Intracavity Rydberg Superatom for Optical Quantum Engineering: Coherent Control, Single-Shot Detection, and Optical $\pi$ Phase Shift

We demonstrate a new versatile building block for optical quantum technologies, enabling deterministic quantum engineering of light by combining the advantages of two complementary approaches: cavity quantum electrodynamics and interacting atomic ensembles. Our system is based on an intracavity Rydberg-blockaded atomic ensemble acting as a single two-level superatom. We coherently control its state and optically detect it in a single shot with 95% efficiency. Crucially, we demonstrate a superatom-state-dependent $\pi$ phase rotation on the light reflected from the cavity. Together with the state manipulation and detection, it is a key ingredient for implementing deterministic photonic entangling gates and for generating highly nonclassical light states.

Popular Summary

As a building block for quantum technologies, photons have many advantages but also one major drawback: They do not interact with each other, which makes them difficult to use in quantum engineering tasks requiring coherent interactions between different components. To make photons interact, a well-explored path is to strongly couple them to individual atoms. A newer approach, which relies on atomic ensembles instead of individual atoms, is to inject photons into an atomic gas and create interactions between the excited atoms, thus making the atom-light coupling dependent on the number of excitations. The two approaches have fundamental and technical limitations. Our experiments show that combining them relieves many of these constraints.

Our new setup features a small atomic cloud within an optical resonator, both operating far from their physical and technical limits. Interactions between excited atoms make the cloud act as a single two-level “superatom,” strongly coupled with light. We demonstrate that the superatom’s state can be coherently manipulated and efficiently detected. Crucially, we observe that changing this state rotates the phase of light fields reflected off the cavity by 180°. This state-dependent phase rotation, together with atomic state control and detection, forms a complete set of tools for deterministic optical quantum engineering.

By sending polarized photons or dim light pulses into the cavity, changing the superatom’s state in between them and measuring this state at the end, one could deterministically prepare highly nonclassical light states for optical quantum sensing or realize two-photon quantum logic operations.

Figure 1

(a) Experimental setup, configured for transmission measurement. (b) Level scheme. A small cloud of rubidium atoms behaves as a two-level superatom thanks to the Rydberg blockade. The $D2$ drive and 109S drive beams allow the excitation of the superatom to the $|R⟩$ state. The atomic ensemble is strongly coupled to a medium-finesse cavity, resonant with the $D1$ line. The control beam enables the transmission of the $D1$ probe through the cavity by Rydberg EIT with the 78S state. (c) Cavity EIT spectra, where the transmission of the $D1$ probe, as a function of the probe detuning, is measured with a single-photon counting module (SPCM) and normalized to the maximum transmission of the empty cavity (light gray). With the strongly coupled atomic ensemble, we observe the vacuum Rabi splitting of the two transmission peaks, located at $\pm g/(2\pi )=\pm 10\mathrm{MHz}$ (red). Adding the EIT control beam opens a transmission window at resonance (blue).

Figure 2

Detection in transmission. The photon flux, detected with a SPCM, switches from (a) a high one without 109S Rydberg excitation to (b) quasizero with a single 109S excitation blockading the EIT (blockade sphere depicted in yellow). The insets show examples of single-shot measures with $2\text{−}\mu \mathrm{s}$ binning. (c) The 109S Rabi oscillations visible on the averaged SPCM counts in the detection windows (black), and the 109S population deduced from single-shot detections (cyan). Each data point represents 200 detections. The dashed line is a best-fit Gaussian decoherence model. (d) SPCM count rate since the beginning of the probe $D1$ pulse without (blue) and with (red) a 109S Rydberg drive $\pi$ pulse. The dark-red line is a best-fit exponential decay with a lifetime ${\tau }_R=42\mu \mathrm{s}$. The dashed vertical line marks the end of the optimized detection window $t_i$. (e) Histograms of the number of photons, without (blue) and with (red) the $\pi$ pulse. The dark curves represent modeled histograms. The dashed vertical line shows the optimized detection threshold $n_t$. (d,e) Data averaged over 400 repetitions. All error bars represent the standard error.

Figure 3

The $\pi$ phase shift in reflection. The reflected probe is combined with a local oscillator beam on a beam splitter, whose output intensity difference is measured by a HD setup. With the addition of a 109S Rydberg excitation blockading the EIT, the reflected probe phase changes from $\pi$ (a) to 0 (b), resulting in a sign flip of the measured HD signal. The insets show examples of single-shot measures with a $2\text{−}\mu \mathrm{s}$ binning. (c) In-phase quadrature $X$ since the beginning of the probe $D1$ pulse without (blue) and with (red) a 109S Rydberg drive $\pi$ pulse. The dashed lines indicate the quadrature means, and the colors represent the measured probability densities. The dashed vertical line gives the upper limit of the optimized detection window. (d) Histograms of the field $X$ quadrature values, without (blue) and with (red) the $\pi$ pulse. The dark curves represent modeled histograms. The dashed vertical line shows the optimized detection threshold. In panels (c) and (d), data are averaged over 400 repetitions. All error bars represent the standard error.

Figure 4

Probe-dependent lifetime. (a) SPCM count rates since the beginning of the probe $D1$ transmitted detection pulse for several probe intensities, after a $\pi$ pulse preparing the $|R⟩$ superatom state. The dashed lines are the best-fit exponential decay with a lifetime ${\tau }_R$. (b) $|R⟩$ decay rate $1/{\tau }_R$ as a function of the transmitted probe SPCM rate ${\varphi }_G$ with the $|G⟩$ state. Colored data points correspond to the curves of panel (a) with identical colors and markers. The dashed line shows the best linear fit to the data. Error bars represent the standard error.

Figure 5

Reflection spectra. The reflection of the $D1$ probe, as a function of the probe detuning, is measured with a SPCM and normalized to the impinging power (light gray). With the strongly coupled atomic ensemble, we observe the vacuum Rabi splitting of the two reflection dips (red). Adding the EIT control beam, with Rabi frequency ${\mathrm{\Omega }}_c/(2\pi )=13\mathrm{MHz}$, creates a reflection dip at resonance (blue).