacement of the CNTs due to the mechanical vibration can be detected via the output from the resonator. Lipson has also published work on graphene-based modulators ([SW4I.4.] 30 GHz Zeno-based Graphene Electro-optic Modulator), and is clearly starting to move into carbon-based materials.

[SM3O.1] Single Nucleic Acid Interactions Monitored with Optical Microcavity Biosensors (Vollmer, MPI)

Invited talk by Vollmer. The main points of this presentation are as follows. First, as was discussed in the workshop held at Keio University last year, prisms are used for the coupling instead of tapers. This is because of its high mechanical stability. Also, we use microspheres instead of toroids to facilitate prismatic coupling. Another point is the use of plasmon microparticles to enhance the electric field and increase sensitivity. As described below, the recent trend in sensing using micro optical resonators seems to be the combination of plasmons, optomechanics, and optofluidics to achieve higher sensitivity and functionality.

[AW3K.1] Fluid Coupled Optomechanical Oscillators (H. Tang, Yale Univ.)

[AW3K.2] Surface Sensitive Microfluidic Optomechanical Ring Resonator Sensors (X. Fan, Michigan Univ.)

The principle of sensing by Optomechanics is very simple. When some particles adhere to the resonator, the effective mass of the resonator changes and the mechanical resonance frequency shifts. By injecting CW light into the resonator and frequency-resolving the output light with an RF spectrometer, the shift in mechanical frequency can be seen. This method has an advantage over conventional sensing methods that use refractive index changes in that it can measure "weight.

The former uses a SiN disk resonator integrated with a waveguide. Water is transparent to visible light, but Si is opaque. Therefore, they fabricated the resonator using SiN, which is transparent in the visible light range. The measurement in liquid is performed by fabricating a flow path, but the problem is the damping caused by the liquid. In liquid, the Q_MegaThe damping in the liquid is said to be reduced to about ~1. In this study, the damping in the liquid is canceled by amplifying optomechanical oscillations with a sufficiently powerful light. As a result, the Q_MegaThe value of ~12, which is extremely high for liquid, was obtained. The input/output of light is performed using a grating coupler.

The latter uses a hollow silica bottle resonator (similar to OIST?). The latter uses a hollow silica bottle resonator (similar to OIST?). This one has a flow channel in the silica bottle, so the taper is not wetted by the liquid. In this study, the target of measurement is HF solution, and the change in effective mass due to the scrapping of the inner wall of the silica bottle by HF is observed.

[STu2I.3] Cascaded four-wave mixing in silicon-on-sapphire microresonators at λ=4.5 μm (Loncar, Harvard Univ.)

[SW4F.2] Quantum cascade laser-based Kerr frequency comb generation (Kippenberg, EPFL)

Finally, I would like to introduce some presentations related to car combs, although I am not very familiar with this field. The above two studies attempted to generate mid-IR combs using QCL as a light source, which is useful for sensing the absorption lines of various gases in the mid-IR region. The former (Harvard) attempted comb generation using Si microrings fabricated on sapphire. The reason for growing on sapphire instead of silica is that silica has absorption in Mid-IR. On the other hand, the latter (EPFL) uses MgF2 resonators as well as combs in the telecommunication wavelength band, but the silica tapers with Mid-IR absorption cannot be used for coupling, so tapered fibers are fabricated from chalcogenide fibers. Prism coupling would have been an option, but was there no material for prisms that is transparent in Mid-IR? In any case, it is necessary to devise some way to produce combs in Mid-IR.