DOE 2/11/05 #1 EDM R&D Progress Steve Lamoreaux, Los Alamos Co-spokesperson for the EDM Project for presentation to The Department of Energy Cost & Schedule Review Feb. 11, 2005 Electric Dipole Moment of the Neutron: e·cm in Superfluid He

DOE 2/11/05 #2 Overview We are developing a new experimental technique to search for the neutron electric dipole moment (EDM) that offers a factor of at least 50 increase in sensitivity over existing experiments when operated at LANSCE (500 fold increase at SNS)

DOE 2/11/05 #3 Collaborating Institutions Hahn-Meitner Institut, Berlin NIST/Gaithersburg Harvard Simon Fraser University Caltech University of Illinois Los Alamos National Laboratory Berkeley Duke Oak Ridge National Laboratory University of Leiden University of New Mexico North Carolina

DOE 2/11/05 #4 Definition of the EDM A permanent EDM d: separation of the charged constituents of the neutron The current experimental technique (ILL) will likely yield d< ecm We hope to obtain roughly < ecm with UCN stored in superfluid 4 He +- s = 1/2dE

DOE 2/11/05 #5 Evolution of Experiments

DOE 2/11/05 #6 Motivation: To Elucidate the Nature of Time Reversal Asymmetry Physics Beyond the Standard Model: CP violation is phenomenologically introduced at present, based on K 0 decay parameters. New sources of CP violation in B decay must be interpreted in background of usual CP violation Supersymmetry Big Bang Baryogenesis: A new source of CP violation is required to explain the observed matter- antimatter asymmetry in the Universe QCD  Parameter (10 -9 by present limits) Motivation for Axion searches

DOE 2/11/05 #7 Supersymmetry

DOE 2/11/05 #8 Physics Beyond the Standard Model

DOE 2/11/05 #9 The Basic Technique + - E s = 1/2 dipole moment d n Look for a precession frequency f n =gB ± 2dE B e cm in a 10 kV/cm electric field corresponds to a shift in frequency of 0.5  Hz

DOE 2/11/05 #10 ILL Experiment 1 UCN/cc 10 kV/cm 100 sec store 2 x e cm 199 Hg Co-magnetometer—Systematic control A new systematic “geometric phase” effect was found in this work (controllable at the e cm level for our experiment)

DOE 2/11/05 #11 Competition: New ILL Experiment Funded by PPRP for construction Also work at PSI—not considered competitive

DOE 2/11/05 #12 Figure of Merit for EDM Experiments  x 180 when operated at LANSCE E  5E  5  N  N By performing the experiment directly in superfluid helium-4 (dielectric properties + superthermal production) that is doped with polarized helium-3 which serves as a magnetometer and spin precession analyzer

DOE 2/11/05 #13 Superthermal Source of UCN

DOE 2/11/05 #14 Liftime  in a Storage Cell Goal: 500 seconds 0.5 cc/sec production rate at LANSCE

DOE 2/11/05 #15 Our Proposed Experiment Light guides Photomultipliers Magnetic shield Vacuum enclosure Superconducting shield HV variable capacitor HV electrode Ground electrodes Neutron beam

DOE 2/11/05 #16 3 He Magnetometry d n dipole moment d 3 =0 Look for a difference in precession frequency f n -f 3 =(  n -  3 )B ± 2dE dependent on E and correct for temporal changes in B 0 by f 3 - EBEB s = 1/2 n 3 He f n =  n B ± 2dEf3=3Bf3=3B

DOE 2/11/05 #17 Co-Magnetometer

DOE 2/11/05 #18 Light Detection 3

DOE 2/11/05 #19 Role of SQUIDs SQUIDs M. Espy, A. Matlachov ~100 cm 2 superconducting pickup coil Flux = 2 x Tm 2 = 0.1  0 Noise = 4 m  0 /Hz 1/2 at 10  Hz ~ T 1/2 Although sensitivity to changing magnetic fields is reduced by a factor of 10, we can do better. Monitor 3 He precession to get volume average of the magnetic field; eliminate (g n -g 3 )B 0 t term to high precision

DOE 2/11/05 #20 Low Field NMR (LANL) Recent progress in P-21 demonstrates SQUID-based NMR in fields of ~ 100 mG Data shown for natural polarization of protons in water pre-polarized in a 50 G field. System uses the same tangential gradiometer proposed for EDM experiment Noise 80 pG/sqrt(Hz), SQUID performance well characterized Gradiometer design will minimize effects of vibration and coupling of holding fields We have developed electronics allowing SQUIDs to survive pulsed fields for NMR Magnetic fields proposed for EDM are ~ 3 mG ; 3 He precession ~10 Hz 3 He precession frequency

DOE 2/11/05 #21 Low Field NMR (LANL)

DOE 2/11/05 #22 Operation of Experiment Fill cell with superfluid helium, doped with freshly polarized 3He Accumulate UCN by downscattering appropriately polarized cold neutrons for about 1000 sec while ramping up HV Flip spins 90o to B 0 by RF pulses Observe scintillation signal and SQUID signal as a function of time for 1000 sec Ramp HV to zero, drain cell of spent 3He

DOE 2/11/05 #23 3 He Atomic Beam Polarizer (LANL) Final testing is in progress Expected flux detected Average velocity < 100 m/s Polarization measurements are consistent with the 100% expectation Differential pump stages will be added soon, final tests will be completed

DOE 2/11/05 #24 Other Technical Achievements The diffusion of helium-3 in superfluid helium-4 has been measured and characterized; this is an important parameter for controlling the geometric phase systematic Ultracold Neutrons were produced at LANSCE by scattering cold neutrons in superfluid helium; 180 second cell lifetime was due to the superfluid fill hole. Production rate extrapolated to improved moderator, higher target current, better guides is 0.5/cc/sec, implying 250/cc UCN density at FP12 of LANSCE A helium isotopic purification apparatus has been operated A realistic technical scheme for operating the experiment has been developed We have acquired and operated a dilution refrigerator

DOE 2/11/05 #25 Gravitational Shift Due to difference in the effective temperature of the UCN and 3He atoms, there can be a displacement between the centers- of–gravity; this places a constraint on systematic magnetic field gradients This is 1.5 mm for UCN, for h=10cm and T=5mK – Systematic magnetic gradient must be less than 10 pG/cm for e cm – 1 nA leakage (1/4 loop) gives a possible systematic of 5 x e cm

DOE 2/11/05 #26 Electric Field Systematic Effects For atoms contained in a cell, However, the effective field adds in quadrature with the applied static magnetic field. The net effect depends on the time between wall collisions, but in the case where there are many precessions between wall collisions,

DOE 2/11/05 #27 Gradient interference with E x v field Radial gradient v x E field Changes sign with direction

DOE 2/11/05 #28 Monte Carlo calculation of shift

DOE 2/11/05 #29 Magnetic gradient requirements e cm requires a< 10  G/m  G/cm easy to achieve Use gravitational offset to tune gradient to zero: 10 Hz Larmor frequency implies 3 ppm difference in ratio of magnetic moments at maximum allowed gradient. The effect changes sign with direction of B 0.

DOE 2/11/05 #30 Pseudomagnetic Field The polarized 3He creates an effective magnetic field for the UCN, corresponding to a Larmor frequency of about 1 mHz The anticipated sensitivity per cycle is about 1  Hz In order to eliminate this potential noise source, the spin flip must be controlled with an accuracy of 0.1%

DOE 2/11/05 #31 Purification of 3 He The McClintock heat flush technique requires that the superfluid helium be above 1 K The per cycle heat energy is too high to bring the 10 liters that must purified to this temperature, then back to 0.3 K There is a narrow temperature window where the propagation of 3 He atoms in the superfluid is sufficiently close to ballistic while there is a sufficient population with velocity sufficiently high to evaporate from the bath (2.8 K binding energy). The range is 0.3 to 0.4 K By using a charcoal pump over a large surface area “vat,” the 3 He atoms can be pumped away on a time scale comparable to the experiment cycle time.

DOE 2/11/05 #32 Dressed Spins By applying a strong non-resonant RF field, the spins can be modified or “dressed” For a particular value of the dressing field, the neutron and 3 He magnetic moments are equal This is an ideal situation: one simply tunes the dressing parameter until the relative precession is zero, and determines how this parameter changes as a function of electric field direction There are a number of difficult technical issues that need to be address, which includes gradients of the RF field and eddy currents

DOE 2/11/05 #33 Principle of Dressed Spin when We want B rf >> B 0 (1-10 mG) so B rf is around 1 G,  rf /2  near 3 kHz RF field must be homogeneous at the 0.1-1% level Heating and gradients due to eddy currents present design challenges Eliminates need for SQUID magnetometers and potentially increases the sensitivity of the experiment

DOE 2/11/05 #34 Conclusion We believe that the systematic effects can be controlled at a level sufficient to achieve a measurement accuracy of e cm Many of the technical issues have been studied, many remain Achieving this level of sensitivity would severely constrain most, if not eliminate all, supersymmetric extensions to the Standard Model The presence of a co-magnetometer is absolutely essential to achieve this level of accuracy