Efficient, high-resolution resonance laser ionization spectroscopy using weak transitions to long-lived excited states

Laser spectroscopic studies on minute samples of exotic radioactive nuclei require very efficient experimental techniques. In addition, high resolving powers are required to allow extraction of nuclear structure information. Here we demonstrate that by using weak atomic transitions, resonance laser ionization spectroscopy is achieved with the required high efficiency (1%–10%) and precision (linewidths of tens of MHz). We illustrate experimentally and through the use of simulations how the narrow experimental linewidths are achieved and how distorted resonance ionization spectroscopy line shapes can be avoided. The role of the delay of the ionization laser pulse with respect to the excitation laser pulse is crucial: the use of a delayed ionization step permits the best resolving powers and line shapes. A high efficiency is maintained if the intermediate level has a lifetime that is at least of the order of the excitation laser pulse width. A model that describes this process reproduces well the observed features and will help to optimize the conditions for future experiments. The simulation code is available upon request to the authors.

Figure 1

An atom with several hyperfine levels in the ground state and in the excited state multiplet. Note the figure is not to scale, since the hyperfine splitting is typically ${10}^6$ times smaller than the transition energy.

Figure 2

Comparison of Eqs. (19) and (20), representing the line broadening $D$ due to the excitation coupling $\mathrm{\Omega }$, for an Einstein $A$ coefficient of ${10}^6$/s for continuous-wave lasers and $T=50$ ns for pulsed lasers.

Figure 3

Two simulations for a two-level atom, using the parameters in Table 1. For each of the figures, the central surface plot shows the excited state population as a function of laser detuning of the first excitation step on the $x$ axis and time on the $y$ axis. The diagram to the left of these central plots schematically displays the laser pulse sequence. On the top of each figure the resonance ionization spectrum is shown. (a) Excited state population when using simultaneous laser pulses. The resonance ionization spectrum is power broadened by both laser pulses. (b) As in (a), but with a delayed ionization pulse. Delaying the ionization step removes the power broadening from the resonance ionization spectrum.

Figure 4

Comparison of the ionization spectra obtained in the top panels of Figs. 3 (red dot-dash line), 3 (green full line), and 9 (blue dashed line). Note how delaying the ionization step removes all power-broadening and line-shape distortions.

Figure 5

Ionization scheme used for ${}^{63}\mathrm{Cu}$ and ${}^{65}\mathrm{Cu}$.

Figure 6

Ionization spectrum of ${}^{63,65}\mathrm{Cu}$. The black box indicates the peaks presented in the zoom-in in Fig. 7.

Figure 7

Two lines in the spectrum of ${}^{63,65}\mathrm{Cu}$ (see Fig. 6 for the full spectrum). Plotted in red, open symbols is a spectrum obtained with temporally overlapping lasers (broad lines), while green full circles are used for a spectrum obtained with a delayed ionization step [40(10) ns delay]. Symbols are the experimental data points; the full lines are the fit. The laser power was 150 $\mathrm{mW}/{\mathrm{cm}}^2$ for both measurements. A sharp reduction in the linewidth can be clearly seen, without loss in efficiency.

Figure 8

Linewidth of the Lorentzian component of the Voigt profile as a function of the delay time of the second laser pulse. The red star corresponds to the red spectrum in Fig. 7; the green triangle corresponds to the green spectrum in Fig. 7. Errors are smaller than the symbols. Also shown is a theoretical calculation of the linewidth as function of the laser delay, using the experimental parameters described in the text.

Figure 9

As Fig. 3, but including a Stark shift due to the ionization laser. (a) Excited state population and ionization spectrum for simultaneous laser pulses. The ionization spectrum is asymmetric, which can also be seen in the population of the excited state. (b) Excited state population and ionization spectrum for delayed laser pulses. By delaying the ionization laser, both the power broadening and asymmetry in the ionization spectrum disappear.

Figure 10

Ionization scheme used to excite and ionize ${}^{221}\mathrm{Fr}$. The size of the hyperfine splittings was calculated using the values in [28].

Figure 11

Resonance ionization spectra of the lowest-frequency components of the hyperfine structure of ${}^{221}\mathrm{Fr}$, obtained with simultaneous laser pulses (red, open symbols) and with a delayed ionization step (green, full symbols). The solid lines are fits using the model for RIS presented in Sec. 2. The asymmetry of the line disappears when the ionization laser is delayed, while the total ionization efficiency does not decrease significantly.

Figure 12

Ionization efficiency for a chopped cw laser with a cw power of 10 mW, overlapped with a 1 mJ/pulse ionization laser and with a variable excitation laser pulse length. The system that was simulated here consists of a ground state doublet and a single excited state. For a weak transition ($A={10}^6$/s) the highest efficiency is reached if the pulse length is $\approx 1\mu \text{s}$. For extremely long laser pulses, the efficiency decreases in all cases due to optical pumping towards the other hyperfine level of the ground state.