Table of Contents
Fetching ...

Rapid axial loading of a grating MOT with a cold-atom beam

Rachel Cannon, Aidan S. Arnold, Paul F. Griffin, Erling Riis, Oliver S. Burrow

Abstract

Laser-cooled atoms are increasingly being used to realise practical quantum devices, motivating the development of compact and robust atom sources. Grating magneto-optical traps (gMOTs) simplify the cold-atom source architecture but are typically vapour-loaded and provide limited atomic flux. Here we explore the loading of gMOTs from cold-atom beams. We numerically simulate loading to show that unbalanced diffracted beams deflect incoming atoms away from the trap centre, thereby strongly constraining radial loading. In contrast, axial loading injects atoms directly into the trapping volume and largely avoids these effects. We experimentally demonstrate rapid axial loading of a gMOT, achieving loading rates of $2.1 \times 10^9$ atoms~s$^{-1}$ using a moving optical molasses to transfer atoms from a 2D MOT into the gMOT. These results establish axial loading as a robust route to high-flux gMOT operation for portable cold-atom systems.

Rapid axial loading of a grating MOT with a cold-atom beam

Abstract

Laser-cooled atoms are increasingly being used to realise practical quantum devices, motivating the development of compact and robust atom sources. Grating magneto-optical traps (gMOTs) simplify the cold-atom source architecture but are typically vapour-loaded and provide limited atomic flux. Here we explore the loading of gMOTs from cold-atom beams. We numerically simulate loading to show that unbalanced diffracted beams deflect incoming atoms away from the trap centre, thereby strongly constraining radial loading. In contrast, axial loading injects atoms directly into the trapping volume and largely avoids these effects. We experimentally demonstrate rapid axial loading of a gMOT, achieving loading rates of atoms~s using a moving optical molasses to transfer atoms from a 2D MOT into the gMOT. These results establish axial loading as a robust route to high-flux gMOT operation for portable cold-atom systems.

Paper Structure

This paper contains 7 sections, 2 equations, 5 figures, 1 table.

Figures (5)

  • Figure 1: (a) Illustration of a QUAD gMOT grating, with the input and diffracted beams in the $x-z$ plane illustrated. Arrows indicate wavevectors of light input and diffracted on the grating in three dimensions, with an ellipsoid representing the position the MOT forms. The gMOT optic has a 3 mm diameter hole in its centre. The symmetry of the QUAD-style chip lends itself for a simplified simulation here. (b) Simulated trajectories for atoms launched radially (Rad) at 10–25 $\mathrm{m\,s^{-1}}$ overlaid on the gMOT average force vectors, evaluated at the arrow midpoint. Slow atoms are deflected by the diffracted beams, while faster atoms overfly the trap. 5 and 20 $\mathrm{m\,s^{-1}}$ axial (Ax) trajectories with a $\mp10^\circ$ angle illustrate how axial loading avoids these transverse deflections. Solid red lines demarcate the regions between different beam overlap geometries.
  • Figure 2: Binary greyscale representation of trapping success in a radially loaded gMOT versus loading speed and input height (dark grey: trapped; light grey: not trapped).
  • Figure 3: Binary greyscale maps of simulated trapping success in a gMOT under radial (top row) and axial (bottom row) loading. Columns left to right show trapping as a function of loading speed versus laser detuning (a,d), intensity (b,e), and axial magnetic field gradient (c,f), respectively. Dark grey regions indicate successful trapping, while light grey regions indicate no trapping.
  • Figure 4: (a) The fraction of atoms that will transmit through a 3 mm diameter aperture at a distance $z$, for different axial beam velocities of 10, 15, 20 and 25 $\mathrm{m\,s^{-1}}$ is calculated by modelling the beam's radial velocity as a Maxwell-Boltzmann distribution. (b) A schematic illustration of the 2D MOT loaded gMOT apparatus, annotated to show how the 2D MOT beams are delivered, the location of the 2D and 3D MOT coils, and the differentiation between the high and low pressure (HP and LP) sides of the vacuum. The differential pumping tube (DPT), gMOT coil former and diffractive optic mount is a single piece of 3D printed titanium, to facilitate the transfer of atoms between the HP and LP sides of the vacuum.
  • Figure 5: (a) Average loading curves (blue) for selected push-beam detunings, each obtained from an average of ten loading measurements. The dark red dashed lines show exponential fits of the form $N(t) = N_0 \left(1 - e^{-t/\tau}\right)$. The detunings shown correspond to the vertical dotted lines in (b) and (c). (b) Loading rate $R = N_0/\tau$ as a function of push-beam detuning. The maximum rate of $2.1 \times 10^9$ atoms s$^{-1}$ occurs at $37$ MHz. The dashed curve is a guide to the eye. (c) Equilibrium atom number $N_0$ extracted from the same fits. The maximum atom number of $6.2 \times 10^8$ Rb atoms occurs at $30$ MHz. (d) 2D-MOT loading rate as a function of push-beam intensity for the two circular polarisation configurations, $\sigma_{-}\sigma_{+}$ and $\sigma_{-}\sigma_{-}$. In(e), the polarisation configurations intensities are normalised to their respective saturation intensities SteckRb87.