A Precision Measurement of the Neutron Spin Structure Functions Using a Polarized HE-3 Target

This thesis describes a precision measurement of the neutron spin dependent structure function, g1n(x). The measurement was made by the E154 collaboration at SLAC using a longitudinally polarized, 48.3 GeV electron beam, and a 3He target polarized by spin exchange with optically pumped rubidium. A target polarization as high as 50% was achieved. The elements of the experiment which pertain to the polarized 3He target will be described in detail in this thesis. To achieve a precision measurement, it has been necessary to minimize the systematic error from the uncertainty in the target parameters. All of the parameters of the target have been carefully measured, and the most important parameters of the target have been measured using multiple techniques. The polarization of the target was measured using nuclear magnetic resonance techniques, and has been calibrated using both proton NMR and by measuring the shift of the Rb Zeeman resonance frequency due to the 3He polarization. The fraction of events which originated in the 3He, as measured by the spectrometers, has been determined using a physical model of the target and the spectrometers. It was also measured during the experiment using a variable pressure 3He reference cell in place of the polarized 3He target. The spin dependent structure function g1n(z) was measured in the Bjorken x range of 0.014 < x < 0.7 with an average Q2 of 5 (GeV/c)2. One of the primary motivations for this experiment was to test the Bjorken sum rule. Because the experiment had smaller statistical errors and a broader kinematic coverage than previous experiments, the behavior of the spin structure function g1n(x) could be studied in detail at low values of the Bjorken scaling variable x. It was found that g1n(x) has a strongly divergent behavior at low values of x, calling into question the methods commonly used to extrapolate the value of g1n(x) to low x. The precision of the measurement made by the E154 collaboration at SLAC puts a tighter constraint on the extrapolation of g1n(x) to low x, which is necessary to evaluate the Bjorken sum rule