Strongly Correlated Growth of Rydberg Aggregates in a Vapor Cell

The observation of strongly interacting many-body phenomena in atomic gases typically requires ultracold samples. Here we show that the strong interaction potentials between Rydberg atoms enable the observation of many-body effects in an atomic vapor, even at room temperature. We excite Rydberg atoms in cesium vapor and observe in real time an out-of-equilibrium excitation dynamics that is consistent with an aggregation mechanism. The experimental observations show qualitative and quantitative agreement with a microscopic theoretical model. Numerical simulations reveal that the strongly correlated growth of the emerging aggregates is reminiscent of soft-matter type systems.

Figure 1

Principle of the aggregation. (a) Interaction induced level shift. An already excited Rydberg atom (red dot) at the origin produces an energy shift for the neighboring atoms. At the facilitation radius $r_{\text{fac}}$, the atoms are exactly shifted in resonance with the excitation laser, red-detuned by $\mathrm{\Delta }$. The grey shaded area symbolizes the excitation bandwidth, a Gaussian whose width is given by the dephasing rate $\gamma$. (b) Left: Typical time evolution of the Rydberg population and of the transmission change of the exciting laser [experimental signal, which is proportional to the derivative of the Rydberg density, see text and Fig. 2 for details] during the aggregation process. Right: Snapshots of the aggregation at times $A,B$, and $C$, illustrated for a frozen gas in 2D. Red dots correspond to Rydberg atoms. Blue dots correspond to resonant atoms. The dark grey line shows the “resonant shell” in which the excitation of Rydberg atoms is facilitated. Inset: Sketch of the density-density correlation function expected for the steady state [corresponding to $C$].

Figure 2

(a) Schematic energy level diagram for the excitation to Rydberg states of cesium atoms. (b) Sketch of the experimental setup. (c) Transmission signals at different atom number densities for the 32S state. The two-photon detuning is $\mathrm{\Delta }\approx 2\pi \times -2200\mathrm{MHz}$. The two-photon Rabi frequency is $\mathrm{\Omega }=2\pi \times 100\mathrm{MHz}$. The grey-shaded area in the background represents the temporal pulse shape of the infrared laser. The time delay $t_{\text{sat}}$ used for the quantitative analysis is also depicted. The black curve (upper panel) is the expected transmission signal (not to scale) for noninteracting atoms (following the approach from Ref. [33]).

Figure 3

Density plots showing the transmission change as a function of time (vertical axis, from top to bottom) and detuning. The parameters are as in Table 1. (a) Simulated data with pure van der Waals interaction. The Rabi frequency was rescaled with a factor 0.38. (b) Simulated data with pure dipole-dipole interaction (for $15^3$ atoms). The Rabi frequency was rescaled with a factor 0.23. (c) Experimental data. The blue (resp. green) line shows the position of the absorption maximum $t_{\text{sat}}$ in the simulated data with van der Waals (resp. dipole-dipole) interaction.