Superfluid Thin Film Optomechanics

• Superfluid thin film optomechanics is the study of nanometer-scale helium films that self-assemble on microphotonic resonators, enabling coupling between confined optical modes and third sound excitations.
• Experimental platforms such as silica microtoroids and silicon nanobeams exploit precise film thickness control and dispersive optical coupling to probe low-dissipation quantum fluids.
• Advanced techniques, including photothermal backaction and phononic crystal confinement, allow for tunable nonlinearity and potential access to quantum regimes in optomechanical systems.

Searching arXiv for recent and foundational papers on superfluid thin-film optomechanics. arxiv_search({"query":"superfluid thin film optomechanics helium third sound photothermal optomechanics", "max_results": 10, "sort_by": "relevance"}) arXiv search results:

1. (Korsch et al., 2024) — "Phononic Crystals in Superfluid Thin-Film Helium"
2. (Baker et al., 2016) — "Thin film superfluid optomechanics"
3. (Harris et al., 2015) — "Laser cooling and control of excitations in superfluid helium"
4. (Sfendla et al., 2020) — "Extreme quantum nonlinearity in superfluid thin-film surface waves"
5. (Korsch et al., 27 Aug 2025) — "A cryogenic chamber setup for superfluid helium experiments with optical fiber and electrical access"
6. (McAuslan et al., 2015) — "Microphotonic Forces From Superfluid Flow"
7. (Kashkanova et al., 2016) — "Photothermal optomechanics in superfluid helium coupled to a fiber-based cavity"
8. (Spence et al., 2020) — "Superfluid Optomechanics with Phononic Nanostructures"
9. (Barroso et al., 12 Sep 2025) — "Digital holographic imaging of superfluid helium free surfaces"
10. (Childress et al., 2017) — "Cavity Optomechanics in a Levitated Helium Drop" Superfluid thin film optomechanics studies the interaction between confined optical modes and collective excitations of nanometer-scale superfluid $^{4}$He films, usually self-assembled on microphotonic resonators. In the thin-film limit, the relevant mechanical excitations are typically third sound modes—oscillations of film thickness with a restoring force set by the van der Waals interaction with the substrate—while the optical mode is supplied by a whispering-gallery or nanobeam cavity whose evanescent field overlaps the film. This combination has been used to probe thermodynamics in real time, to cool and amplify superfluid excitations, and to engineer strongly confined phononic modes with large intrinsic nonlinearity (Harris et al., 2015, Baker et al., 2016, Korsch et al., 2024).

1. Emergence of the field

A central motivation for superfluid optomechanics is that superfluid $^{4}$He offers low mechanical dissipation, low optical absorption, vanishing viscosity, and high thermal conductivity, while remaining compatible with cryogenic electromagnetic cavities (Kashkanova et al., 2016, DeLorenzo et al., 2013). In thin films, these advantages are combined with self-assembly on microstructured substrates and direct overlap between the film and the optical near field. A few-nanometer film can therefore serve as the mechanical degree of freedom of an optomechanical resonator rather than merely as a surrounding fluid (Harris et al., 2015, Baker et al., 2016).

The early thin-film program developed along two complementary lines. One line was experimental: superfluid He-4 films of about $10$ nm thickness were assembled on silica microtoroids with major/minor diameters $70/7\,\mu\mathrm{m}$, and third sound modes with frequencies $10$ kHz–$5$ MHz were measured, cooled, and amplified optomechanically (Harris et al., 2015). The other line was analytical and numerical: thin self-assembled films covering microfabricated high-$Q$ whispering gallery mode resonators were analyzed as optomechanical systems, with predicted single-photon coupling rates $g_0>2\pi \times 100$ kHz and single-photon cooperativities $C_0>10$ under optimized conditions (Baker et al., 2016).

This thin-film development sits within a broader superfluid-optomechanics landscape. Bulk or cavity-filled helium systems established that superfluid acoustic modes can be read out with high precision and can exhibit exceptionally low dissipation (DeLorenzo et al., 2013, Kashkanova et al., 2016). That broader context is important because later thin-film work repeatedly showed that some intuition imported from solid-state optomechanics—especially about dominant force mechanisms and loss channels—must be modified for quantum fluids.

2. Hydrodynamic basis and optical coupling

The characteristic mechanical mode of a superfluid helium film is third sound, a surface wave in which the film thickness is modulated while the restoring force is provided by the van der Waals interaction between helium and substrate (Harris et al., 2015, Korsch et al., 2024). In the analytical thin-film treatment, the third sound speed is written as

$$c_3 = \sqrt{ 3\frac{\rho_s}{\rho} \frac{\alpha_\text{vdW}}{d^3} },$$

with $^{4}$0 the superfluid fraction, $^{4}$1 the van der Waals coefficient, and $^{4}$2 the film thickness (Baker et al., 2016). In the nanobeam phononic-crystal formulation, the restoring force per unit area is

$^{4}$3

and the hydrodynamics are described by a linearized Euler equation in the thickness modulation $^{4}$4 (Korsch et al., 2024).

Confinement quantizes these waves. For disks, the mode shapes are expressed in terms of Bessel functions,

$^{4}$5

with $^{4}$6 fixed by the boundary conditions (Baker et al., 2016). The resulting mechanical behavior differs from that of solid resonators because the fluid motion is not localized to a rigid body displacement; instead, the useful optomechanical coordinate is a surface-thickness field.

Optomechanical coupling is dispersive. Variations in film thickness modulate the local refractive index and therefore the optical resonance frequency. In the nanobeam realization this is summarized by

$^{4}$7

and mechanical motion is transduced by homodyne optical detection through the phase shift imprinted on the light field (Korsch et al., 2024). In the general thin-film model, the single-photon coupling is $^{4}$8, where $^{4}$9 is the cavity frequency shift per displacement and

$10$0

The single-photon cooperativity is

$10$1

These expressions make explicit that strong coupling depends jointly on field overlap and on the unconventional effective mass of third sound modes (Baker et al., 2016).

3. Device platforms, film formation, and metrology

Two experimental platforms have been especially prominent. The first is the silica microtoroid, on which superfluid He-4 films condense via van der Waals attraction and support third sound modes coupled to a high-$10$2 optical whispering-gallery mode (Harris et al., 2015). The second is the silicon nanobeam optical resonator, fabricated from SOI wafers with $10$3 Si, patterned and undercut to produce a suspended nanobeam, then passivated with atomic layer deposition alumina; a few nanometer thick superfluid film self-assembles on top and couples to a co-localized optical mode (Korsch et al., 2024).

A recurring technical theme is in situ control of film thickness. In the cryogenic chamber reported for millikelvin operation, an automated gas handling system with pneumatic valves and a piezo gauge controls the introduction of ultrapure $10$4 into a hermetically sealed chamber with optical fiber and electrical access (Korsch et al., 27 Aug 2025). Using suspended silicon nanobeam photonic crystal resonators, the optical cavity red-shift is calibrated to the helium film thickness: for thin films, $10$5, the resonance shift is $10$6 per nm of helium, and the control system allows steps as small as $10$7 nm in film thickness (Korsch et al., 27 Aug 2025). The upper practical thickness is set by the gravitational–van der Waals balance,

$10$8

beyond which excess helium forms a bulk puddle at the chamber bottom (Korsch et al., 27 Aug 2025).

Measurement has evolved from local cavity transduction to broader spatial metrology. Thin-film third sound has been read out with shot-noise limited homodyne detection, allowing thermal motion to be resolved on timescales $10$9, more than $70/7\,\mu\mathrm{m}$0 faster than the mechanical decay time $70/7\,\mu\mathrm{m}$1 ms (Harris et al., 2015). In a different but related direction, off-axis digital holography has been used for full-field imaging of superfluid $70/7\,\mu\mathrm{m}$2He free surfaces, including thick films, enabling reconstruction of gravity-capillary mode structure and extraction of film thickness from the measured dispersion (Barroso et al., 12 Sep 2025). This suggests a route toward combining local optomechanical readout with spatially resolved surface metrology.

4. Backaction, photothermal forces, and mode control

A common simplification is that radiation pressure should dominate in cavity optomechanics. Superfluid helium experiments have repeatedly shown that this need not be the case. In thin-film microtoroids, optical backaction involves both radiation pressure and photothermal forces, but the photothermal response—heating due to optical absorption followed by superfluid fountain flow—was observed to dominate over pure radiation pressure (Harris et al., 2015). Its sign can even differ for nearly degenerate modes depending on overlap with defects or spatial position, so the same optical detuning can cool one mode while heating another (Harris et al., 2015).

The dynamical consequence is the familiar detuning-dependent modification of linewidth. For the thin-film modes under dynamical control,

$70/7\,\mu\mathrm{m}$3

and cooling by a factor of $70/7\,\mu\mathrm{m}$4, corresponding to $70/7\,\mu\mathrm{m}$5, was demonstrated for certain modes (Harris et al., 2015). The photothermal response was also unusually fast, with $70/7\,\mu\mathrm{m}$6 ns, attributed to the very high thermal conductivity of superfluid helium (Harris et al., 2015).

The nanobeam phononic-crystal experiment reached the same qualitative conclusion. Photothermal backaction was dominant over radiation pressure, and linewidth broadening or narrowing depended on laser detuning and mode symmetry, with coupling constants $70/7\,\mu\mathrm{m}$7 of $70/7\,\mu\mathrm{m}$8 and $70/7\,\mu\mathrm{m}$9 for the fundamental and higher-order localized modes, respectively; fits yielded a photothermal time constant $10$0 (Korsch et al., 2024).

Related helium optomechanics in a fiber-based cavity provides an important control case. There, for a radial mode of superfluid helium surrounding the cavity, the optomechanical response was found to be predominantly photothermal rather than electrostrictive, with $10$1 and photothermal bandwidths of order $10$2 MHz (Kashkanova et al., 2016). Although that geometry is not a thin film, it reinforces that photothermal coupling is a general and experimentally consequential feature of helium optomechanics.

Superfluid-mediated optical forcing can also be realized outside the standard cavity-backaction picture. In a microtoroid coated by a thin $10$3 helium film, optical heating drove superfluid flow and evaporation, producing a force of $10$4 nN, roughly one order of magnitude larger than the radiation pressure force, and enabling feedback cooling of a mechanical mode to $10$5 mK (McAuslan et al., 2015). This demonstrates that superfluid transport itself can be an actuator.

5. Confinement engineering, dissipation, and phononic crystals

Strong confinement has become a defining theme of the field because it simultaneously raises mode frequency, reshapes dissipation channels, and enhances intrinsic nonlinearity. In the silicon nanobeam architecture, periodic patterning of the silicon material creates a periodic modulation of the superfluid film and therefore a phononic band structure for third sound (Korsch et al., 2024). For the design shown, symmetric third sound modes exhibit a band gap between $10$6 and $10$7 MHz, and a defect in the center of the beam produces a localized mode inside that gap (Korsch et al., 2024). The defect mode is confined to a region on the scale of the third sound wavelength, $10$8, nearly two orders of magnitude smaller than in previous whispering-gallery resonator implementations (Korsch et al., 2024).

Experimentally, with no helium no mechanical resonance was detected, while deposition of a $10$9 nm film produced discrete peaks between $5$0 and $5$1 MHz corresponding to localized third sound modes (Korsch et al., 2024). The fundamental defect mode reached $5$2 at $5$3 nm thickness, and $5$4 fell with increasing thickness due to enhanced radiation and internal loss (Korsch et al., 2024). The same experiment also detected extended low-frequency third sound modes below $5$5 MHz and showed that in the bulk superfluid regime the Si nanobeam’s $5$6 GHz acoustic breathing mode was strongly damped, with $5$7 dropping from $5$8 to $5$9 ns (Korsch et al., 2024). That observation directly contradicts the notion that vanishing viscosity alone guarantees negligible acoustic radiation.

Loss engineering has also been treated at the level of architecture. In the nanofluidic proposal based on a 2D sonic crystal, superfluid $Q$0He is confined in channels of height $Q$1; removing a pillar creates a point defect whose acoustic mode lies in a complete phononic bandgap (Spence et al., 2020). Numerical simulations indicate that planar radiation loss decreases exponentially with sonic-crystal size and that for $Q$2 crystals one can obtain $Q$3 (Spence et al., 2020). The total quality factor is decomposed as

$Q$4

emphasizing that superfluid devices are often limited by extrinsic radiation and substrate channels rather than by the intrinsic dissipation of helium itself (Spence et al., 2020).

6. Scaling laws, nonlinearity, and quantum regimes

Thin-film superfluid optomechanics is unusual even at the level of scaling. In the analytical thin-film model, the effective mass of a third sound mode scales as $Q$5, while the mechanical frequency scales as $Q$6 and the zero-point motion scales as $Q$7 (Baker et al., 2016). Consequently, thicker and heavier films can exhibit smaller effective mass and larger zero-point motion, which is opposite to the standard intuition developed for solid resonators (Baker et al., 2016). The same analysis predicts access to the regime $Q$8 for microdisk resonators with $Q$9 and sufficiently thick films (Baker et al., 2016).

The nonlinearity of third sound is likewise unconventional because it originates directly from the substrate van der Waals interaction rather than primarily from geometric elasticity. For confined disk modes, the potential energy can be written as

$g_0>2\pi \times 100$0

with cubic and quartic coefficients derived from the thickness dependence of the van der Waals potential (Sfendla et al., 2020). For a $g_0>2\pi \times 100$1 thick film confined to $g_0>2\pi \times 100$2, the predicted single-phonon shift is $g_0>2\pi \times 100$3, and the paper argues that this exceeds state-of-the-art values in contemporary solid-state nonlinear resonators by three orders of magnitude (Sfendla et al., 2020). The condition for phonon blockade is $g_0>2\pi \times 100$4, in which case the oscillator acquires two-level-system-like behavior (Sfendla et al., 2020).

Current phononic-crystal experiments have not yet reached that extreme regime, but they move in the same direction. For the localized third sound cavity modes realized in the nanobeam platform, theoretical estimates give single-phonon nonlinearities $g_0>2\pi \times 100$5 for the fundamental mode and up to $g_0>2\pi \times 100$6 for tightly confined higher-order modes; at the reported parameters $g_0>2\pi \times 100$7 is still insufficient for the quantum nonlinear regime, but only a moderate increase in $g_0>2\pi \times 100$8 and stronger confinement are identified as necessary (Korsch et al., 2024).

Thickness tunability is now being used to drive the system into qualitatively different dynamical states. In the millikelvin cryogenic chamber platform, optomechanically induced phonon lasing of phononic crystal cavity third sound modes was demonstrated, and the lasing threshold depended crucially on film thickness: at $g_0>2\pi \times 100$9, $C_0>10$0, while increasing the thickness to $C_0>10$1 nm reduced the threshold to $C_0>10$2 (Korsch et al., 27 Aug 2025). The conceptual threshold condition is

$C_0>10$3

This thickness dependence follows directly from the $C_0>10$4 scaling of the van der Waals restoring force (Korsch et al., 27 Aug 2025).

Taken together, these results define superfluid thin film optomechanics as a branch of cavity optomechanics in which geometry, hydrodynamics, and substrate interactions are inseparable. The field now includes real-time thermomechanical readout, photothermal cooling and amplification, phononic-bandgap confinement, sub-nanometer thickness control, and experimentally relevant pathways toward single-phonon nonlinear behavior (Harris et al., 2015, Korsch et al., 2024, Korsch et al., 27 Aug 2025). A plausible implication is that future progress will depend less on discovering new coupling mechanisms than on combining already demonstrated ingredients—strong confinement, low extrinsic loss, and precise film-thickness tuning—within a single platform.