Spin Coherence Time studies for the storage ring EDM search

This thesis is part of the feasibility studies for a search for an Electric Dipole Moment (EDM) of charged particles in a storage ring. The evidence for a non-vanishing EDM at the sensitivity of present or planned experiments would clearly prove the existence of new CP violating meachanisms beyond the Standard Model. The proposed solution to measure the EDM of charged particles is the use of a storage ring where the polarized charged particle beam can be kept circulating while interacting with a radial electric field. Starting with a longitudinally polarized beam, the EDM signal would be detected as a polarization precession starting from the horizontal plane and rotating toward the vertical direction. A long horizontal polarization lifetime, called spin coherence time, is required since it represents the time available to observe the EDM signal. In order to have a sensitivity about 10−29 e·cm to the deuteron EDM, the spin coherence time should reach 1000 s while the measurement of a vertical polarization change should detect angles as small as micro-radians. The aim of this work is the analysis of the mechanisms which control the spin coherence time in a storage ring. The measurements presented here were made at the COSY (COoler SYnchrotron) ring located at the Forschungszentrum-J¨ulich GmbH (Germany). There are two set of measurements presented in this thesis: the first is a study of a spin resonance induced by a radio-frequency (rf) solenoid and the second shows the results from the first direct measurement of the horizontal polarization as a function of time. The first experiment sought to estimate the spin coherence time by measuring the width of a deuteron spin resonance induced by an rf-solenoid. Since the width of the resonance depends on the spin tune spread and thus on particle momentum distribution, each mechanism that can change the particle velocity in the beam could contribute to the spin tune spread. In particular, these mechanisms are betatron oscillations related to the beam emittance and synchrotron oscillations that are present only in a bunched beam. The experiment consisted in the measurement of the vertical polarization measurements with the rf-solenoid running at fixed frequency on and off resonance, for both uncooled and cooled bunched beam. In order to interpret the data, a simple “no-lattice” model was developed based on two rotation matrices for the spin precession about the vertical axis and the solenoid kick about the longitudinal axis; synchrotron oscillations were included as simple harmonic motion. The model demonstrated that the effect of synchrotron oscillations on the induced spin resonance were large enough to hide any dependence on emittance. The second experiment was the direct measurement of the horizontal polarization as a function of time. This task was accomplished through the development of a dedicated data acquisition system synchronized with the revolution frequency of the beam. By changing the horizontal beam emittance with a white noise electric field, the measurements gave the first experimental evidence of a dependence of the spin coherence time on the horizontal beam size. The dependence is due to the path lengthening introduced by betatron oscillations which forces the particles to go faster in order to respect the isochronous condition in a bunched beam. A possible method to correct for emittance effects is to use sextupole magnets. In fact the field varies as the square of the radius from the center and provides an adjustment to the particle orbit to remove the term driving the spin tune change. It has been demonstrated that for a particular value of sextupole strength the contribution from the horizontal emittance was canceled, reaching a spin coherence time of a hundred seconds.