ORIGINAL

Abstract

Large-scale quantum networks will require reliable quantum memory nodes capable of dis- tributing entanglement over long distances. The Er3+ ion is a good candidate for this purpose as it has an optically accessible transition at 1550 nm, the so-called “telecom” wave- length at which light propagates through optical fibers with minimal loss. However, this transition is electric-dipole forbidden, and so spontaneous emission from it is slow (around 100 Hz). Such a bandwidth constraint would be crippling in a real photonic quantum system. Previously, the Thompson lab has used Er3+-doped Y2SiO5 coupled to silicon nanophotonic cavities, enabling the observation of single ions and enhancing the spontaneous emission rate by a factor of 600 through the Purcell effect. However, the bulk-doped Er3+:Y2SiO5 sys- tem has two flaws. One is that Y2SiO5 has a high concentration of nuclear spins which limit the T1 and T2 spin coherence times of the Er3+ ions. The other is that the lower bound on the density of Er3+ possible through the bulk-doping process is still too high for individual Er3+ ions to be addressed optically. These issues have motivated a search for new host crystals which can be doped with Er3+ through ion implantation, hopefully resulting in substantially lower concentrations of Er3+ ions than would be possible through bulk doping. In this work, we characterize the spectral properties of Er3+ implanted in the promising new host crystal MgO, and discuss new techniques for improving the resolution of future such measurements.