Single-photon excitation of $K\alpha$ in heliumlike ${\mathrm{Kr}}^{34+}$: Results supporting quantum electrodynamics predictions

We study two fundamental transitions from the ground state ${}^{1}S_0$ to ${}^{1}P_1$ ($w$ line) and ${}^{3}P_1$ ($y$ line) in heliumlike ${\mathrm{Kr}}^{34+}$ by resonant single-photon excitation using an electron-beam ion trap and monochromatic x rays at PETRA III. Our results for the transition energies $E\left(w\right)=13114.47\left(14\right)$ eV and $E\left(y\right)=13026.15\left(14\right)$ eV are in excellent agreement with quantum electrodynamics calculations, but disagree for $w$ with the average of the hitherto reported experimental results ${\overline{E}}_{\mathrm{Lit}}\left(w\right)=13115.15\left(17\right)$ eV. Independently of any energy calibration, we obtain a value $E\left(w\right)/E\left(y\right)=1.006780\left(7\right)$, also in consistency with predictions.

Figure 1

Experimental raw data [Si(333)] of the $w$ and $y$ line along with the Pb $L_3$-edge absorption spectrum (PbL) for energy calibration. The $w$ data points show a result from a single scan (No. 6 data point in red) lasting about 15 min. A Gaussian fit to this scan provides a reduced chi-squared (RCS) of $1.3$ and a center Bragg angle of ${\theta }_{\exp }=26.89274{\left(4\right)}^{\circ }$. A Voigt fit, as depicted, produces even more favorable RCSs. The $y$ line data represent an overlay of 25 single scans (background of 73 counts subtracted). The PbL (open circles) shows our data scaled to the (dashed) reference data [29, 30, 31]. Inset (blue): Centroids from 12 individual scans of $w$. Error bars include counting statistics, angular encoder errors (1 ppm), and errors due to normalization to the ion fill factor of the trap and beam overlap mismatch $(\sim 3\mathrm{ppm})$.

Figure 2

Comparison of our $w$ and $y$ line results with available theoretical and experimental values. Measurements No. 1 to No. 5 indicate results from Refs. [32, 33, 34, 35, 36], respectively. The black solid line is the weighted average of measurements No. 1 to No. 5. Theoretical data from Artemyev et al. [26], Cheng et al. [27], and Plante et al. [28] are included.