Superfluid-Helium Converter for Accumulation and Extraction of Ultracold Neutrons

We report the first successful extraction of accumulated ultracold neutrons (UCN) from a converter of superfluid helium, in which they were produced by downscattering neutrons of a cold beam from the Munich research reactor. Windowless UCN extraction is performed in vertical direction through a mechanical cold valve. This prototype of a versatile UCN source is comprised of a novel cryostat designed to keep the source portable and to allow for rapid cooldown. We measured time constants for UCN storage and extraction into a detector at room temperature, with the converter held at various temperatures between 0.7 and 1.3 K. The UCN production rate inferred from the count rate of extracted UCN is close to the theoretical expectation.

Figure 1

Schematics of the apparatus for UCN production in superfluid ${}^4\mathrm{He}$: 1 cold neutron beam; 2 beam collimation; 3 stainless steel converter vessel; 4 flap valve for UCN extraction; 5 UCN extraction guide; 6 UCN detector; 7 cold neutron converted into a UCN; 8 trapped UCN, 9 Ni foils.

Figure 2

Time histograms of UCN count rates, measured in buildup mode at 1.26 K for various UCN accumulation times $t_0$. Cold neutron irradiation starts at $t=0$. $N$ denotes the integrated UCN counts for each of the histograms and ${\tau }_e$ the corresponding emptying time constant (see text).

Figure 3

Time histograms of UCN count rates, measured in buildup mode at 0.72 K for various accumulation times $t_0$. The corresponding emptying time constants ${\tau }_e$ are shown in Fig. 4. Note that the UCN detector was always recording events starting from $t=0$. The few counts for $0<t<7\mathrm{s}$ are due to all measurements in this time interval, which demonstrates the excellent background conditions.

Figure 4

Emptying time constants ${\tau }_e$ as a function of the UCN accumulation time $t_0$ in “storage mode” measurements for various temperatures. Note that in the measurements at 0.93 K the converter vessel was only partly filled, leading to an increased ${\tau }_e$ and larger counting statistical uncertainties.