A Search for Neutron to Mirror-Neutron Oscillations

Abstract

Lee and Yang had noted, in their Nobel prize winning paper, that observation of apparent parity violation in the weak interaction of particles could be mitigated with the introduction of a parity conjugated copy of the same particles. The existence of a parity conjugated copy in terms of the weak interaction of normal matter, called mirror matter, that does not interact with normal matter through known forces, has long been theorized. Baryon number violation is required for baryogenesis in order to explain the observed asymmetry between matter and antimatter in the universe. Neutron to mirror-neutron oscillations could be an observable baryon number violating process. In late 2000s, it was pointed out that the oscillation time of such neutron to mirror-neutron oscillation could be of the order of a few seconds, and that a magnetic field dependent ultracold neutron storage measurement could be sensitive to such oscillations. Furthermore, it was also shown that the ambient mirror magnetic field on the surface of the Earth could be as high as the Earth's magnetic field. Subsequently, two separate groups performed experiments at the Institut Laue-Langevin in search of neutron to mirror-neutron oscillations and reported having found no evidence. The limit set on the oscillation time was $\tau_{nn'} > 414~$s (90\% C.L.) for the case where the mirror magnetic field, $B'=0$, and $\tau_{nn'}>12~$s (95\% C.L.) for the case where $B'\in(0,12.5)~\mu$T. These constraints have since been improved to $\tau_{nn'} > 448~$s (90\% C.L., $B'=0$), $\tau_{nn'} > 17~\text{s}~\forall~B'\in(8,17)~\mu\text{T (95\% C.L.)}$, and $\tau_{nn'} > 27~\text{s}~\forall~B'\in(6,25)~\mu\text{T (95\% C.L.)}$. Soon after, when the results of these experiments were further analyzed by Berezhiani et al., $5\sigma$, $3\sigma$, and $2.5\sigma$ statistically significant signals for mirror-neutron oscillation in the presence of a non zero mirror magnetic field were reported. The current leading constraints upon $\tau_{nn'}$ do not exclude these signals. Thus a new experiment was required to test these claimed signals. The neutron electric dipole moment experiment based at the Paul Scherrer Institute performed a dedicated search to investigate these signals and found no evidence of neutron mirror-neutron oscillations. We thereby impose the following new lower limits on the oscillation time: $\tau_{nn'} > 388~$s (90\% C.L., $B'=0$), $\tau_{nn'} > 6~\text{s}~\forall~B'\in(0.38,25.66)~\mu\text{T (95\% C.L.)}$, and $\left(\tau_{nn'}/\sqrt{\cos(\beta)}\right) > 9~\text{s}~\forall~B'\in(5.04,25.39)~\mu\text{T (95\% C.L.)}$, where $\beta$ is the fixed angle between the applied magnetic field and the ambient mirror magnetic field, which is bound to the reference frame of the Earth. For the case when the mirror magnetic field is fixed in the cosmos, we have also for the first time placed the following constraints: $\tau^{\Omega_{\bigoplus}}_{nn'} > 7~\text{s (95\% C.L.)}~\forall~B'\in(5.20,24.43)~\mu\text{T}$, $\tau^{\Omega_{2\bigoplus}}_{nn'}~>~6~\text{s}~\forall~B'\in(5.51,25.78)~\mu\text{T}~(95\%~\text{C.L.})$, $\tau^{\Omega_{\bigodot}}_{nn'}~>~4~\text{s}~\forall~B'\in(6.77,27.18)~\mu\text{T}~(95\%~\text{C.L.})$, and $\tau^{\Omega_{2\bigodot}}_{nn'}~>~5~\text{s}~\forall~B'\in(4.20,24.17)~\mu\text{T}~(95\%~\text{C.L.})$, for modulation frequencies associated with one sidereal day, half a sidereal day, an annual year, and half an annual year, respectively. Our new constraints in the assumption of a mirror magnetic field bound to Earth are the best constraints around $B'\sim10~\mu$T, and also in the region where the mirror magnet field falls in the range of $B'>37~\mu$T. While this result excludes large portions of the three statistically significant signals indicated by Berezhiani et al., especially where at least two signal regions overlap, it does not exclude all the signals. We therefore need further tests of these signals in the vicinity of $B'\in(4,37)~\mu$T.