New source for ultracold neutrons at the Institut Laue-Langevin

A new intense superthermal source for ultracold neutrons (UCN) was installed at a dedicated beam line at the Institut Laue-Langevin. Incident neutrons with a wavelength of 0.89 nm are converted to UCN in a 5-liter volume filled with superfluid ${}^4\mathrm{He}$ at a temperature of about 0.7 K. The UCN can be extracted to room temperature experiments. We present the cryogenic setup of the source, a characterization of the cold neutron beam, and UCN production measurements, where a UCN density in the production volume of at least 55 per ${\mathrm{cm}}^3$ was determined.

Figure 1

Schematic drawing of the cryostat of the UCN source: (1) isolation vacuum vessel, (2) cold traps and heat exchangers for incoming helium gas thermally anchored to the first stage of the cryocooler at about 45 K, (3) regenerator heat exchangers, (4) second stage of the cryocooler with heat exchangers, (5) cold needle valve of the ${}^4$ He circuit, (6) 1-K ${}^4\mathrm{He}$ evaporation pot, (7) cold needle valve of the ${}^3\mathrm{He}$ circuit, (8) thermal baffles, (9) outer heat shield attached to the first stage of the cryocooler, (10) inner heat shield attached to the second stage of the cryocooler, (11) superleak filter, (12) ${}^3\mathrm{He}$ evaporation pot, (13) copper heat exchanger, (14) “U”-shaped connection pipe, (15) aluminum converter vessel filled with superfluid ${}^4\mathrm{He}$, (16) UCN storage and production volume, (17) vertical stainless steel UCN guide, (18) UCN shutter valve thermally anchored to the inner heat shield, (19) thin aluminum cold neutron beam window, and (20) stainless steel UCN guide at room temperature.

Figure 2

Copper heat exchanger between the ${}^3\mathrm{He}$ pot and the superfluid ${}^4$ He bath in the converter vessel. The eroded fins (28-mm long and 0.5-mm thick) result in a surface area of about 300 ${\mathrm{cm}}^2$ on each side.

Figure 3

(a) One of the five 20-cm long BeO ceramic pieces forming the 5-liter UCN storage volume. Its inner cross section is $70\times 70$ ${\mathrm{mm}}^2$. The hole in the piece shown serves for extraction of the UCN via a stainless steel guide (see text). (b) Front view of the opened aluminum converter vessel with the installed white BeO ceramic pieces. The storage volume is closed off with two 1-mm thick beryllium sheets (right) serving as entrance and exit windows for the cold neutron beam. The spring mechanism used to press the ceramic parts together is presented on the left. (c) Closed converter vessel with an indium-sealed thin aluminum entrance window (approximate converter vessel length, 1 m).

Figure 4

Time plot of a cool-down of the UCN source cryostat starting from room temperature. After about 7 days the filled converter reaches a temperature of 0.85 K (measured outside the converter at the UCN extraction side; see text), while the temperature in the ${}^3\mathrm{He}$ pot is 0.55 K.

Figure 5

Schematic top view of the beam line: (1) neutron guide H172 coming from the vertical cold source of the ILL reactor, (2) monochromator crystals (dashed lines) attached to triangular shaped holder, (3) long neutron guide section of H172a, converging from $80\times 80$ ${\mathrm{mm}}^2$ to $70\times 70$ ${\mathrm{mm}}^2$ and tilted by $\vartheta =61.2$° with respect to the incoming neutron beam, (4) secondary beam shutter, (5) short neutron guide section of H172a, (6) UCN source with helium filled converter vessel, (7) lead shielding with a thickness of 7 cm, (8) beam stops, and (9) concrete wall for biological shielding.

Figure 6

Time-of-flight spectrum taken at the exit of the secondary neutron guide in the center of the beam (corrected for neutron scattering and absorption in air and normalized to a reactor power of 53.2 MW). The positions and widths of the four peaks have been fitted with independent Gauss functions (red lines): ${\lambda }_1=(0.91\pm 0.02)$ nm, ${\sigma }_1=0.0284\left(9\right)$ nm, ${\lambda }_2=(0.46\pm 0.01)$ nm, ${\sigma }_2=0.0134\left(4\right)$ nm, ${\lambda }_3=(0.31\pm 0.01)$ nm, ${\sigma }_3=0.0114\left(9\right)$ nm and ${\lambda }_4=(0.56\pm 0.01)$ nm, ${\sigma }_4=0.0142\left(5\right)$ nm. The given errors for the peak positions are dominated by the 2% uncertainty in the wavelength calibration (compare text). The errors on the peak width are a combination of statistical uncertainties and the wavelength calibration.

Figure 7

Pinhole image of the beam of neutron guide H172a $(320\times 240$ pixel). Dark (light) pixel correspond to a high (low) neutron flux intensity, i.e., a high (low) reflectivity of the crystals. The dashed lines indicate the central square which represents the image of the direct beam. The eight surrounding images are from mirror reflections in the secondary neutron guide.

Figure 8

Setup of the new UCN source at the beam line H172a. The cryostat containing the superfluid helium converter is housed in a lead castle. Shown is the setup with sidewards UCN extraction as used in the experiments described here. The shutter for the cold neutron beam is situated in the lower left corner of the picture.

Figure 9

(a) UCN leakage rate from the UCN storage volume to the detector with a slightly opened UCN shutter valve for different temperatures of the superfluid ${}^4\mathrm{He}$. The converter vessel is irradiated with cold neutrons for different durations with starting time set to zero. For clarity 4 neighboring data points have been averaged and recombined to one. (b) Saturated UCN leakage rate through the slightly opened shutter valve and the corresponding buildup time constant for different ${}^4\mathrm{He}$ temperatures. The latter are obtained from a fit to the exponential rise of the data points in (a). The line represents a fit to the data points of the buildup time (compare text).

Figure 10

(a) Time plots of the delayed extraction measurements at a ${}^4\mathrm{He}$ temperature of about 0.7 K. The plot depicts the detected UCN count rate as a function of time and corresponding single exponential fit curves (red lines). The converter vessel is irradiated with the cold neutron beam for a 300-s long UCN accumulation period. After the irradiation and a variable delay time the UCN shutter valve is opened and the UCN are extracted towards the detector. (b) The integral counts of the extracted UCN as a function of the delay time. The line shows the fit with a double exponential decay function. (c) Extraction time constant ${\tau }_{\mathrm{e}}$ as a function of the delay time, obtained from a single exponential decay fit to the data points in (a) during emptying of the storage volume.

Figure 11

Time plot of a delayed extraction measurement with zero delay time. Here, the UCN shutter valve is opened after approximately 800 s of irradiation with cold neutrons and accumulation of UCN. The total number of detected neutrons during the 125-s long extraction is approximately $2.74\times {10}^5$. The double exponential fit to the extraction data yields two time constants: $(13\pm 1)$ s and $(34\pm 1)$ s.

Figure 12

Repeated extraction of UCN during permanent irradiation of cold neutrons with a duty cycle of 10% (defined as fraction of time with UCN shutter valve open per accumulation-extraction cycle). In each emptying curve a total amount of approximately $1.4\times {10}^5$ UCN is extracted (shutter valve open for 10 s, counting during 20 s after start of shutter opening).