Full text loading...
Scanning electron micrographs (a), (b) and photothermal maps (c), (d) of the spatial dependence of resonance shift for a standard (a), (c) and re-etched (b), (d) toroid. Each map is constructed from 13 × 13 grid with 4 μm increment, where each shift at constant (4.2 mW absorbed) pump power is referenced to the shift with no external pump. An optical micrograph is superimposed on the map, showing the correlation between maximum shift and the silicon pillar.
Photothermal linescans of a standard and re-etched toroid on two scales (a), (b). The laser spot is scanned laterally across the center of the toroid at constant power (4.2 mW absorbed) and the resonance shift measured as a function of laser position. The profile of the silicon pillar is shown for the standard toroid (a). The greater sensitivity of the re-etched toroid is conspicuous in (b). (inset) Optical micrograph taken during a linescan, with the pump laser spot marked.
Power dependence of photothermal shift. The shift is seen to be quadratic for a re-etched toroid and to extend to over half the FSR of the resonator. Simulations capture the quadratic dependence and were performed by calculating equilibrium temperature at the silica rim from the measured pump power and extrapolating resonance shift with Eq. (1) . Error bars are present but too small to see on this scale. (inset) A standard toroid shows a shallower slope and is linear over the range examined.
Simulations of the temperature profile of the toroid upon excitation at its center for a standard toroid (a) and re-etched toroid (b). The heating source is shown in black wireframe (see supplementary material 35 for details of the thermal modelling). Absorbed power is 4.2 mW in both simulations.
Scan of toroid resonance via photothermal heating and Lorentzian fit. The probe wavelength is fixed at 1566.93 nm while the pump laser power is scanned with a triangular waveform. The wavelength is calculated from the measured temperature dependence of the toroid resonance (3200 fm/mW).
Frequency response of a standard toroid during pump beam modulation (circles). The toroid has a cutoff frequency of 4200 Hz and is fit with exponential time dependence (dashed line, 1-e(−t/τ)). Simulation (solid line) predicts a cutoff frequency of 5000 Hz. (inset) Scan of resonance with pump laser off (black) and on (grey).
Article metrics loading...