Inverse-designed release-free optomechanical crystal with high photon–phonon coupling Preprint: APS/123-QED

David Hambraeus, Paul Burger, Johan Kolvik, Philippe Tassin, Raphaël Van Laer · 2026

Read Inverse-designed release-free optomechanical crystal with high photon–phonon coupling Preprint: APS/123-QED by David Hambraeus, Paul Burger, Johan Kolvik, Philippe Tassin, Raphaël Van Laer (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-03910.txt.

Introduction

David Hambraeus and four co-authors demonstrate an inverse-designed optomechanical crystal that largely closes a known performance gap. Optomechanical devices couple light to mechanical motion, linking optical and microwave signals. Release-free crystals anchor heat better than suspended nanobeams. But they have historically shown weaker photon-phonon coupling. Strength and thermal robustness seemed to trade off against each other.

The team combined physics-guided human intuition with a multiphysics inverse-design algorithm for resonant structures. They built a release-free silicon crystal with a record vacuum optomechanical coupling rate for its class. That rate is comparable to suspended state-of-the-art devices. Its scattering rate is nearly twice previous release-free results. The design framework itself is presented as reusable for co-optimizing optical and mechanical properties in other devices.

Catalog record

CREATORDavid Hambraeus, Paul Burger, Johan Kolvik, Philippe Tassin, Raphaël Van Laer
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE55 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.03910

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Inverse-designed release-free optomechanical crystal with high photon–phonon coupling Preprint: APS/123-QED David Hambraeus, Paul Burger, Johan Kolvik, Philippe Tassin, Raphaël Van Laer Abstract Interaction between light and mechanics provide a powerful interface between optical and microwave signals, with applications spanning classical signal processing and quantum technologies. High-performance optomechanical devices require both strong photon–phonon coupling and tolerance to parasitic laser heating. Release-free optomechanical crystals provide improved thermal anchoring compared to suspended nanobeams, but have so far exhibited weaker vacuum optomechanical coupling rates, leaving a trade-off between coupling strength and thermal robustness. Here, we largely close this gap: we design and experimentally demonstrate a release-free silicon optomechanical crystal with a record vacuum optomechanical coupling rate of about g OM / ( 2 ​ π ) = 800 kHz g_{\text{OM}}/(2\pi)=$800\text{\,}\mathrm{kHz}$ , comparable to suspended state-of-the-art devices. The resulting optomechanical scattering rate Γ OM / ( 2 ​ π ) = 1.1 kHz \Gamma_{\text{OM}}/(2\pi)=$1.1\text{\,}\mathrm{kHz}$ is nearly twice that of previous release-free implementations. This performance is achieved by combining physics-guided human intuition with a multiphysics inverse-design algorithm introduced here for resonant optomechanical structures. Beyond the specific device demonstrated, the inverse-design framework is applicable to co-optimizing optical and mechanical resonances and eigenmodes more broadly. These results strengthen release-free optomechanical crystals as a platform for fast, low-noise classical and quantum optomechanics. I Introduction Optomechanical interaction in solid-state systems enable strong coupling between optical fields and mechanical motion 3 , with applications ranging from precision sensing 39 ; 38 ; 11 to microwave–optical transduction 26 ; 16 and quantum information processing 41 . These interactions have enabled key demonstrations such as slow light 33 , laser cooling of mechanical resonators to their ground state 5 , and the generation of squeezed states of light and motion 34 . High-performance optomechanical devices require both large photon–phonon coupling rates and robustness against optical absorption–induced thermal noise. Optomechanical crystals (OMCs)achieve strong coupling by co-localizing optical and mechanical modes within wavelength-scale nanostructures 6 ; 45 ; 37 . To suppress mechanical radiation loss, these devices are typically implemented as suspended nanobeams. While suspension enables tight confinement, it also limits thermal anchoring, rendering suspended OMCssusceptible to excess noise under optical pumping 24 ; 4 ; 45 . Several approaches have been explored to mitigate this limitation, including two-dimensional suspended geometries 32 ; 23 ; 45 ; 35 and active cooling using superfluid helium 31 . Release-free optomechanical crystals provide an alternative architecture in which the device remains mechanically attached to the substrate, allowing heat generated by optical absorption to dissipate directly into the substrate 19 . This geometry substantially improves tolerance to optical heating 19 ; 20 . However, first-generation release-free OMCsexhibited vacuum optomechanical coupling rates of approximately g_{\text{OM}}/(2\pi)=$500\text{\,}\mathrm{kHz}$ , leading to optomechanical scattering rates \Gamma_{\text{OM}}=4g_{\text{OM}}^{2}/\kappa , where \kappa is the optical linewidth, below those of state-of-the-art suspended devices 6 ; 32 ; 23 ; 45 . This limitation arose because release-free operation requires a high-wavevector mechanical mode and an optical mode near the half-X-point rather than at a band edge 19 ; 4 – a configuration in which the optical field extends gradually into the mirror, producing sub-optimal spatial overlap with the tightly confined mechanical mode. In this work, we close this gap and demonstrate a release-free silicon optomechanical crystal with a vacuum optomechanical coupling rate of g_{\text{OM}}/(2\pi)=$800\text{\,}\mathrm{kHz}$ , comparable to state-of-the-art suspended devices. The device is realized through two complementary contributions: a physically motivated mode-engineering technique (X-point matching by harmonic period extension, or X-HOPE) that restores spatial overlap between the optical and mechanical modes, and a multiphysics inverse-design algorithm for resonant optomechanical structures that recovers the radiation-limited quality factors X-HOPEalone leaves too low. Combined with the intrinsic light-resilience of the release-free architecture 20 , these results establish release-free OMCsas a platform for fast, low-noise classical and quantum optomechanics, and provide a key ingredient for release-free piezo-optomechanical microwave-to-optical transducers 4 . Figure 1: Release-free optomechanical crystal cavity with improved photon-phonon overlap. (a) The release-free OMCwith X-HOPEdesign pattern showing the transverse component of the normalized electric field E_{x} of the optical mode (top) and the longitudinal normalized displacement u_{y} of the mechanical mode (bottom). (b) Diagram of unit cell geometry. The basic cell is an elliptical hole with major and minor radii r_{x} and r_{y} . The double cell used for the mirror consists of two holes that have been moved toward each other by a distance s . When s is large, the holes are merged to prevent bridges too thin to fabricate. (c) Optical intensity along the center of the nanobeam for a conventional mirror transition 19 and our design. The displacement of the X-HOPEmechanical mode is shown in green with arbitrary units (the mechanical mode shape is very similar for the conventional case). The X-HOPEdesign has an increased optical intensity in the center of the beam, resulting in improved photon-phonon overlap. (d) Unit cell band structure for the double cell (shades of blue) and single cell (dashed orange). The x-axis is showing the k in units of \pi/a , which would be the X-point of the single cell beam with periodicity a . The lighter (right) gray light-cone only appears when considering the double cell. At s=0 , meaning no deviation from regularly spaced holes, the bands of the double cell are the combination of the single cell bands and a folded copy, effectively mirrored about k=0.5\pi/a . For s>0 , a quasi-bandgap opens at k=0.5\pi/a , the new X-point. Figure 2: Inverse-design of the optomechanical cavity enhances the optomechanical scattering rate. (a) A diagram depicting the inverse design algorithm. (i) The geometry is imported into the simulation software. (ii) An eigenmode simulation is performed, yielding the eigenmode fields and frequencies. (iii) The gradient of the figure-of-meritw.r.t. the fields is computed. (iv) An adjoint simulation is performed, yielding the adjoint fields. (v) Combining the adjoint and eigenmode fields, a geometric gradient is obtained. (vi) From the geometric gradient, the gradient w.r.t. the design parameters is computed, which in turn is used to obtain an updated set of parameters and the cycle repeats until performance is satisfactory. (b) Comparison of the defect and transition region before (red) and after (green) optimization. (c) Evolution of simulated optomechanical interaction strength, optical and mechanical quality factors, and the optomechanical scattering rate \Gamma_{\text{OM}} during optimization. II Results II.1 Design II.1.1 X-HOPE design technique Optomechanical crystal cavities enable strong light–sound interactions by confining both optical and acoustic fields to a wavelength-scale region. The central design challenge is to realize mirror regions, where a bandgap prevents the modes’ propagation, and defect regions, where the modes can interact, simultaneously for near-infrared light and gigahertz sound while maintaining strong spatial overlap between the optical and mechanical modes. In addition, release-free OMCsrely on mechanical modes with large wavevectors to suppress leakage into the substrate and enable confinement without suspension, which increases the design complexity 19 ; 4 . Strong optomechanical interactions further require phase-matching between the optical and mechanical waves. In existing release-free OMCs, this constraint prohibits the use of an optical mode at the X-point 19 ; 4 . Instead, the release-free OMCcan be designed with an optical mode with wavevector near the half-X-point ( k=0.5\pi/a , where a is the periodicity) and a mechanical mode at the X-point (wavevector

Read the complete text