Suspension-Free Integrated Cavity Brillouin Optomechanics on a Chip

Abstract

Cavity optomechanical systems enable coherent photon-phonon interactions essential for quantum technologies, yet high-performance devices have been limited to suspended structures. Here, we overcome this limitation by demonstrating cavity Brillouin optomechanics in a suspension-free racetrack microring resonator on a lithium-niobate-on-sapphire chip, a platform that merits high stability and scalability. We demonstrate coherent coupling between telecom-band optical modes and a 9.6-GHz phonon mode, achieving a maximum cooperativity of $0.41$ and a phonon quality-factor-frequency product of $10^{13}\,\mathrm{Hz}$. The momentum-matching condition inherent to traveling-wave Brillouin interactions establishes a one-to-one mapping between optical wavelength and phonon frequency, enabling multi-channel parallel operations across nearly $300\,\mathrm{MHz}$ in phonon frequency and $40\,\mathrm{nm}$ in optical wavelength. Our suspension-free architecture provides a coherent photon-phonon interface compatible with wafer-scale integration, opening pathways toward hybrid quantum circuits that unite photonic, phononic, and superconducting components on a single chip.

Figures (4)

• Figure 1: (a) Schematic of the racetrack microring resonator for cavity Brillouin optomechanics (BOM). Red, orange, and green arrows denote the optical pump ($a_{\text{r}}$), probe ($a_0$), and phonon ($b_{\text{cw}}$) mode, respectively. Inset: ridge waveguide cross-section. Green dashed boxes mark the straight sections deminating BOM coupling. (b) Mode profiles of traveling-wave optical (electric field) and phonon (displacement field) modes. (c) Transmission spectrum of three participated optical modes. The red line shows the fitting result, with intrinsic and external quality factors are $2.3\times10^5$ and $2.8\times10^5$, respectively. (d) Energy level diagram for BOM interaction between optical signal ($a_0$), clockwise ($b_\mathrm{cw}$) and counter-clockwise ($b_\mathrm{ccw}$) phonon modes, under the resonantly-enhanced ($a_{\text{r}}$ and $a_{\text{b}}$) optical pump fields. OMIA/OMIT: optomechanically induced amplification/transparency; $\kappa$, $\gamma$: optical and phonon decay rates; $g_0$: vacuum BOM coupling strength.
• Figure 2: (a) Experimental setup for BOM characterization. EDFA: erbium-doped fiber amplifier; PC: polarization controller; PD: photo-detector; RF source: radio-frequency source; SSBM: single-sideband modulator. (b) and (c) Schematics and measured spectra of OMIA and OMIT, respectively, with an on-chip pump power of $121\,\mathrm{mW}$. Red lines denote the fits; red, blue, and orange arrows represent OMIA pump, OMIT pump, and probe light, respectively; gray dashed lines indicate the probe sweep range. (d) Mode profiles of phonon modes $A$, $B$, $C$ and $D$, corresponding to labeled peaks in (b) and (c).
• Figure 3: (a) OMIA and OMIT spectra with different pump powers. On-chip $P_{\text{pump}} = 40.01\,\text{mW}$ for (I) and (II), and $121.41\,\text{mW}$ for (III) and (IV). (b) Dependence between cooperativity $C$ and $P_{\text{pump}}$ extracted from OMIA and OMIT spectra, respectively. The dots represent the fitted $C$, and the error bars represent the corresponding fitting errors. The solid lines represent the linear fits. (c) OMIA and OMIT spectra of modes with different optical wavelength $\lambda$. (d) Frequency mismatch between FSR and phonon mode ($\Omega$) and corresponding $C$ vary with optical wavelength. Blue and red dots: the fitted results of OMIA and OMIT, respectively. Solid line: a linear fit. Orange dots: fitted C factor.
• Figure 4: (a) Experimental setup for noise power spectrum density (PSD) characterization. ESA: electrical spectrum analyzer. (b) Normalized PSD of the phonon mode with varied $P_{\text{pump}}$. Solid lines with dark and light colors represent experimental and fitting results, respectively. (c) Full-width-at-half-maximum (FWHM) varies with $P_{\text{pump}}$. The blue dots and solid line represent the experimental and fitting results, respectively. (d) Normalized PSD with different optical wavelength. (e) Dependence of the phonon mode frequency on optical wavelength, extracted from noise PSD spectra (solid blue dots) and OMIT spectra (hollow blue dots), respectively. The shaded region corresponds to the simulated results, for which we calculate structures with a waveguide width of $1.2\pm0.1\,\mu \text{m}$ to factor in experimental fabrication tolerances.