Search

We describe a method for recording the Rabi nutation of nuclear spin polarized 3He by optically pumped cesium magnetometers. The measurement is performed by detecting the time-dependent magnetic field produced by the 3He magnetization. The observed signals are compared to theoretical models and the results are used to precisely trace the evolution of the magnetization. This procedure represents a convenient way to control and measure the Rabi flip angle and the degree of spin polarization in experiments using 3He magnetometers. The method requires only very coarse knowledge of the applied magnetic field’s magnitude.

We report on extensive studies on the intrinsic sensitivity of a combined 3 He/Cs magnetometer. The magnetometer relies on the detection of the free spin precession of nuclear spin polarized 3 He by optically pumped cesium magnetometers. We characterize the relevant processes involved in the detection and quantify their impact on the total sensitivity of the magnetometer. An expression is derived that predicts the sensitivity of this magnetometer scheme and the results are compared to experiments. Excellent agreement is found between theory and experiments, and implications for an application of a 3 He/Cs magnetometer in an experiment searching for a permanent neutron electric dipole moment are discussed.

In the first in a series of three papers on wall relaxation of spin polarized 3He we have reported on a breakdown of relaxation times which is observed after exposing the 3He containing glass cells to a strong magnetizing field. In this third paper we give a quantitative analysis of this phenomenon, based on magnetic signal detection by means of SQUIDs, on the pressure dependence of relaxation times in magnetized cells, as well as on Monte Carlo simulations of 3He-relaxation in a macroscopic dipole field. Our analysis allows to identify the contaminants as being aggregates of dust-like Fe3O4 particles (magnetite) with a radius \(R \approx 10~\mu\)m and a remanent magnetic moment of the order of m ≈O(10\(^{-10}~\)A m2). The particles are located at or close to the inner glass surface.

In this first in a series of three papers on wall relaxation of spin polarized, gaseous 3He we investigate both by theory and by experiment surface-induced spin relaxation due to paramagnetic sites in the containing glass. We present experimental and theoretical evidence that — contrary to the traditional opinion — distant dipolar coupling to paramagnetic impurities in the glass, in particular iron ions, cannot be the dominant relaxation mechanism of 3He-spins, although iron dominates the bulk static permeability. Instead dangling-bond type defects in the glass matrix are found to interact much stronger via the isotropic Fermi contact interaction. A model of paramagnetic site controlled 3He relaxation including the Fermi contact interaction is presented. With reasonable semi-empirical assumptions our model allows to describe satisfactorily the measured relaxivities, both in the dissolution-dominated regime of fused silica or borosilicate glasses of the Pyrex type as well as in the surface dominated situation of aluminosilicate glasses which have only a low permeability for He atoms. In a large sample of 1.1 litre cells, built from various aluminosilicate glasses, an average relaxation time of 150 h is reached in case contaminant ferromagnetic particles have been demagnetized beforehand. From the maximum observed value of 250 h we derive after subtraction of dipolar relaxation in the gas phase a paramagnetic surface relaxivity of ρ

In this second in a series of three papers on wall relaxation of 3He-spins we discuss relaxation in metal-coated glass cells in terms of hyperfine coupling to paramagnetic conduction electrons at the Fermi surface. This scales with the square of the work function of the coating and thereby also with its He-adsorption energy. In this sense we investigate coatings with particularly low work function and adsorption energy, namely Cs and Cs-suboxides. Although we observe a suppression of relaxation rates by two orders of magnitude as compared to bare Pyrex and fused silica walls, their temperature dependence still shows the same Arrhenius dependence as observed for bare substrates, instead of a T 3/2 dependence expected for a metallic surface. From this finding we conclude that, on one hand, the surface coverage is not complete and, on the other hand, the relaxation at the alkali surface itself is extremely slow. This finding is supported, too, by a semi-empirical estimate based on measured relaxation rates at ordinary metal surfaces, rescaled then with the respective dependence on adsorption energy.