Figure 1

Two amplitudes interfere in the excitation of the $m_j^{\prime }=0$ sublevel in a $J=0\to J^{\prime }=1$ two-photon atomic transition. Similar diagrams can be drawn for excitation to the $m_j^{\prime }=\pm 1$ sublevels.Reuse & Permissions

Figure 2

Plot of $|f_{\pm }{|}^2$ on resonance (${\Omega }_1+{\Omega }_2={\omega }_{eg}$) as a function of one photon’s frequency ${\Omega }_1$. The two peaks correspond to the cases where the frequency of one of the photons matches the ground-to-intermediate transition.Reuse & Permissions

Figure 3

Partial energy-level diagram of barium. Even (odd) parity states are in the left (right) column. Energies are in ${\mathrm{cm}}^{-1}$. The close proximity of $6s6p{}^{1}P_1$ to the energy midpoint ($\Delta =18060-35934/2=93{\mathrm{cm}}^{-1}$) enhances the two-photon transition rate.Reuse & Permissions

Figure 4

Experimental apparatus. An atomic beam of barium interacts with the light fields of two dye lasers ($\lambda \cong 557\mathrm{nm}$) at the center of a tunable confocal optical power buildup cavity (PBC) of length $L\cong 30\mathrm{cm}$. The lasers are coupled into the cavity to excite only symmetric transverse modes, making the effective free spectral range (FSR) $c/2L\cong 500\mathrm{MHz}$. Cavity finesse is 340. The lasers drive the $6s^2{}^{1}S_0\to 5d6d{}^{3}D_1$ two-photon transition of barium; see Fig. 3. Fluorescence from $5d6d{}^{3}D_1\to 6s6p{}^{3}P_{0,1,2}$ ($\lambda \cong 420\mathrm{nm}$) is collected through a light pipe and bandpass filters (not shown) and detected by a low-noise photomultiplier tube and single-photon counter. Each dye laser is locked to the PBC with an electro-optic modulator (EOM) and rf electronics, employing the Pound-Drever-Hall method [13]. Polarizing beam splitters and Faraday rotators isolate the lasers from each other and from their own cavity-rejected light. The lasers are also coupled into a single-mode fiber and interference is detected with a fast photodiode. A stabilized cavity, of $\mathrm{FSR}\cong 50\mathrm{MHz}$, records transmission peaks from the Scylla laser.Reuse & Permissions

Figure 5

Laser tuning path during run. The lasers are tuned in concert with the PBC, either separated by $m=4$ longitudinal cavity modes or in the same mode.Reuse & Permissions

Figure 6

Processed data from one run. Fluorescence (photons/second), from $5d6d{}^{3}D_1\to 6s6p{}^{3}P_{0,1,2}$. ${\Omega }_1^{\prime }\equiv {\Omega }_1-{\omega }_{eg}^{(138)}/2+{\delta }_{\mathrm{cal}}$, measured in reference-cavity FSRs ($\approx 50\mathrm{MHz}$). In the region L1 (Fig. 5), $-4\le {\Omega }_1^{\prime }<11$, ${\Omega }_2-{\Omega }_1=2\pi \times 2\mathrm{GHz}$, and a calibration spectrum is acquired containing both even and odd isotopes. Solid lines mark their approximate locations, known from a separate measurement to be presented elsewhere. Light shifts are responsible for the broad line shapes. In the region L2, $13<{\Omega }_1^{\prime }\le 25$, ${\Omega }_2-{\Omega }_1\approx 0$, and longer integration improved the signal-to-noise ratio. The spectrum of even isotopes (dashed lines) is suppressed below detection by BE statistics. In the spectrum of odd isotopes, however, the hyperfine interaction enables the transition. This is the first observation of this effect.Reuse & Permissions

Figure 7

$\nu$ from 12 runs. The error bars decrease over the lifetime of the experiment as improvements in excitation and detection efficiency were achieved.Reuse & Permissions