Comparison of ultracold neutron sources for fundamental physics measurements

Ultracold neutrons (UCNs) are key for precision studies of fundamental parameters of the neutron and in searches for new charge-parity-violating processes or exotic interactions beyond the Standard Model of particle physics. The most prominent example is the search for a permanent electric-dipole moment of the neutron (nEDM). We have performed an experimental comparison of the leading UCN sources currently operating. We have used a “standard” UCN storage bottle with a volume of 32 liters, comparable in size to nEDM experiments, which allows us to compare the UCN density available at a given beam port.

Figure 1

Side view of the UCN storage bottle with two different detector mounting options: (a) the horizontal extraction with straight UCN guide, (b) the vertical extraction with L-shaped UCN guide and the detector located 1 m lower. Visible from left to right are connection flange to source, shutter 1 (S-1), storage bottle, shutter 2 (S-2), connection flange to detector, UCN guide, Cascade detector (all dimensions in mm).

Figure 2

Sketch of the setup at SUN-2 (not to scale). Indicated are the position of the beam port (BP), the length, diameter, and materials of the connecting guides, and the direction of the UCNs. The storage vessel is sketched with shutters S-1 and S-2, storage volume SV, and detector D, as detailed in Fig. 1 for the horizontal UCN extraction setup.

Figure 3

Setup installed at the SUN-2 source with vertical extraction towards the detector. The source cryostat is visible on the left.

Figure 4

Time spectrum of UCN counts per 1 s, starting with the opening of the UCN valve (rising count rate) which defines time = 0. Shown are measured values for 600 s of accumulation, averaged over four UCN extractions with the vertical extraction system behind the external storage vessel (two with a storage time of 100 s and two with 200 s, respectively, each visible as a separate extraction peak). One observes a good saturation of the leakage rate during filling at 50 s and its decrease after closure of shutter 1 at the beginning of the storage time, i.e., at $t=60\mathrm{s}$. A fit to the leakage rate was performed within the 95 s long fit window indicated by the vertical green lines and extended to the duration of the whole period of storage and extraction (blue line).

Figure 5

UCN counts per measurement after a storage time of 5 s for various filling times of the storage bottle with 600 s accumulation time of SUN-2 and vertical extraction. The two data points at $t=50\mathrm{s}$ differ because of the decline of UCN source output with time and also part of the trend of the other data points is due to this effect.

Figure 6

UCN counts per measurement after various storage times for a SUN-2 UCN accumulation time of 600 s and vertical and horizontal UCN extractions to the detector. The measurement sequence was started with a storage time of 2 s and increased up to 200 s. Then one more measurement was done at 2 s and afterwards the storage times were decreased from 200 s back to 2 s. The differences observed for each pair of measurements are due to the gradual loss of intensity (see text). The results of the fits to the storage time constant measured via the shutter leakage rate (see Table 1) are indicated by the continuous lines. The dashed lines show the $1\sigma$ error bands. The data point at 200 s was out of the fit window.

Figure 7

Same as shown in Fig. 6 but for a SUN-2 UCN accumulation time of 300 s.

Figure 8

Sketch of the setup at PF-2 (not to scale): (a) at the height of the turbine exit and (b) on the EDM platform. Components as described in Fig. 2. In addition, the location of the safety foil after the beam port is indicated.

Figure 9

PF2 EDM platform: Storage bottle setup with the long UCN guiding section to the EDM platform at PF2. The blue UCN turbine is visible on the left side in the back.

Figure 10

UCN counts per measurement after 5 s of storage using various filling times measured at turbine exit height with vertical extraction and with the vacuum separation foil in place.

Figure 11

Same as in Fig. 10 but measured on top of the EDM platform with vertical extraction, without vacuum separation foil.

Figure 12

UCN counts per measurement after various storage times measured at turbine exit height. The results of the fits to the leakage rates are indicated by the continuous lines for the determined storage times of 27.6, 32.0, 32.1, and 41.8 s, respectively, see Table 3. The dashed lines show the $1\sigma$ error bands. The data point at 200 s was out of the fit window. Error bands are smaller at later times, as the UCN counts are then dominated by leakage. UCN count rates are higher for both measurements without aluminum safety foil than with foil.

Figure 13

Same as shown in Fig. 12, but measured on the EDM platform. The results of the fits to the leakage rates are indicated by the continuous lines for the determined storage times of 33.1, 44.8, 36.5, and 45.9 s, respectively, see Table 4. The dashed lines show the $1\sigma$ error bands. The data point at 200 s was out of the fit window. Error bands are smaller at later times, as the UCN counts are then dominated by leakage. Measurements with and without aluminum foil are interspersed.

Figure 14

Time spectrum of UCN counts in the detector measured with vertical extraction and no safety foil at turbine exit height. Sum of four measurements with 100 s storage time and one measurement with 200 s storage time. Starting from a very low UCN rate when the turbine's extraction tube is at a different beam line position, the rate of UCNs leaking through shutter 2 during filling rises and saturates after about 50 s. After shutter 1 is closed at $t=70\mathrm{s}$, the UCN rate drops for the storage time until the first emptying peak at $t=170\mathrm{s}$. A second emptying peak (200 s storage time) can be seen at $t=270\mathrm{s}$. Afterwards, the count rate stabilizes at the high leakage rate of about 2 kHz until the turbine extraction tube moves away at $t=350\mathrm{s}$. The fit to the leakage rate (blue line) of UCNs during these long storage measurements is indicated, the vertical green lines show the fit window.

Figure 15

Sketch of the setup at TRIGA Mainz (not to scale): Components are as described in Fig. 2. In addition, the location of the safety foil is indicated, which is about 1 m upstream from the beam port.

Figure 16

Photo of the storage setup with vertical extraction at beam tube D of the TRIGA Mainz reactor.

Figure 17

UCN counts per measurement after 2 s of storage for mounting positions of the storage bottle at various heights above the beam port. The filling times were optimized for each height. Due to space restrictions the setup could not be mounted at larger heights.

Figure 18

UCN counts per measurement after 5 s storage versus filling time measured with vertical extraction.

Figure 19

Same as Fig. 18 but measured with horizontal extraction.

Figure 20

Leaking UCN counts versus time after the reactor pulse measured with vertical extraction. Plotted is the sum of four measurements with 100 and 200 s of storage time, therefore two emptying peaks corresponding to these two times are visible. The fit window for the evaluation of the storage time constants is indicated with vertical green lines.

Figure 21

UCN counts per measurement after various storage times measured with filling times of 4 s (vertical extraction) and 3 s (horizontal extraction) at the TRIGA Mainz UCN source. The results of the fits to the leakage rates (see Table 7) are indicated by the continuous lines. The data point at 200 s was out of the fit window. The dashed lines show the $1\sigma$ error bands.

Figure 22

Sketch of the setup at PSI (not to scale): (a) at the height of the beam port and (b) at elevated height. Components are as described in Fig. 2. In addition, the location of the safety foil about 50 cm before the beam port is indicated.

Figure 23

PSI area West-1: Setup with additional stainless-steel guide section leading to the storage bottle mounted half a meter above the beam port with horizontal extraction to the detector.

Figure 24

UCN counts per measurement after 5 s of storage time for different filling times at beam height, at the PSI West-1 beam port, measured with vertical extraction. Statistical errors are below symbol size.

Figure 25

UCN counts per measurement after 5 s of storage time for different filling times at 500 mm above beam port West-1, measured with vertical extraction. Statistical errors are below symbol size.

Figure 26

UCN counts per measurement at beam height, after various storage times, measured with filling times of 26 s (at beam height). The results of the fits to the leakage rates (see Table 9) are indicated by the continuous lines. The data point at 200 s was out of the fit window. The dashed lines show the $1\sigma$ error bands.

Figure 27

Same as Fig. 26 but measured at 500 mm height above beam port with 24 s filling time. The data point at 200 s was out of the fit window. The large error band in the horizontal extraction measurement is caused by a strong correlation between the two storage time constants in the fit.

Figure 28

Time spectrum of UCN counts per 1 s in the detector, measured at a height of 500 mm, with vertical extraction. Sum of four measurements with 100 s and 200 s storage time, therefore two emptying peaks are visible. The double-exponential fit to the leakage rate of UCN during the storage time (see Table 9) is indicated (blue line), the vertical green lines show the fit window. $\text{Time}=0$ is given by the accelerator signal before the main beam pulse.

Figure 29

Measured largest UCN density in the standard storage bottle at a given UCN source plotted versus the measured storage time constant. The measurement conditions are explained in the text. The PF2 value is without safety foil, which is not a standard user configuration. Errors on the UCN density are smaller than symbol size.

Figure 30

Calculated ratio of measured UCN density in horizontal and vertical extraction, “h/v ratio,” versus UCN density measured in vertical extraction for the given source in a 2 s storage measurement. Labels are explained in the text. Errors are smaller than symbol size.

Figure 31

Measured largest UCN densities in the standard storage bottle after a storage time of 50 s (filled symbols) and 100 s (open symbols) at a given UCN source, plotted versus the measured storage time constant. The measurement conditions are explained in the text and we use the same measurements as shown in Fig. 29. Due to the leakage issues as discussed in the PF2 section a considerable leakage subtraction had to be applied to the counts shown in Fig. 12.

Figure 32

Simulated UCN energy spectra in the storage bottle after a storage time of 2 and 100 s. The dashed lines reflect the different UCN mean energies of 97 and 76 neV for the two energy distributions.

Figure 33

Simulation of storage measurements with horizontal and vertical extraction for the PSI source.

Figure 34

Simulated h/v ratios at different storage times for the PSI source. The lower values at later times reflect a larger low-energy neutron fraction at later storage times, when faster UCNs have already been lost.