Optical-fiber-based measurement of an ultrasmall volume high-$Q$ photonic crystal microcavity

A two-dimensional photonic crystal semiconductor microcavity with a quality factor $Q\sim \mathrm{40,000}$ and a modal volume $V_{\mathrm{eff}}\sim 0.9$ cubic wavelengths is demonstrated. A micron-scale optical fiber taper is used as a means to probe both the spectral and spatial properties of the cavity modes, allowing not only measurement of modal loss, but also the ability to ascertain the in-plane localization of the cavity modes. This simultaneous demonstration of high-$Q$ and ultrasmall $V_{\mathrm{eff}}$ in an optical microcavity is of potential interest in nonlinear optics, optoelectronics, and quantum optics, where the measured $Q$ and $V_{\mathrm{eff}}$ values could enable strong coupling to both atomic and quantum dot systems.

Figure 1

(Color online) Design of the photonic crystal (PC) membrane microcavity and experimental measurement setup. (a) Schematic of the undercut, two-dimensional PC microcavity geometry. (b) Magnetic field amplitude $\left(\mid \mathbf{B}\mid \right)$ in the center of the PC membrane for the $A_2^0$ mode. The dashed curves show the grade in hole radius $(r∕a)$ along the central $\widehat{x}$ and $\widehat{y}$ axes of the cavity (marked by dashed lines through the dielectric lattice), and the solid curves are slices of the field component $B_z$ along these directions. The hole radius $\left(r\right)$ varies between $r∕a=0.23$ in the center to $r∕a=0.35$ at the edges of the cavity, and for the mode of interest, $a∕{\lambda }_c=0.245$. (c) Schematic of the fiber taper probe measurement setup.

Figure 2

SEM micrographs of a fully fabricated PC microcavity. (a) Cross-sectional view. (b) Top view of the portion of the cavity contained within the dashed lines in (a). Total cavity dimensions are $\sim 13\mu \mathrm{m}\times 16\mu \mathrm{m}$. (c) Zoomed in angled view of the dashed line region in (b) showing the smoothness and verticality of the etched air hole sidewalls, necessary to limit scattering loss and radiative coupling to TM-like modes.

Figure 3

(Color online) Fiber taper transmission measurements of the PC microcavities. (a) Taper transmission spectrum for a cavity with $a=409\mathrm{nm}$, where the data have been normalized to a background spectrum when the taper is far above the cavity. The highlighted long wavelength mode is the $A_2^0$ resonant mode. (b) Measured linewidth (dots) versus taper-cavity gap ($\Delta z$) for the $A_2^0$ mode $(a∕{\lambda }_c\sim 0.263)$ in a sample with $a=425\mathrm{nm}$. The taper is vertically positioned by a stepper motor with $50\mathrm{nm}$ encoder resolution. The solid curve is a fit to the experimental data. (inset) Normalized taper transmission versus wavelength when the taper is $650\mathrm{nm}$ above the cavity surface.

Figure 4

(Color online) Mode localization data. Measured normalized taper transmission depth (dots) as a function of taper displacement along the (a) $\widehat{x}$ axis and (b) $\widehat{y}$ axis of the cavity. The dashed line in (a)-(b) is a Gaussian fit to the data while the solid line is a numerically calculated coupling curve based upon the FDTD cavity field and analytically determined fundamental fiber taper mode (taper diameter $\sim 1.7\mu \mathrm{m}$). The Gaussian fit to the data gives a full-width at half-maximum (FWHM) of 1.2 and $1.3\mu \mathrm{m}$ along the $\widehat{x}$ and $\widehat{y}$ axes, respectively. The insets in (a)-(b) are optical micrographs of the taper aligned along the $\widehat{y}$ and $\widehat{x}$ axes of the cavity, respectively. The cavity is the central rectangular region.