Observation of Backaction and Self-Induced Trapping in a Planar Hollow Photonic Crystal Cavity

The optomechanical coupling between a resonant optical field and a nanoparticle through trapping forces is demonstrated. Resonant optical trapping, when achieved in a hollow photonic crystal cavity is accompanied by cavity backaction effects that result from two mechanisms. First, the effect of the particle on the resonant field is measured as a shift in the cavity eigenfrequency. Second, the effect of the resonant field on the particle is shown as a wavelength-dependent trapping strength. The existence of two distinct trapping regimes, intrinsically particle specific, is also revealed. Long optical trapping ($>10\min$) of 500 nm dielectric particles is achieved with very low intracavity powers ($<120\mu \mathrm{W}$).

Figure 1

Hollow photonic crystal cavity. (a) Computed electric field distribution in a hollow circular cavity in the presence of a 500 nm dielectric particle. (b) Scanning electron micrograph of the PhC device showing both the circular cavity and the coupling waveguide. Inset: $10\times \mathrm{\text{magnification}}$ of the circular defect (700 nm in diameter). (c) Distribution of the electric field in a vertical cross section of the photonic crystal, centered on the particle, as computed with 3D finite elements (COMSOL).

Figure 2

Resonant optical trapping inside the hollow photonic crystal cavity. (a) A single 500 nm fluorescent particle is resonantly trapped inside the photonic crystal cavity region with a waveguide power lower than $120\mu \mathrm{W}$. (b), (c) The particle remains strongly trapped for 10 min while other fluorescent particles are seen in the flow near the trapped particle. (d) The particle immediately leaves the cavity when the excitation laser is turned off after 10 min of continuous trapping.

Figure 3

Demonstration of particle-induced resonance frequency shift. (a) Record of the evolution of the cavity spectrum in time while a particle is trapped in the cavity and after it is released. (b) A single snapshot from (a), displaying an average resonance shift of 1.8 nm from the unloaded cavity resonance. (c) Computed resonance wavelength shifts as a function of particle size as it is moved along the vertical direction from the center of the cavity. The shaded area (in grey) shows the extent of the silicon slab. A maximum shift of 2.9 nm is found for a 500 nm particle placed in the center of the cavity.

Figure 4

Demonstration of particle-cavity coupling. Record of trapping threshold powers as a function of the estimated guided power in the access waveguide (left axis) and the laser diode output power (right axis). Experimental points correspond to escape trapping times of the order of 10 s.

Figure 5

(a) Theoretical computations for a 500 nm particle including the particle-cavity renormalization effect: Trapping forces as a function of the position of the particle for various detuning values along the horizontal axis inside the cavity (from $-100\mathrm{nm}$ to 0) and along the vertical axis (0 to 800 nm). The shaded area corresponds to the half-thickness of the silicon slab. (b) Illustration of the first trapping regime. The eigenmode of the particle located in the cavity is weakly coupled to the excitation laser (${\lambda }_{\mathrm{Ex}1}$). An out-of-plane displacement of the particle couples the eigenmode back to the excitation resulting in strong gradient forces. (c) Illustration of the second trapping regime. A particle outside the cavity weakly couples the eigenmode to the excitation laser (${\lambda }_{\mathrm{Ex}2}$). As soon as the particle enters further into the cavity, the eigenmode is strongly coupled to the excitation and hence increases the trapping strength.