Precision measurement of the ionization energy of Cs i

We present absolute-frequency measurements for the transitions from the $6s_{1/2}$ ground state of ${}^{133}\mathrm{Cs}$ to $n{p}_{1/2}$ and $n{p}_{3/2}$ Rydberg states. The transition frequencies are determined by one-photon UV spectroscopy in ultracold samples of Cs atoms using a narrow-band laser system referenced to a frequency comb. From a global fit of the ionization energy $E_{\mathrm{I}}$ and the quantum defects of the two series we determine an improved value of $E_{\mathrm{I}}/hc=31406.4677325\left(14\right){\mathrm{cm}}^{-1}$ for the ionization energy of Cs with a relative uncertainty of $5\times {10}^{-11}$. We also report improved values for the quantum defects of the $n{p}_{1/2}, n{p}_{3/2}, n{s}_{1/2}$, and $n{d}_{5/2}$ series.

Figure 1

Previous measurements of the first ionization energy of ${}^{133}\mathrm{Cs}$. Shown is the deviation of values reported in Refs. [23, 24, 25, 26] from the value determined in this work (last point on the right). Error bars indicate the $1\sigma$ uncertainties given in the respective references. Open symbols indicate that no uncertainty was given. For clarity the regions marked with different shades of gray have been scaled differently as indicated at the top of each region.

Figure 2

(a) Spectrum of the $27p_{1/2}\leftarrow 6s_{1/2}(F=4)$ transition with partially resolved hyperfine structure: (red lines) positions and relative weights of hyperfine components, (dashed line) fit of the line model of Eq. (1) to the raw data, (black points) binned raw data for visualization (error bars show the standard error of the mean). (b) Spectrum of the $42p_{1/2}\leftarrow 6s_{1/2}(F=4)$ transition with unresolved hyperfine structure. Legend as in (a).

Figure 3

The ac Stark shift of the $70p_{3/2}\leftarrow 6s_{1/2}$ transition of Cs measured in the ODT at different power levels of the ODT laser during excitation. Experimentally determined peak positions are displayed as full circles. The horizontal and vertical error bars indicate the estimated standard deviation of the peak position and a 3% uncertainty in the determination of the laser power, respectively. The full line represents the fitted linear model. The upper and lower insets show experimental spectra (black points) and a fit of Eq. (2) to the spectrum (solid lines) for ODT-laser powers of 1.1 and 10.0 W, respectively. All frequencies are given relative to the extrapolated zero-power peak position.

Figure 4

Fit residuals of the global fit based on Eqs. (3) and (5) to the observed frequencies of the $n{p}_{3/2}\leftarrow 6s_{1/2}$ (black triangles) and $n{p}_{1/2}\leftarrow 6s_{1/2}$ (red squares) transitions. The error bars give the estimated standard deviations of the center frequencies resulting from the fit of Eq. (1) to the raw data (see Sec. 2).