The confinement in material storage vessels - as in previous experiments - includes the disadvantage of the neutrons interacting with the walls.
In addition to the decay the neutrons can be lost by absorption or so call thermal upscattering, where they gain kinetic energy by thermal vibrations and leave the storage vessel. The result would be a shorter measured lifetime. Therefore τSPECT uses magnetic "walls" instead of material. Since the neutrons are fermions (spin 1/2), their resulting magnetic moment interacts with magnetic fields. Via the Stern-Gerlach force the neutrons experience an accerlation or rather a deceleration depending on the orientation of the spin in the magnetic field. The preferred location in high or low fields gives them either the name high field seeker or low field seeker. In τSPECT the storage field consists of a longitudinal field produced by superconducting coils (this configuration comes from the former experiment aSPECT) and an octupole made of permanent magnets which produces a radially rising field. In total neutrons with kinetic energies up to 60 neV are storable.
From the UCN source the neutrons are guided to the experiment by stainless steel tubes where they are spin polarised as soon as they reach the longitudinal magnetic field. The high field seekers are accelerated into the first high field area. But since these neutrons cannot be stored they have to be converted into low field seekers by a spin flip. This is done with a transversally irradiated magnetic field produced by an RF coil (spin flipper). As soon as the neutrons are confined, the spin flipper is pulled out of the edge of the storage volume. At this point the storage time begins. Then a neutron detector is driven into the storage volume from the other side of the experiment and counts the remaining neutrons. From the exponential decay of the measured neutrons with rising storage time the lifetime can be extracted.
Optimisation of the filling process
Phase I of the experiment targets to measure the neutron lifetime with an uncertainty of 1 s. The required total measurement time is here essentially determined by the number of neutrons that can be stored in the trap in each measurement run. Therefore, the filling process, i.e. the efficiency in the transformation of high field seekers to low field seekers, needs to be optimised. This includes an optimisation of the spin flipper RF power or the polarisation of the irradiated magnetic field.
The installation of an additional RF coil by the end of 2020 lead to a further increase in the number of storable UCN as now low field seeking neutrons from a more densely occupied higher energetic range are decelerated in a two-stage process.
Neutrons with a maximum kinetic energy determined by the strength of the magnetic field can be stored in the trap as these neutrons are reflected from the potential walls under any angles of incidence. However, neutrons with slightly higher energies can be reflected as well if they fly under sufficiently small angles with respect to the magnetic field. Such so-called marginally trapped neutrons can remain in the trap for comparably long times (order of magnitude: 100 s) until they eventually escape. This leads to an additional escape channel apart from β-decay because these marginally trapped neutrons are included in measurements with short storage times but have vanished already in measurements with long storage times. Therefore, one either has to precisely determine the time constant with which marginally trapped neutrons leave the trap or one has to remove them from the trap before the actual storage period.
In the τSPECT experiment this is done by so-called 'spectral cleaning': For this purpose the detector is moved to the fringe region of the magnetic trap. Only such neutrons with sufficiently high kinetic energies are able to reach the detector in this position and can thus be removed from the trap. Neutrons with lower kinetic energies remain in the trap. Afterwards the detector is retracted again and the actual storage time begins.
The quality of spectral cleaning depends primarily on three parameters: The position of the detector in the magnetic field, the cleaning duration, and the trajectories of the marginally trapped neutrons.
First 3D magnetic storage measurements were carried out successfully in September 2019. Since then the filling process utilising the single spin flip method was optimised for maximum efficiency. Current optimisations include the filling process using the double spin flip technique as well as the parameters for spectral cleaning.