First-hand information for everyone
How it works (method and key physics) The membrane’s transverse motion scatters cavity illumination into optical sidebands with large tangential wavevectors. These are evanescent components that decay away from the surface and can be phase‑matched to electrons traveling in vacuum. Phase matching means the electron samples an optical phase that stays nearly stationary along its path so contributions add coherently.
Model The membrane is modeled as a square, fixed‑edge, tension‑dominated membrane. The analysis focuses on transverse (m,1) modes with spatial profile ψ_m(x,y)=sin(mπx/L) sin(πy/L) and modal wave number K_m = mπ/L.
Mechanically generated evanescent sidebands The mechanically generated upper optical sideband at frequency ω_L+Ω_m carries tangential components ±K_m. It is evanescent when K_m > (ω_L+Ω_m)/c, with decay rate κ_m = sqrt(K_m^2 − (ω_L+Ω_m)^2/c^2) and spatial factor e^{−κ_m(L_c−z)}.
Electron–field coupling The electron–field coupling amplitude is obtained by integrating the mechanically generated upper‑sideband vector potential along the electron trajectory. For a Gaussian interaction envelope the coupling can be written
g_m ≃ g_ref (q_M/q_ref) e^{−κ_m d} exp[−v^2 Δ_m^2 τ_int^2 /4].
Here q_M is the mechanical amplitude, d the electron–surface separation, Δ_m = K_m cosθ − (ω_L+Ω_m)/v is the phase mismatch, and τ_int is the effective interaction time. The three multiplicative factors describe mechanical‑amplitude control, evanescent attenuation, and finite‑time phase‑matching selectivity, respectively.
Concrete electron response When a quasi‑monoenergetic electron is phase‑matched, resolved recoil‑order populations follow P_ℓ = J_ℓ^2(g_m), where J_ℓ is a Bessel function. The root‑mean‑square recoil‑order width scales as √⟨ℓ^2⟩ = g_m/√2. Thus the mechanical amplitude q_M controls the strength of the electron–light interaction and redistributes electron populations among recoil orders. External forces that change q_M are therefore imprinted on the electron recoil spectrum.
Mode preparation Two optical strategies are proposed to excite high‑order transverse mechanical modes: (i) a focused intensity‑modulated beam localized near a mechanical antinode to maximize spatial overlap; (ii) two coherent non‑collinear beams that produce a traveling optical beat with wavevector K_D matched to K_m, driving the standing mechanical mode across many periods.
Sensing performance under the modeled conditions Sensing comparisons treat modes m=49 and m=50 separately as single‑mode benchmarks with the electron energy set for phase matching. Calculations use equal detected probe rates R_e = R_ph = 10^9 s^−1 and a mechanical quality factor Q_M = 200.
Under these reference conditions and at an integration bandwidth f_B = 5 GHz, the modeled integrated signal‑to‑noise ratio (SNR) for electron readout exceeds a Gaussian optical homodyne benchmark by ≈16 dB for m=49 and ≈34 dB for near‑dark m=50. The larger advantage for m=50 reflects its strongly suppressed optical overlap while phase‑matched electron coupling remains appreciable. The absolute SNR remains below unity with the prescribed signal calibration.
Force‑referred imprecision At their resonances and for the reference parameters, the force‑referred readout imprecision (resonant minima) is approximately 0.57 and 0.64 nN Hz^{−1/2} for electron detection of modes m=49 and m=50, respectively. The corresponding Gaussian homodyne imprecisions are about 4.0 and 34 nN Hz^{−1/2}. These values quantify relative sensitivity for the chosen, calibrated scenario.
Caveats and limitations The absolute moving‑boundary conversion coefficient η_{m,+} and related device‑dependent factors are not derived from first principles in this work; they are absorbed into an empirical calibrated reference coupling g_ref. Absolute conversion efficiencies therefore depend on material and device details.
The reference operating conditions assume a high‑order GHz mechanical mode. For L = 20 μm the modeled (50,1) mode lies near 3.27 GHz; realizing such modes may require engineered membranes or phononic structures. Cavity filtering can strongly attenuate GHz interrogation, so preparation and device engineering are nontrivial.
The analysis assumes large coherent mechanical displacement (q_M ≫ zero‑point and thermal fluctuations) and negligible electron‑induced backaction during each passage. Extensions to a full quantum treatment that include decoherence, loss, and backaction are beyond the present calculations and would be required to assess any quantum sensing advantage.
Summary This theoretical analysis predicts that transverse vibrations of an optomechanical membrane can convert mechanical motion into evanescent optical sidebands that phase‑match to free electrons, enabling mechanical control of electron recoil and the encoding of localized forces in the electron recoil spectrum, subject to the device‑dependent conversion, preparation challenges, and the large‑amplitude assumptions stated above.