Realization of an all-optical effective negative-mass oscillator for coherent quantum noise cancellation

Abstract

We report the realization of an all-optical, tabletop effective-negative-mass oscillator (ENMO) scheme capable of canceling quantum noise when cascaded with an opto-mechanical sensor susceptible to (quantum) radiation pressure noise. Our coherent quantum noise cancellation (CQNC) scheme offers a broadband cancellation capability with a tunable, wavelength-flexible, and compact system. This is achieved through the implementation of an optical equivalent of an opto-mechanical interaction, facilitated by a down-conversion and a beam-splitting process. The intricate nature of the system and its multiple interacting components made characterizing the interdependent parameters with conventional methods ineffective, leading to the development of an in-situ characterization scheme. The obtained parameters meet the targets for CQNC set in previous studies. With our current realization, we project a broadband quantum noise reduction of 3.6 dB, corresponding to a 77% reduction in quantum back-action noise at the optimal frequency of maximum reduction, indicating the readiness of the ENMO for application. We discuss the prospects for new applications in quantum information and communication using the same platform.

Figures (6)

• Figure 1: a) The cascaded CQNC system with the ENMO in series to an OMS and its corresponding quantum noise ellipses at different frequencies are shown for each system. The tailored frequency-dependent inversely squeezed state (blue) destructively interferes with the ponderomotively squeezed state (orange), resulting in vacuum states (green). The scheme reveals the force signal (black arrow), which was masked by the additional phase quadrature noise due to ponderomotive squeezing. b) shows the schematic plots for the ideal susceptibilities of the oscillators.
• Figure 2: a) Representation of a typical opto-mechanical cavity. Here, the optical resonator (meter cavity) field shown in red interacts with the mechanical oscillator mirror shown in purple. The output field from the opto-mechanical cavity has QBA noise imprinted (ponderomotive squeezing), which is detected using a homodyne detection scheme. b) Here, the opto-mechanical oscillator is replaced by a high-finesse optical resonator (ancilla cavity) shown in purple. The power beam-splitter couples the two oscillators through the beam-splitting process. The down-conversion process is not present here. c) In this realization, the high-finesse ancilla cavity is an orthogonal polarization mode instead of a spatially separated cavity. The opto-mechanical interaction is then mimicked by an optical-optical interaction by using a type II down-conversion crystal (down-conversion process) and a waveplate (beam-splitter process), and careful selection of the oscillator configuration.
• Figure 3: Schematic of the optical realization of the ENMO. It consists of a polarization-coupled cavity system with a type II PPKTP crystal. The cavity is 1.52m long and has an FSR of 197.4MHz. The input/output coupler has a reflectivity of 97 for p-polarization and reflectivity of 99.995 for s-polarization. The cavity is pumped with a single-pass 532nm light field in s-polarization. The dashed lines represent the vacuum states in p-polarization that are squeezed by the setup. Homodyne detection with a local oscillator in p-polarization detects the inversely squeezed states in the p-polarized light.
• Figure 4: a) Measured variance (normalized to the shot noise, with dark noise subtracted) of the output field as a function of frequency for the ancilla cavity detuning $\Delta_\mathrm{a}$ of -465kHz (light blue), -710kHz (light red) and -1050kHz (light green). For each detuning, the variance is shown for two approximately orthogonal detection-locked quadratures. The corresponding fits are shown as dashed lines in matching dark colors. The squeezing measurements correspond to an ENMO for a Si$_3$N$_4$ membrane opto-mechanical system as compared in Table. \ref{['tab: exp output']} b) Tomographical reconstruction of the quantum squeezed ellipse over frequency for $\Delta_\mathrm{a}$ of -710kHz.
• Figure 5: a) Simulated variance of the noise in the phase quadrature due to ponderomotive squeezing (orange line) and the potential broadband cancellation with the current ENMO (green line), normalized to shot noise. These are shown for the OMS and ENMO parameters given in Tab. \ref{['tab: exp output']}. The green-shaded area shows the reduction in noise. The projected noise with a potential future ENMO with improvement of $\kappa_\mathrm{a} =$10kHz is shown in purple. The corresponding dashed lines show the canceled noise. b) shows the percentage of QBA noise that could be canceled with the current and future projected ENMO realization.