Determination of the neutron lifetime using magnetically trapped neutrons

We report progres upon an experiment to measure the neutron lifetime using magnetically trapped neutron Neutron are loaded into a 11 T of great depth superconducting Ioffe-type trap by scattering 089 nm neutron in isotopically fair superfluid [.sup.4]He. Neutron decays are exposeed in real time using the scintillation light produc in the helium by means of the beta-decay electrons. The measured trap lifetime at a helium temperature of 300 mK and with no ameliorative magnetic ramping is substantially shorter than the independent neutron lifetime. This is attributed to the vicinity of neutrons with energies higher than the magnetic potential of the trap. Magnetic field ramping is implemented to eliminate these neutron resulting in an [833sub-63sup+74] s trap lifetime, consistent with the popularly accepted value of the independent neutron lifetime.

lock opener words: magnetic trapping; neutron lifetime; superthermal neutron production; ultracold neutrons

1 Introduction

We not away a progress report on an experimental program to improve the measurement of the neutron lifetime, [[tau].sub.n], using a technique with completely different systematic results than previous measurements [1]. Ultracold neutron (UCN) are produc by dint of inelastic scattering of cold (089 nm) neutron in a reservoir of superfluid [.sup.4]He (the "superthermal" process) These neutron are then confined through a three-dimensional magnetic trap. As the trapped neutron beta decay, the resulting energetic electron generate scintillations in the liquid He. Each decay is detectable with high effciency. Thus, [[tau].sub.n] can be directly determined by means of measuring the scintillation rate as a function of time.

For detailed information upon the experiment, the reader is directed to the graduate thesis of s N. Dzhosyuk [2] and Ref [3] and [4] This paper summarizes more [i]or[/i] less highlights of the neutron trapping/lifetime data gathered at the National nstitute of Standards and Technology (NIST) from the fall of 2002 until the summer of 2003

2 Experimental Procedure

In order to correct our data for the pair time-dependent and time-independent backgrounds, data is assembleed in what we refer to as "trapping" and "non-trapping" move swiftlys In a trapping run, the magnet is energized while neutron are loaded into the trap. After the beam has been make go rounded off, the neutron decay incidents are recorded. In non-trapping (or background) races the magnet is deenergized while the beam is upon then raised to the replete value as the neutron beam is revolveed off. In this non-trapping case, the background facts arising from neutron activation, neutron-induced luminescence, etc should be the same. A difference in the enumerate rate versus time between trapping (trapped UCN + backgrounds) and non-trapping (backgrounds only) races should arise solely due to magnetically trapped UCN If for more [i]or[/i] less reason the backgrounds are not identical in the trapping and non-trapping move swiftlys then the subtraction process will leave a residual difference that could mimic a trapping signal. Measurements made with natural abundance helium are used to conclusively determine that a putative trapping signal is, in fact, to be paid to trapped neutrons.

[FIGURE 1 OMITTED]

the couple trapping and non-trapping data was bring togethered in a number of configurations that will be described below. Analysis of each data plant is performed by integrating the pulsation area of each digitized photomultiplier tube (PMT) signal and applying appropriate lower horizontal threshold cuts on the area of the pulsations Since neutron-induced luminescence (occurring as single uncorrelated photons) is known to be not away in the coincidence data with doorsills at single photoelectron levels, doorsills are set to require an area in each throb equivalent to at least three photoelectrons.

A representative station of data (from approximately 8 weeks of data collection) is shown as the upper bend in Fig. 1. The bend is obtained by taking the difference between the trapping and non-trapping move swiftlys and then fit to the function y = [ysub0] + A exp(-t/[tau]) with parameter estimates [ysub0] = 004 [+ or -] 001 [ssup-1] A = (194 [+ or -] 003) [ssup-1] and [tau] = ([621sub-17sup+18]) s with [chi square] = 096 where [chi square] is the reduc chi-squared value. As single can see, the lifetime obtained from this data is substantially shorter than the forthwith accepted value of the neutron lifetime (8857 [+ or -] 08) s (1 [sigma] uncertainty) [1]. after runs have led us to identify this systematic result in our system as arising from marginally trapped neutrons

3 inquiry of Systematics

A wide range of experimental configurations has been explored to the couple measure the lifetime of UCN in our magnetic trap and to understand the observ shift in the measured trap lifetime to be paid to systematic effects. The origin of the systematic shift is not completely understood, on the contrary as will be shown below, it appears to arise from a combination of marginally trapped neutron and material bottling. Lifetime measurements beneath different experimental conditions yield values for the lifetime in the range of 600 s to 900 s, a spread larger than the statistical uncertainty.