Production of ultracold neutrons from a cold neutron beam on a ${}^2{H}_2$ target

The production rates of ultracold neutrons (UCN) from cold neutrons on gaseous, liquid, and solid deuterium targets have been measured. The comparison of the measured and calculated UCN production on gaseous ${}^2{H}_2$ is used to calibrate the simulated target extraction and transport efficiencies of the experimental apparatus. The production cross section in solid ${}^2{H}_2$ at 8 K for UCN with energies between 0 and 250 neV is $R_{\mathrm{solid},8\hspace{1em}K}={\sigma }_{\mathrm{solid},8\hspace{1em}K}^{\mathrm{CN}\to \mathrm{UCN}}\hspace{1em}\rho =(1.11\pm 0.23)\times {10}^{-8}{\mathrm{cm}}^{-1}$. This value is consistent with other experiments in which UCN had been extracted from ${}^2{H}_2$. The value also agrees with calculations using the incoherent approximation and a simple Debye model and corroborates predictions for UCN densities expected at the high-intensity UCN source at the Paul Scherrer Institut. The temperature dependence of the UCN production in solid ${}^2{H}_2$ down to 8 K can be explained within the same model when multiple-phonon excitation is included.

Figure 1

(Color online) Schematical setup (not to scale): The CN beam of diameter 40 mm is collimated over a length of about 1.5 m before hitting the ${}^2{H}_2$ target sitting behind a 38 mm aperture. UCN can leave the target and reach the guide system. A mirror reflects UCN upward where they can reach the storage bottle at about 1 m height after being reflected by a second mirror. Two UCN shutters allow storage measurements. The UCN setup is rotated by ${90}^{°}$ around the vertical UCN guide, moving the detector out of the CN beam axis.

Figure 2

Cold neutron flux vs neutron energy at the exit of the FUNSPIN neutron guide [17].

Figure 3

(Color online) Sideview of the Al target cell mounted on the He-flow cryostat. The optical windows allow inspection of the full target volume accessible to the neutron beam. Three thermometers and two heaters are installed on the target for the control of the solidification and cooling process. The 80 K thermal shield has been removed for the photograph. The neutron beam axis is horizontal and perpendicular to the optical axis. For details, see [18].

Figure 4

(Color online) Number of neutrons detected in flow-through mode per 1 C of proton charge onto the SINQ target as a function of ${}^2{H}_2$ temperature. The data have been corrected for background. The measurement at 85 K was with an empty cell to confirm the background correction. One data point typically required 1000 s with 1.2 mA onto the SINQ target.

Figure 5

(Color online) Same as Fig. 4 but zoomed in on the liquid and solid ${}^2{H}_2$ region. Circles are for the slowly cooled crystal; squares are for the rapidly cooled crystal. The solid line was obtained using the simulation for UCN extraction and taking into account the temperature dependences of both thermal upscattering losses and UCN production (see Sec. 4c).

Figure 6

(Color online) Background-subtracted time spectra for UCN counts during filling, storing, and emptying times of 5 s each.

Figure 7

(Color online) Background-subtracted time spectra for UCN counts during 5 s filling, 20 s storing, and 5 s emptying.

Figure 8

(Color online) Decay of the UCN density in the bottle as a function of storage time. The four measured points are given along with an exponential fit. The storage lifetime is found to be 24 ± 8 s; the equilibrium population after filling is 0.21 ± 0.03 UCN per fill.

Figure 9

Simulated efficiencies for extraction ${\varepsilon }_{\mathrm{ext}}$ and transport ${\varepsilon }_{\mathrm{tra}}$ as functions of neutron velocity inside the ${}^2{H}_2$ gas target at 24 K. Ultracold neutrons with velocities below 4 m/s in the gas target cannot reach the detector.

Figure 10

The product $P(v)$ of the simulated efficiencies ${\varepsilon }_{\mathrm{ext}},{\varepsilon }_{\mathrm{tra}}$ (see Fig. 9) and the velocity distribution $f(v)$ is displayed [see Eq. (8)] for the ${}^2{H}_2$ gas target at 24 K as a function of neutron velocity inside the target. The normalization condition for $f(v)$ is explained in the text.

Figure 11

MCNPX [23] simulation of the attenuation and spectral change of the cold neutron flux per incident source particle in liquid ${}^2{H}_2$. The spectra displayed are averages over 10 mm diameter, 5 mm long cylinders on the beam axis. The intensity varies systematically with distance, with the highest values by the entrance window and the lowest by the UCN exit window. The simulation used the IKE Stuttgart scattering kernel [24].

Figure 12

The product $P(v)$ of the simulated efficiencies ${\varepsilon }_{\mathrm{ext}}$ and ${\varepsilon }_{\mathrm{tra}}$ and the velocity distribution $f(v)$ is displayed [see Eq. (8)] for liquid and solid ${}^2{H}_2$, respectively. The different normalization conditions for $f(v)$ are explained in the text.