Doppler-free Rydberg Spectroscopy in Warm Vapor

Abstract

The common approach for producing Rydberg atoms in warm vapor cells is with lasers arranged in a counter-propagating, collinear configuration. Doppler effects in these configurations reduce the efficiency of excitation to the Rydberg state while also producing broadened spectral features. In this work, we demonstrate a three-laser Doppler-free excitation using laser beams whose k-vectors sum to zero, resulting in an enhancement in the Rydberg density and narrowed spectral features. A three-times enhancement to Rydberg density along with a near four-times reduction in spectroscopic line-widths are observed compared to a collinear configuration. This Doppler-free configuration could prove beneficial to Rydberg atomic technologies, such as electric field sensing with small volumes or deterministic photon sources.

Figures (7)

• Figure 1: (a) CL configuration with all lasers having parallel linear polarization. (b) DF star configuration where $\theta$ and $\phi$ are the angles of the Rydberg and dressing lasers with respect to the probe. (c) Excitation level diagram used for both the DF and CL measurements. (d) Rydberg decay diagram for fluorescence measurements. We collect the $\approx$480 nm light.
• Figure 2: Top: One-dimensional simulation of the absorption of the probe laser per velocity class in a non-DF (a) and DF (b) configuration versus Rydberg laser detuning in a three-laser excitation scheme. Bottom: Simulated EIT spectra in the non-DF (c) and DF (d) configuration. Laser Rabi frequencies are $\Omega_{\rm p} = 2\pi\times 2$ MHz, $\Omega_{\rm d} = 2\pi\times10$ MHz, and $\Omega_{\rm R} = 2\pi\times 1$ MHz.
• Figure 3: Amplitude and linewidth of EIT spectroscopy signal as the Rydberg-laser beam angle varies. Theory predicted values are solid closed symbols, and measured values are open symbols. The max signal amplitude and narrowest linewidth occur at the DF angle, denoted 0 mrad. For clarity of visual comparison of the profiles, a systematic offset in the experiment values has been compensated for by adding $13$ mrad. This should not be interpreted as absolute agreement of the optimum angular shift between theory and experiment.
• Figure 4: (a) Transmission spectra showing splitting and power broadening in the CL configuration as a function of Rydberg laser detuning and dressing laser Rabi frequency. The inset is a slice from the contour plot depicting experimental EIT spectra (black circles) at the lowest dressing Rabi frequency of $2\pi\times2$ MHz. The red solid line is a fit to the experimental data, and the blue dashed line is the simulated EIT spectra. (b) Simulated contour plot of power broadening and splitting in the CL configuration. For both (a) and (b) $\Omega_{p} = 2\pi\times 2.4$ MHz and $\Omega_{R} = 2\pi\times1.2$ MHz.
• Figure 5: (a) FWHM of the EIT linewidth in the DF configuration versus dressing laser Rabi frequency with Rydberg laser Rabi frequency held constant $\Omega_{R} = 1.2$ MHz and four probe laser Rabi frequencies represented by symbol color. (b) EIT spectra with the smallest linewidth from (a) showing FWHM of 1.18(8) MHz. Solid red curve is a Voigt fit. Comparable CL data is shown with blue squares along with a double Voigt fit (dashed red curve). For both CL and DF, the amplitudes have been scaled to a value of 1, and both use similar Rabi frequenices: DF probe Rabi frequency is $2\pi\times 2.2$ MHz and dressing Rabi frequency $2\pi \times21.4$ MHz, while the CL had $\Omega_{p} = 2\pi\times2.4$ MHz and $\Omega_{d} = 2\pi\times 20.5$ MHz