Reversing hydride-ion formation in quantum-information experiments with ${\mathrm{Be}}^{+}$

We demonstrate photodissociation of ${\mathrm{BeH}}^{+}$ ions within a Coulomb crystal of thousands of ${}^9{\mathrm{Be}}^{+}$ ions confined in a Penning trap. Because ${\mathrm{BeH}}^{+}$ ions are created via exothermic reactions between trapped, laser-cooled ${\mathrm{Be}}^{+}\left({}^2{P}_{3/2}\right)$ and background ${\mathrm{H}}_2$ within the vacuum chamber, they represent a major contaminant species responsible for infidelities in large-scale trapped-ion quantum-information experiments. The rotational-state-insensitive dissociation scheme described here makes use of 157-nm photons to produce ${\mathrm{Be}}^{+}$ and H as products, thereby restoring ${\mathrm{Be}}^{+}$ ions without the need for reloading. This technique facilitates longer experiment runtimes at a given background ${\mathrm{H}}_2$ pressure and may be adapted for removal of ${\mathrm{MgH}}^{+}$ and ${\mathrm{AlH}}^{+}$ impurities.

Figure 1

(a) Illustration of the experiment trap electrode assembly in cross section with routing of laser beams for Doppler cooling of ${\mathrm{Be}}^{+}$ and photodissociation of ${\mathrm{BeH}}^{+}$. A spheroidal ion cloud is depicted at the center of the trap. The vertical axis of the electrode stack is aligned to the external, uniform magnetic field of 4.46 T. Also shown are side view images of the trapped Coulomb crystal of $\sim 1.2\times {10}^4$ ions (b) before and (c) after 9000 pulses of 157-nm photodissociation light at 1.6 mJ/pulse. We measure an increase in the ${\mathrm{Be}}^{+}$ fraction from $81\%$ to $97\%$ of the constant total ion number in the crystal. Each image is $1.2\times 1.2$ ${\mathrm{mm}}^2$.

Figure 2

(a) Low-lying electronic states of the ${\mathrm{BeH}}^{+}$ molecular ion as calculated in Refs. [25, 26]. We excite the $B{}^1\Pi \leftarrow X{}^1{\Sigma }^{+}(v^{\prime \prime }=0)$ transition above the dissociation threshold to produce ${\mathrm{Be}}^{+}$ and H as given by the $B{}^1\Pi$ asymptote. Relevant wave functions calculated as numerical solutions to the radial Schrödinger equation are depicted in blue (not drawn to scale). The vertical dashed arrow indicates the 157-nm excitation wavelength used for this work. (b) Calculated thermally averaged photodissociation cross section versus photon wavelength for the transition of interest. The vertical dashed line indicates the 157-nm excitation wavelength used for this work. The calculation is truncated at 170 nm, which is $\sim 0.1$ eV above the dissociation threshold of the $B{}^1\Pi$ potential.

Figure 3

Measured (points with error bars) and fit (solid lines) ${\mathrm{Be}}^{+}$ exponential decay plotted on a logarithmic vertical scale. Each of the three data sets consists of one experimental run and vertical error bars represent one standard deviation. With a saturation parameter $s$ of 0.03 and a 3% duty cycle (weak cooling), we measure a ${\mathrm{Be}}^{+}$ lifetime of 11.4(6) h that is limited by chemistry with background gas molecules other than ${\mathrm{H}}_2$ $\left(\blacksquare \right)$. Using a greater than $300\times$ larger photon flux for Doppler cooling $(•)$, the ${\mathrm{Be}}^{+}$ number decays with a 2.19(2)-h time constant that is consistent with an ${\mathrm{H}}_2$ background pressure of $7\times {10}^{-9}$ Pa ($5\times {10}^{-11}$ Torr). To mitigate ${\mathrm{BeH}}^{+}$ production, we trigger 200 pulses (500-Hz repetition rate) of a 157-nm excimer laser every 2 min $\left(▴\right)$. The resulting rate of ${\mathrm{BeH}}^{+}$ photodissociation is sufficient to increase the ${\mathrm{Be}}^{+}$ lifetime by a factor of 4.4(3) to be roughly consistent with the weak-cooling measurements.