nEDM experiments


Neutron is a convenient probe for EDM. Because of its neutral charge, neutrons can be subject to large electric fields (in a beam or in a bottle), without impacting their motions, to measure the EDM. Neutrons have a spin of 1/2 and can be readily spin-polarized using polarizing mirrors (for neutron beam) or superconducting magnets via the µ·B potential (for ultracold neutrons). Using polarized neutrons, many important studies on the parity-violating phenomena in formulating the electroweak theory have been carried out, including the search for the neutron EDM (nEDM) [Purcell1950]. Although the initial attempt on nEDM failed to reveal the anticipated parity non-conservation, the non-observation of the nEDM severely suppresses the QCD θ term, giving rise to the strong CP problem. Various BSM models are being constrained by comparing to the ever-increasing precision of nEDM measurements. Today, motivated by the need to find new sources of CP violation for baryogenesis, the community of fundamental neutron physics has been stepping up the effort worldwide to search for nEDM.

Our research on nEDM is highlighted in a short NPR radio program: Why Corned Beef Sandwiches — And The Rest Of The Universe — Exist

The EDM violates both the Parity-inversion symmetry and the Time-reversal symmetry. To experience the time-reversal symmetry, click here.


Beam experiments

Searches for the neutron EDM (nEDM) began in 1949 by Purcell and Ramsey [Smith1957] at the Oak Ridge reactor. A beam of neutrons, polarized using a magnetized iron mirror, passed through a region of magnetic and electric fields; upon exit, the spin state of the neutrons was analyzed by reflecting off another magnetized iron mirror. The nEDM was measured through the nuclear magnetic resonance (NMR): EDM interactions induced a shift in the Larmour precession frequency when the neutron was subject to a strong external electric field in parallel with a weak magnetic field.  In the attempt to improve the NMR signal, Ramsey developed the well-known method of separated oscillatory fields [Ramsey1950]: by separating the oscillating field that rotated the neutron spin into two regions, a much narrower resonance was achieved. The resulting interferogram contained narrow fringes, the width of which scaled inversely with the duration of free spin precession. The precision in the precession frequency measurements could be enhanced by increasing the length of the region of the interaction fields and by using slower neutrons.

Continued beam experiments in 1969 in the Brookhaven high flux beam reactor revealed a systematic effect associated with the motional field [Cohen69, Baird69].  As polarized neutrons moved through the interaction region of static electric and magnetic fields, they experienced an additional magnetic field, v/c × E, arising from the relativistic effect, first pointed out in the context of an atomic EDM search [Sandars1964]. Coupled with a small, but inevitable misalignment (of a few mrad) between the E and B fields, the strength of the total magnetic field experienced by the neutrons varied upon the field reversals. This effect gave rise to a systematic shift to the precession frequency that limited the EDM sensitivity to 10-21 e-cm. Subsequent improvements involved routinely rotating the whole apparatus to reverse the direction of the neutron beam through the apparatus; the best nEDM sensitivity using neutron beams was carried out at the ILL high flux reactor in 1977 and achieved 3×10-24 e-cm [Dress77].

UCN experiments

Ultracold neutrons (UCN) have velocities up to 5 m/s, wavelengths of 500 Å, and an effective temperature of order 3 mK. These neutrons can be easily polarized and stored inside room-temperature material bottles for several hundreds of seconds. Ramsey's method of separated oscillatory fields, applied to a bottle of stored UCN in a precisely controlled magnetic field, resulted in superb interferograms with extremely narrow fringes, giving orders of magnitude enhancement in the precision of resonant frequency measurements. To measure the nEDM, an additional large electric field (E) is applied parallel to the magnetic field (B); the neutron precession speeds up (or slows down) in proportion to the strength of both the applied electric field and the size of the nEDM.  The precession frequency is ν = -2 (μ ⋅ B + d ⋅ E)/h. The change in the Larmor precession frequency between the two configurations with the B and E fields parallel and anti-parallel measures the nEDM. The nEDM sensitivity per cycle is 
σ dn ≈ ħ/(2 α E Tfp √N),
where ħ is Planck's constant, α is the visibility of the Ramsey fringes, E is the strength of the electric field, Tfp is the time allowed for neutron free precession, and N is the number of neutrons per cycle counted by the UCN detector in the spin analyzer. Because the velocity of the UCN in a bottle averages to 0, i.e., ⟨ v ⟩ =0, the dominant systematic effect of motional field associated with nonparallel E and B fields is greatly suppressed.

In my group, we are involved with two efforts in the US to search for the nEDM:



Despite the many advantages of UCN to improve nEDM measurements, the main bottleneck limiting the progress of nEDM experiments worldwide is the density of UCN.  In the US, its first UCN facility was developed in 2000 and became operational in 2005 at the Los Alamos Neutron Science Center (LANSCE). Other UCN-producing facilities hosting active physics measurement programs are the ILL (in France) and PSI (in Switzerland). Another UCN facility coming online is at TRIUMF (in Canada).

At LANSCE, the newly-completed upgrade of the UCN facility [Ito2018] provides the necessary UCN density to meet the demand of an nEDM experiment with tenfold sensitivity improvement. A factor of 5--6 increase in the UCN output has already been achieved (as measured both in the UCNτ experiment [Gonzalez2021] and in an nEDM test apparatus [Ito2018]). The LANL nEDM experiment takes the same Ramsey approach by using a room-temperature apparatus coupled to the newly-upgraded, solid deuterium-based UCN source. The apparatus operates in a vacuum and uses Ramsey's method of separated oscillatory fields, which is a mature technology developed in prior nEDM experiments [Pendlebury2000, Baker2006}. The low-risk technology together with the high-yield UCN source at LANL opens up a timely opportunity to substantially increase the nEDM sensitivity before the nEDM@SNS experiment becomes fully operational.

Click here for more details.



The nEDM@SNS experiment has been under development for the past two decades as it involves many technological innovations to enable nEDM breakthroughs. In 1994, Golub and Lamoreaux proposed a new method [GOLUB1994] to improve EDM measurements. It calls for the innovative use of superfluid helium as the UCN production target as well as a noble-liquid detector to measure the neutron precession. Performing an experiment immersed in a bath of superfluid helium, a significant improvement in all of E, N, and Tfp is expected, with a goal sensitivity of δ dn = 3 × 10-28 e-cm. The nEDM@SNS apparatus also has double chambers, with the E fields pointing in opposite directions. In this (sub-Kelvin) cryogenic environment, 3He atoms replace 199Hg as the comagnetometer. One major advantage of using 3He dissolved in superfluid LHe as a comagnetometer is its ability to study the systematic effects due to the geometric phase effect (the known leading systematic effect) by varying the temperature: changing the temperature changes the phonon density in LHe, thereby changing the mean free path of He atoms, which in turn changes the aforementioned correlation function. The spin-dependent cross-sections for neutron absorption on 3He is used to analyze the neutron spin precession. Scintillation in liquid He, produced when the reaction products of the neutron capture on 3He, travels in the medium signals the time at which the spins of the neutrons and 3He are anti-parallel.  

The nEDM@SNS apparatus will be capable of two different readout techniques: the dressed-spin method originally proposed by Golub and Lamoreaux [GOLUB1994], and a free-precession method without spin dressing in which the 3He magnetization is directly detected by SQUID magnetometers. In the dressed-spin method, a strong (~0.1 mT), 1 kHz magnetic field (“dressing field”) is applied transverse to the static electric and magnetic fields, with the dressing field parameters set such that the neutron and 3He have the same effective gyromagnetic ratio (“critical dressing”). At critical dressing, the difference between neutron and 3He effective precession rates are insensitive to changes in the (near-static) magnetic field. Dressing field parameters are varied around this critical dressing condition and a feedback-null technique is designed to be sensitive to a neutron EDM.


To find more in-depth discussions on the EDM research, read this EDM whitepaper.