State-of-the-art experiments are able to trap a few dozen individual ultracold molecules in optical tweezers, while scaling to hundreds or even thousands is desired. This scalability can ultimately be limited by the time it takes to rearrange these molecules to defect-free arrays, where every tweezer contains one molecule. As few defects as possible before rearranging is needed and this can be solved by controlling collisions, although collisions between molecules are generally hard to control.
An array of ultracold molecular qubits
By cooling molecules down to ultracold temperatures, they move slow enough to be trapped in optical tweezers. A large number of these tweezers can then be arranged into a two-dimensional defect-free array. Every molecule would then represent a qubit, the quantum analogue of a computer bit. The ultracold temperature also allows for ultimate control of these qubits, needed to build logical operations between qubits that can create an advantage over classical computers through, for example, quantum entanglement. One of the current hurdles in the creation process is loading single molecules efficiently into the tweezers. The standard procedure now is to load stochastically, which fills only 50% of all tweezers. “The rearrangement into defect-free arrays is time consuming and will put a limit on the scalability,” explains Tijs Karman, assistant professor in Theoretical Chemistry. “Our goal is to deterministically load with 100% success rate.”
Scaling up scalability
The idea is to load molecules into optical tweezers in multiple cycles that each load 50%. Between cycles, loaded molecules are transferred to ‘storage states’ that are rotationally excited by two additional quanta. “We have found intriguing interactions between dipolar molecules in different rotational states,” says Ph.D. student Etienne Walraven. “When far apart, these are van der Waals interactions, although they are repulsive! This is very uncommon and, in our scheme, very useful.” If in a later cycle a second molecule is loaded, this interaction decreases the probability of colliding – and for CaF even reacting – with the ‘storage’ molecule, thereby keeping it intact throughout the procedure. The interaction with stored molecules even prevents the accumulation of multiple storage molecules in the same tweezer. This scheme improves the loading efficiency for molecules to 80%, comparable to the state-of-the-art for atoms. Hopefully in the near future, this scheme will open a path for larger molecular arrays, and eventually for larger molecular quantum computers.
