Research and Development Progress toward a New Search for the Electric Dipole Moment

M. Cooper (P-25), S. Lamoreaux (P-23) representing the EDM Collaboration
Excerpted from LA-14202-PR

Introduction

A nonzero electric dipole moment of the neutron (nEDM) would be an exciting discovery because it would either solve the “strong CP” problem (see below) or reveal new physics beyond the standard model of electroweak interactions. In the neutron, an EDM would arise from a slight separation of positive and negative charges along the spin axis. Such a separation allows for an interaction with an applied electric field that has space-time transformations that break the time-reversal symmetry (T). Symmetries of nature have their origin in conservation laws. No such law exists for the strong interaction, the origin of the nuclear force, and time reversal is expected to be broken. However, extremely sensitive searches for a nonzero nEDM have given null results. Furthermore, great experimental effort has been devoted to the search for a particle called the axion that could be the signature of an undiscovered conservation law; these searches have also provided negative results. A substantially more sensitive search for a nEDM would sharpen this conflict in our understanding.

A nonzero value for the nEDM is expected to arise at a very small value (~ 10^-31 e·cm) due to the violation of time reversal discovered in the strange- and bottom-quark systems. Between this (nearly) immeasurably small value and the current experimental limit of 6 × 10^-26 e·cm,¹ there exists a window for discovery of new phenomena in the standard model. Physicists have two exciting motivations to look for the nEDM: the highly popular idea of supersymmetry, which remains to be proven, predicts a value in the range 10^-26–10^-28 e·cm; the dominance of matter over antimatter in the universe resists quantitative explanation, and another source of T violation could resolve this puzzle.

An international team of scientists has proposed a new method to search for the nEDM that promises to increase sensitivity by two orders of magnitude.² To measure a nEDM, polarized neutrons are placed in a weak magnetic field and strong electric field. The magnetic field causes neutron precession about the field direction with a predictable frequency. An EDM slightly modifies the frequency in proportion to the value and sign of the applied electric field. In the new method, a small quantity of another species, polarized helium-3, is placed in the same container as the neutrons. The helium-3 has a very similar magnetic dipole moment to the neutron but is known not to have an EDM greater than 10^-29 e·cm. The helium-3 precession occurs at nearly the same rate as the neutron precession, and any shift of the neutron frequency with respect to the helium-3 frequency that is proportional to the electric field will be the signature of an nEDM.

Both the neutrons and helium-3 are contained in a measurement cell filled with superfluid helium-4. Neutrons can be bottled in a cell if their energy is low enough for them to be reflected by the Fermi potential of the walls; they are called ultracold neutrons (UCNs). UCNs are produced³ in the cell by down scattering cold neutrons off the helium-4. The cross section for absorption of neutrons by helium-3 is highly spin-dependent, and the helium-3 serves as an analyzer of the relative precession frequency between the two species. If a neutron is absorbed, the reaction products lose their kinetic energy in the helium-4, which in turn scintillates. The detection of this scintillation light measures the beat frequency between the two species. As a control measurement, the precession frequency of the helium-3 is measured with superconducting quantum interference devices. The helium-3 is referred to as a comagnetometer because it occupies the same volume as the neutrons and measures the magnetic field.

The proposal has been reviewed by the Nuclear Science Advisory Committee and has been deemed to be the experiment with the greatest discovery potential for the new fundamental-neutron-science beam line at the Spallation Neutron Source.⁴ The committee recommended a vigorous program of research and development (R&D) to validate the measurement technique and to work out the most significant engineering challenges as well as eventual funding for the project. In order to control a variety of systematic errors that could produce a false result, the experimental apparatus must meet stringent design requirements. Significant progress has been made towards validation of the method.

The Electric Field

Whereas the sensitivity of the measurement is proportional to the magnitude of the electric field, this experiment has a goal of 50 kV·cm?^-1, roughly five times that of previous measurements. Liquid helium-4 is a very good dielectric, i.e., will break down only at quite a high voltage. However, all previous measurements of its dielectric strength were performed with small surface areas and small gaps between the electrodes. The proposed experiment will have electrodes exceeding 600 cm² and a separation of 7.5 cm. We felt it necessary to demonstrate that we can achieve the desired voltage because the breakdown voltage is believed to scale as the square root of the separation.

A full-scale apparatus has been built at LANL to measure the dielectric strength in the relevant geometry. The electrodes are constructed so that the gap between them is variable. The high-voltage electrode is charged to about 50 kV with a power supply and then disconnected from the supply. As the electrode gap is increased, the voltage is multiplied to keep the field constant. Figure 1 shows a plot of the breakdown voltage as a function of separation. The curve in the figure is the breakdown voltage calculated assuming a square-root-of-the-gap dependence. The curve is normalized to the work of other investigators, all of which occurred below 10 mm. Our value of 570 ± 70 kV is the point at 73 mm. The voltage was stable for 11 h. In order to achieve 50 kV·cm?^-1 in the actual experiment, a larger variable capacitor will be attached in parallel with the fixed electrodes of the measurement cell.

The sign of the change in precession frequency with electric field for a real EDM should reverse if the sign of the field is reversed. Furthermore, at the very highest densities of UCNs, ionization produced by the decay or absorption of neutrons will discharge the electrodes slowly during the measurement. Thus, the value of the electric field needs to be known for both polarities. As part of our collaboration, Berkeley has proposed to use the Kerr effect to make an in situ measurement of the electric field. The Kerr effect is the rotation of polarized light passing through a high electric field. The size of the effect is proportional to the square of the electric field, and the proportionality constant is called the Kerr constant. Figure 2 shows the first measurement of the Kerr effect in superfluid liquid helium-4. The quadratic dependence is clear. The extracted value⁵ of the Kerr constant is (1.43 ± 0.02^(stat) ± 0.04^(sys)) × 10^-20 (cm/V)² at T = 1.5 K. The measurement at each voltage was made with roughly 3.5 cm of liquid helium-4 and took 1000 s. With a 50 cm path that is expected in the real application, the time for a measurement should be short enough to meet our needs.

The Magnetic Field

To preserve the polarization of the neutrons and the helium-3, magnetic-field uniformity must be 10^-3. The measurement cell must be shielded against μG external fields. The scheme we have selected consists of a multilayer shield of highly permeable material that surrounds a superconducting shield. The constant field is produced by a cosθ coil inside the superconducting shield. The magnetic-field boundary conditions of the superconducing shield are incompatible with the coil. The coil will produce a very uniform field if it is wound inside an additional ferromagnetic layer. Because the experiment runs at cryogenic temperatures, it is necessary to select a ferromagnetic material that preserves its permeability at such temperatures. We have studied Metglas, a commercially available amorphous metal. Figure 3 shows the measured inductance of the Metglas as a function of temperature to be sufficient to meet our needs.

The Helium-3 Comagnetometer

The handling of the polarized helium-3 requires many complicated steps. We must
• produce highly polarized helium-3,
• inject it into a superfluid helium-4 bath while maintaining the polarization,
• maintain the polarization during the measurement, and
• remove the helium-3 once it depolarizes.

The production step is accomplished with an atomic beam source. Helium-3 atoms at 1 K are injected into a quadrupole magnet. This magnet configuration produces a gradient field that is appropriate to focus only one spin through the apparatus. The other spin is removed with vacuum pumps. This type of device is known to produce very high polarization but sufficient intensity to meet our needs.

An atomic beam source has been built at LANL to verify its performance. In order to know the polarization, a second quadrupole magnet has also been built to act as an analyzer of the emitted atomic beam. The two magnets are separated by a region where radio frequency coils can manipulate the spins to measure the polarization. The polarization of the source has been measured to be 99.5 ± 0.25%. The flux was approximately 4 × 10¹⁴ atoms's.

The polarization must be maintained until the measuring cycle is complete. The problem has been broken into two parts, one where the atoms are injected into the system and one where the polarization is maintained in the measurement cell. The former is the subject of future R&D. The polarization relaxation time depends on the wall material of the container. The mechanisms for depolarizing the atoms vary significantly over the temperature range between 0.1–4.3 K. The expectation is that the relaxation time will decrease as the temperature is lowered to 1 K and then rise rapidly below 1 K. The operating temperature of the experiment is 0.3 K. The cell walls have been chosen to be deuterated styrene impregnated with a deuterated wavelength shifter in order to minimize UCN absorption on the walls and to aid in the detection of the scintillation light.

Duke University is studying relaxation times at temperatures above 2 K. An ideal relaxation time is in excess of 8 h, which implies a depolarization of a few percent during an EDM measuring cycle. The Duke investigators are comparing the relaxation times in glass cells to those in cells where beads coated with deuterated styrene have been added. Thus far, they have achieved times in excess of 7 h in pure glass cells. They see no degradation when the coated beads are added.

The EDM collaboration has measured the diffusion coefficient for helium-3 in superfluid helium-4 between 0.45–0.95 K.⁶ The measurement was carried out at Los Alamos Neutron Science Center employing neutron tomography. The results are shown in Figure 4. The measurement verifies predictions that the diffusion coefficient varies as T^-7 below 0.7 K and allowed us to determine the temperature where the ballistic velocity of neutrons equals the diffusion velocity of the helium-3. Between 0.1–0.6 K, the average velocity of the helium-3 is sufficient to allow escape from the superfluid helium-4. This realization leads towards the possible design of a system to remove depolarized helium-3 from a mixture by simply pumping on the bath.

The heat-wind technique can produce our initial charge of ultrapure helium-4 (< 10^-13 atoms of helium-3).⁷ Phonons couple strongly to helium-3 and weakly to helium-4. The phonons can be used to blow the helium-3 away from a source of heat. When liquid helium-4 at ~ 1 K is passed through a capillary tube surrounded by a resistive heater, only pure helium-4 passes. A cryostat designed for this purification process has been built at the Hahn-Meitner Institut and was operated at LANL. The first sample has been analyzed at Argonne National Laboratory and shown to have a purity of at least 10^-12.

Neutron Absorption Identification

The possibility to identify neutron absorption in liquid helium-4 has been reported by the method of after-pulses.⁸ If the scintillation light produced by other mechanisms can be discriminated against, the scintillation signal has the possibility of being nearly background free, increasing the sensitivity of the measurement. The method of after-pulses consists of plotting the number of single photoelectrons detected in the 10 μs following the main scintillation pulse plotted versus the total scintillation light. The events due to absorption produce more highly ionizing particles than background beta particles. Both the highly ionizing particles and betas excite dimers in the helium-4, but the absorption events have a longer de-excitation time and thus, more late after pulses. Figure 5 shows the effect in data taken by the collaboration at the Hahn-Meitner Insitut. These data show the promise of greatly reducing a variety of backgrounds in the experiment.

Future R&D

A challenging task is to build a single piece of apparatus to contain all these different elements simultaneously that may be assembled and serviced. The engineering staff at LANL has made great progress toward reaching a reference design.

The EDM Collaboration has made great strides in developing a new experiment for a nEDM search. The collaboration has many other projects underway. Most of the crucial performance criteria of the detector components have been demonstrated. The future R&D is focused on bringing new concepts to the experiment to reduce backgrounds and systematic errors.

References

1. P.G. Harris et al., “New experimental limit on the electric dipole moment of the neutron,” Physical Review Letters 82, 904–907 (1999).
2. R. Golub and S.K. Lamoreaux, “Neutron electric-dipole moment, ultracold neutrons and polarized ³He,” Physics Reports 237, 1–62 (1994); M.D. Cooper et al., “A new search for the neutron electric dipole moment,” Los Alamos National Laboratory report LA-UR-02-2331 (April 2002).
3. E. Golub and J.M. Pendlebury, Contemporary Physics 13, 519 (1972).
4. Report to NSAC, Subcommittee on Fundamental Physics with Neutrons (2003) 1.
5. A.O. Sushkov et al., “Kerr effect in liquid helium at temperatures below the superfluid transition,” Physical Review Letters 93, 153003-1–153003-4 (2004).
6. S.K. Lamoreaux et al., “Measurement of the 3He mass coefficient in superfluid 4He over the 0.45–0.95 K temperature range,” Europhysics Letters 58, 718–724 (2002).
7. P.V.E. McClintock, “An apparatus for preparing isotopically pure 4He,” Cryogenics 18, 201–208 (1978).
8. K. Habricht, PhD Thesis, Technischen Universtat Berlin (1998).

Acknowledgment

The authors would like to acknowledge the participation of our LANL colleagues and our offsite collaborators from the University of California, Berkeley; California Institute of Technology; Duke University; Hahn-Meitner Institut; Harvard University; Hungarian Academy of Sciences; University of Illinois, Urbana-Champaign; University of Kentucky; Air Liquide, Advanced Technology Division; University of Leiden; National Institute of Standards and Technology, Gaithersburg; North Carolina State University; Oak Ridge National Laboratory; Simon-Fraser University; and Yale University. It is only through broad international collaboration that this experimental effort has achieved the results that it has.

Some aspects of this work are funded by LDRD funds from LANL, and other aspects are funded by a grant from the U.S. DOE.

For further information, contact Martin Cooper, 505-667-2929, mcooper@lanl.gov.