Fig. 2: Schematic of the experimental setup.

A cold nozzle followed by a skimmer is used to form the helium beam. A hexapole–dipole assembly is used as a spin polariser resulting in a beam polarised along the direction of the dipole field (\(\hat{{{{\boldsymbol{x}}}}}\)). A homogeneous \({{{{\boldsymbol{B}}}}}_{1}\) field along \(\hat{{{{\boldsymbol{z}}}}}\) leads to Larmor precessions in the \({xy}\) plane used to both reverse velocity-dependent spin dephasing in \({{{{\boldsymbol{B}}}}}_{2}\) and to produce an additional 90° rotation for measuring both components of the complex signal. A moveable 100 μm wire was used to modify the beam profile imaged by magnetic encoding and also perform reference measurements of the beam profile. The encoding device uses a 12-wire geometry to produce a magnetic field along the \(\hat{{{{\boldsymbol{y}}}}}\) direction, which changes linearly as a function of the \(x\) coordinate. The encoding device also includes a stronger homogeneous dipole field along the \(\hat{{{{\boldsymbol{y}}}}}\) direction, to eliminate the effect of the unwanted gradient field. A dipole–hexapole pair is used to focus particles with spins oriented along the +\(\hat{{{{\boldsymbol{x}}}}}\)-axis towards a mass spectrometer particle detector at the end of the beamline. The field directions of the polariser (dipole), \({{{{\boldsymbol{B}}}}}_{1}\), \({{{{\boldsymbol{B}}}}}_{2}\) and the analyser (dipole) are illustrated above with green arrows.