Machine learner optimization of optical nanofiber-based dipole traps

Abstract

We use a machine learning optimizer to increase the number of rubidium-87 atoms trapped in an optical nanofiber-based two-color evanescent dipole trap array. Collisional blockade limits the average number of atoms per trap to about 0.5, and a typical uncompensated rubidium trap has even lower occupancy due to challenges in simultaneously cooling atoms and loading them in the traps. Here, we report on the implementation of an in-loop stochastic artificial neural network machine learner to optimize this loading by optimizing the absorption of a near-resonant, nanofiber-guided, probe beam. By giving the neural network control of the laser cooling process, we observe an increase in peak optical depth of 66% from 3.2 ± 0.2 to 5.3 ± 0.3. We use a microscopic model of the atomic absorption to infer an increase in the number of dipole-trapped atoms from 300 ± 60 to 450 ± 90 and a small decrease in their average temperature from 150 to 140 μK. The machine learner is able to quickly and effectively explore the large parameter space of the laser cooling control process so as to find optimal parameters for loading the dipole traps. The increased number of atoms should facilitate studies of collective atom–light interactions mediated via the evanescent field.

Interaction of two Rydberg atoms in the vicinity of an optical nanofibre

“…These values allow us to account for the behaviour of the potential U AB represented in figure 17 (left panel). For R = 350 nm, A 1 < A 2 , and, as expected from equation (12), the local maximum in Θ = 0 is less marked than the maximum in Θ = π 2 . By contrast, for R = 250 nm, A 1 > A 2 , and the opposite behaviour is observed.…”

“…In particular optical nanofibres (ONFs) received much attention within the past two decades [6][7][8]. Their evanescent guided modes have been used to trap [9][10][11][12][13] and detect atoms and related phenomena [14][15][16][17][18]. Recently, a super-extended guided mode, which resides almost entirely outside the fibre [19], could be achieved by using an extremely thin ONF.…”

An Expressive Ansatz for Low-Depth Quantum Optimisation

“…The surrogate model is then further trained to fit the landscape. The effectiveness of SANNs has been shown on various complex problems [105][106][107][108], and we believe that it will allow the XQAOA X=Y 1 ansatz to surpass the GW algorithm for the MaxCut problem on all graph instances. However, testing SANNs on the XQAOA X=Y 1 ansatz is out of this paper's scope and remains a topic for future research.…”

Atom-light interactions using optical nanofibres—a perspective

“…A theoretical study on atom heating in nanophotonic traps developed a general model based on particle-phonon interactions to determine the effect of mechanical vibrations of waveguides on guided light fields [110]. To optimize the number of atoms trapped in fibre-based dipole traps, a machine learning optimisation algorithm was implemented to quickly and effectively search the large experimental parameter space for laser-cooling and trap loading of 87 Rb [111]. While the initial outcomes were promising (increasing the number of trapped atoms from about 300 to 450), further improvements could be made by increasing the parameter space to include a wider selection of the experimental controls.…”