Atom-Light Hybrid Interferometer

A new type of hybrid atom-light interferometer is demonstrated with atomic Raman amplification processes replacing the beam splitting elements in a traditional interferometer. This nonconventional interferometer involves correlated optical and atomic waves in the two arms. The correlation between atoms and light developed with the Raman process makes this interferometer different from conventional interferometers with linear beam splitters. It is observed that the high-contrast interference fringes are sensitive to the optical phase via a path change as well as the atomic phase via a magnetic field change. This new atom-light correlated hybrid interferometer is a sensitive probe of the atomic internal state and should find wide applications in precision measurement and quantum control with atoms and photons.

Figure 1

Experimental sketch for the hybrid atom-light interferometer. A strong Raman write beam ($W_1$, red) and a Stokes input field ($S_0$, blue) in orthogonal polarization interact with a $\mathrm{\Lambda }$-shaped atomic system to generate an amplified Stokes field ($S_1$) and a correlated atomic spin wave $S_a$ that stays in the atomic system. The amplified Stokes beam ($S_1$), after reflection ($S_1^{\prime }$), is sent back together with another strong write beam ($W_2$) to the atomic cell after some optical delay to recombine with the waiting atomic spin wave $S_a$ for superposition. The final Stokes field ($S_2$) (generated by $W_{\mathbf{2}}$ and $S_1^{\prime }$) is detected by $D$, together with a delayed anti-Stokes field (AS) due to a strong delayed reading beam ($R$, red) for the atomic spin wave readout. Both $S_{\mathbf{2}}$ and AS show interference fringes. Top inset: an interferometer with nonlinear processes (NP) for wave splitting and recombination. The labels correspond to the main figure. Bottom inset: atomic energy levels.

Figure 2

Observed interference fringes in two output signals ($S_2$ and AS) as the optical phase is scanned. (a) The input field is the light field $S_0$. The Stokes light output has a slightly smaller visibility (94%) than the readout anti-Stokes (96.5%). (b) The input field is the atomic spin wave ($S_{a0}$) (see the Supplemental Material [31] for this arrangement). The Stokes light output has a slightly higher visibility (96.3%) than the anti-Stokes readout (93.6%). In both figures, the red circles are for the Stokes light output and the blue squares are for the anti-Stokes light output (2.9 times magnified for comparison with the Stokes), which represents the atomic spin wave.

Figure 3

Sagnac loop for the optical delay. (a) We use the Sagnac configuration for the write and Stokes delay to stabilize the optical phase of the interferometer: The write and Stokes fields travel in the opposite directions inside the Sagnac loop to cancel any optical phase change due to mirror vibration. Red: Raman write field; blue: Stokes field; $\lambda /2$: half wave plate to rotate the polarization angles by 90 degrees; FC: fiber coupler; SMF: single-mode fiber; Stage: linear translational stage. (b) Magnetic sublevels for optical and multiple atomic spin waves.

Figure 4

Interference fringes by scanning the atomic phase via magnetic field. Optical delay is (a) 60, (b) 100, and (c) 160 m, respectively. Red dot: the real interference fringe; black line: the fitting curve of the interference fringe; blue line: the ramp scan of the magnetic field. (d) The phase sensitivity with respect to the magnetic field change as a function of the fiber delay length.

Figure 5

Optical phase shift as a function of the stage displacement. (a) Interference fringes as the atomic phase is scanned via the magnetic field at two different locations of the translational stage. (b) Relative phase shifts derived from (a) as a function of the position of the translational stage. The solid red line is a linear fit. The fit slope is $16.9\pm 0.7\mathrm{degree}/\mathrm{mm}$ and in accordance with the theoretical value of $16.3\mathrm{degree}/\mathrm{mm}$ found via Eq. (4).