Lawrence Berkeley National Laboratory

The beta-neutrino correlation coefficient, a_beta nu, is measured in 21Na by detecting the time-of-flight of the recoil nucleus detected in coincidence with the atomic electrons shaken off in beta decay. The sample of 21Na is confined in a magneto-optic trap. High detection efficiency allows low trap density, which suppresses the photoassociation of molecular sodium, which can cause a large systematic error. Suppressing the fraction of trapped atoms in the excited state using a dark trap also reduces the photoassociation process, and data taken with this technique are consistent. The main remaining systematic uncertainties come from the measurement of the position and size of the atom trap, and the subtraction of background. We find mbox a_beta nu=0.5502(60), in agreement with the Standard Model prediction of mbox a_beta nu=0.553(2), and disagreeing with a previous measurement which was susceptible to an error introduced by the presence of molecular sodium.