Highly charged ions (HCI) are highly sensitive probes for testing fundamental physics, since effects originating from special relativity, quantum electrodynamics and nuclear interactions are enhanced by orders of magnitude compared to neutral or singly-charged atomic systems. Furthermore, they are promising candidates for next-generation atomic clocks as they are also less sensitive to external electric field perturbations. A great variety of forbidden optical transitions exist, some of which would be extremely sensitive to any changes of the fine-structure constant or the electron-proton mass ratio .
For a long time, accurate and high-resolution optical spectroscopy of HCIs was limited to the ppm level due to Doppler broadenings of several 10 GHz, caused by the typical megakelvin plasma temperatures inherent to HCI sources. Transporting HCIs into Penning or Paul traps, recently accomplished in the cryogenic Paul trap experiment CryPTEx at the Max-Planck-Institut für Kernphysik (MPIK) , allows for advanced cooling techniques. But still sensitive detection techniques are required.
At the German national metrology institute (PTB), we have commissioned in collaboration with MPIK an evolved version of CryPTEx with improved magnetic shielding and vibration damping. Following the MPIK experiment [3,4], Ar13+ is produced by an electron beam ion trap , extracted, decelerated, transferred to a linear Paul trap and there retrapped in a Coulomb crystal of laser-cooled beryllium ions. Next, a two-ion crystal composed of a single Ar13+ and an auxiliary Be+ ion is prepared. After Doppler-cooling, the crystal is sideband-cooled to its motional ground state. Using the quantum logic technique , we then demonstrated for the first time coherent laser spectroscopy on HCIs. With a pre-stabilized clock laser, we could resolve the 2P1/2 – 2P3/2 M1 transition at 441 nm on a sub-kHz level and improve the previously best reported resolution  by about seven orders of magnitude. In a next step our clock laser will be transfer-stabilized to an ultra-stable laser of sub-10 mHz linewidth . This should allow us to resolve the natural linewidth of 17 Hz and evaluate minuscule systematic shifts to determine the unperturbed transition frequency with sub-Hz accuracy, by directly linking it to the SI second via a femtosecond frequency comb. The corresponding fractional frequency inaccuracy of better than 10-15 paves the way to an HCI-based optical atomic clock.
Following this proof-of-principle experiment, we can easily switch the HCI species to a more promising clock candidate, allowing for spectroscopy of forbidden transitions with proposed inaccuracies competitive with the most accurate atomic clocks in the world.
 M. G. Kozlov et al., arXiv:1803.06532 Phys. (2018).
 L. Schmöger et al., Science 347, 1233 (2015).
 M. Schwarz et al., Rev. Sci. Instrum. 83, 083115 (2012).
 L. Schmöger et al., Rev. Sci. Instrum. 86, 103111 (2015).
 P. Micke et al., Rev. Sci. Instrum. 89, 063109 (2018).
 P. O. Schmidt et al., Science 309, 749 (2005).
 V. Mäckel et al., Phys. Scr. T156 014004 (2013).
 D. G. Matei, Phys. Rev. Lett. 118, 263202 (2017).