Squeezed Light from a Levitated Nanoparticle at Room Temperature

Quantum measurements of mechanical systems can generate optical squeezing via ponderomotive forces. Its observation requires high environmental isolation and efficient detection, typically achieved by using cryogenic cooling and optical cavities. Here, we realize these conditions by measuring the position of an optically levitated nanoparticle at room temperature and without the overhead of an optical cavity. We use a fast heterodyne detection to reconstruct simultaneously orthogonal optical quadratures, and observe a noise reduction of $9\%\pm 0.5\%$ below shot noise. Our experiment offers a novel, cavityless platform for squeezed-light enhanced sensing. At the same time it delineates a clear and simple strategy toward observation of stationary optomechanical entanglement.

Figure 1

Experimental setup. A silica particle is trapped in an optical tweezer (optical frequency: ${\omega }_L$) in ultrahigh vacuum (UHV). The backscattered light is isolated by a Faraday rotator (FR) in combination with a polarizing beam splitter (PBS), and focused onto a single mode fiber (SMF) for confocal detection. A $10:90$ beam splitter distributes the collected light between the homodyne and heterodyne detectors. The homodyne measurement is used for electrostatic feedback (fb). The heterodyne signal is numerically demodulated and low-passed (LP) to obtain orthogonal quadrature components.

Figure 2

Ponderomotive squeezing. Spectrum of the optical quadrature at an angle $\theta =\pi /20$ (blue) and the optical shot noise of the heterodyne measurement (red), which is independently measured by blocking the signal path. Close to resonance we observe a noise reduction of about 9% below shot noise. The black dashed line shows the theoretical plot calculated for our experimental parameters.

Figure 3

Ponderomotive spectrogram. Spectrogram of the measured output quadratures as a function of the analyzer angle, where the warm (logarithmic scale) and cold (linear scale) colors represent noise values above and below shot noise, respectively. (a) The signal at $\mathrm{\Omega }\sim 0$ arises from the low frequency phase noise (minimized at $\theta =0$). For $\mathrm{\Omega }/2\pi =76.9\mathrm{kHz}$, the negative contribution of the squeezed light outweighs the other noise contributions in this frequency range, taking the spectrum below the shot noise level, in accordance with Eq. (2). No squeezing can be observed for the transverse modes at ${\mathrm{\Omega }}_x$ and ${\mathrm{\Omega }}_y$. (b) Detail of the measured spectrogram around the main resonance ${\mathrm{\Omega }}_z$. The inset shows the theoretical plot expected for our experimental conditions.