Precision Determination of the Neutron Spin Structure Function g1n

Abstract

We report on a precision measurement of the neutron spin structure function $g^n_1$ using deep inelastic scattering of polarized electrons by polarized ^3He. For the kinematic range 0.014<x<0.7 and 1 (GeV/c)^2< Q^2< 17 (GeV/c)^2, we obtain $\int^{0.7}_{0.014} g^n_1(x)dx = -0.036 \pm 0.004 (stat) \pm 0.005 (syst)$ at an average $Q^2=5 (GeV/c)^2$. We find relatively large negative values for $g^n_1$ at low $x$. The results call into question the usual Regge theory method for extrapolating to x=0 to find the full neutron integral $\int^1_0 g^n_1(x)dx$, needed for testing quark-parton model and QCD sum rules.

Figures (3)

• Figure 1: Results for $g_1^n$ versus $x$ from SLAC Experiment E154 compared to Experiment E142 evaluated at $Q^2$ = 5 (GeV/c)$^2$. Shaded region corresponds to 1 $\sigma$ systematic uncertainties.
• Figure 2: Difference between the measured proton [5,6] and neutron [This experiment] integrals calculated from a minimum $x$ value, $x_{\rm min}$ up to $x$ of 1. The value is compared to the theoretical prediction from the Bjorken sum rule which makes a prediction over the full $x$ range. For the prediction, the Bjorken sum rule is evaluated up to third order in $\alpha_s$Lar:91PDG:96 and at $Q^2$ = 5 (GeV/c)$^2$. Error bars on the data are dominated by systematic uncertainties and are highly correlated point-to-point.
• Figure 3: Results for $g_1^n$ versus $x$ for the low $x$ region from SLAC experiment E154 compared to the CERN SMC experiment. The data is evolved to $Q^2$ = 5 (GeV$^2$/c$^2$). Fits that impact the low $x$ extrapolation (discussed in the text) are presented.