Collective inhibition of light scattering from atoms into an optical cavity at a magic frequency

TL;DR

This work demonstrates a new collective, interference-driven suppression of light scattering into a high‑finesse cavity by a cold ${}^{87}$Rb ensemble. The authors identify a magic frequency near $ ext{detuning } \Delta oughly -185 ext{ MHz}$ where both Rayleigh and Raman scattering are inhibited due to destructive interference among polariton excitations in the strongly coupled system, with the phenomenon captured by a polariton mean‑field model and characterized by an effective atom number $N_{ ext{eff}}$. They also corroborate a prior single‑atom magic frequency near $ ext{detuning } \Delta^* oughly -506 ext{ MHz}$ that suppresses Raman scattering alone. The results reveal a robust collective mechanism, largely insensitive to inhomogeneities, and suggest potential applications as a precise atomic reference and for cavity‑QED based quantum control.

Abstract

We report on the observation of a new magic frequency within the hyperfine structure of the D2 line of ${}^{87}$Rb atoms at which the scattering of light into a high-finesse cavity is suppressed by an interplay between quantum interference and the strong collective coupling of atoms to the cavity. Scattering from a cloud of laser-driven cold atoms into the cavity was measured in a polarization sensitive way. We have found that both the Rayleigh and Raman scattering processes into the near-resonant cavity modes are extinguished at 185 MHz below the F=2$\leftrightarrow$F'=3 transition frequency. This coincidence together with the shape of the observed spectral dip imply that the effect relies on a quantum interference in the polariton excitations of the strongly coupled combined atom-photon system. We have also demonstrated the existence of a magic frequency around -506 MHz, where only the Raman scattering is suppressed due to a quantum interference effect at the single-atom level.

Figures (4)

• Figure 1: Scheme of the experiment with $\sigma^+$-$\sigma^-$ laser drives on the D2 line of ${}^{87}$Rb atoms in the cavity and the polarization-resolved photon detection. The two-photon transitions induced by the coherent drives $\sigma^\pm$ are illustrated in the example for an initial state $m_F=0$. Cavity photons in three possible scattering channels are created: $\Delta m=0 \;$ (Rayleigh), and $\Delta m=\pm 1, \pm2 \;$ (Raman) transitions with respect to the quantization axis $\vec{z}$, the corresponding cavity mode polarizations are indicated, which are directed to separate detectors.
• Figure 2: Polarization-resolved photon count rates as a function of the drive detuning. The measurement data (solid lines with shaded area representing the width of the distribution at half maximum) are compared with the single-atom scattering rate model (dashed line). Vertical dashed lines indicate the transition resonances to excited hyperfine states. The arrow indicates the collective inhibition of scattering at the magic frequency.
• Figure 3: Photon scattering rate into the $\vec{y}$-polarized mode near the dip at $-185$ MHz. The average of measured data are plotted by the orange curve, the shaded area representing the standard deviation of 100 measurements. Dashed lines give the polariton theory curves for different effective atom numbers $N_{\text{eff}}$ indicated in the legend. These curves are plotted with a frequency offset to match the experimental dip position. The cavity detuning was $\Delta_c=0$.
• Figure 4: Polariton-mediated coherent photon scattering rate into the $\vec{y}$-polarized cavity mode as a function of the drive and cavity frequencies near the interference dip at –185 MHz. 2D color-coded plot presents the steady-state solution of \ref{['eq:phnumCoh']}, the two sections of thick red lines indicate the maxima as a function of the drive $\omega-\omega_3$. This was fit to the experimentally measured maxima of the cavity photon number for selected cavity frequencies, given by orange points. For some of them, the drive frequency dependence is shown (orange line with shaded area representing the standard deviation) in the right column of panels. Red dashed lines in these panels represent the theory curves, for which the maxima are not well defined between –181 MHz and –190 MHz.