Introduction

Neutrons, as non-ionizing particles are detectable indirectly by secondary processes. As active detectors the most popular are proportional counters or ionizing chambers modified using thermal neutron sensitive materials like 3He, 6Li, 10B or 235U. Thus, the detection occurs due to specific nuclear reactions, which take place inside counter:

  • Helium counter: \({}^{1} n + {}^{3} {\text{He}} \to {}^{3} {\text{H}} + {}^{1} {\text{H}}\)

  • Lithium counter: \({}^{1} n + {}^{6} {\text{Li}} \to {}^{3} {\text{He}} + {}^{4} {\text{He}}\)

  • BF3 or B lined counter: \({}^{1} n + {}^{10} {\text{B}} \to {}^{4} {\text{He}} + {}^{7} {\text{Li}}\)

  • Uranium counter: \({}^{1} n + {}^{235} {\text{U}} \to fission\)

The second class of active detectors is based on elastic scattering (proton recoil), which is sensitive to fast neutrons. They are filled with hydrogen or hydrogen rich material.

Other kinds of active detectors are modified scintillators (anthracene or other organic) sensitive to fast neutrons, and semiconductors like silicone detectors with LiF converters.

Next to active methods passive ones are widely used. They are based on neutron activation of different materials followed by gamma-ray spectrometry or they are dose meters as for instance thermoluminescence detectors (TLD) using lithium fluoride pellets (enriched with 6Li).

Eventually, since the neutron background at sea level is about 0.015 n/(cm2s) [1] also neutron induced effects like inelastic neutron scattering with gamma-ray emission or radiative capture processes are observed in low-background gamma-ray spectrometry. In such processes gamma-ray quanta are emitted. Therefore, if these effects exists at reasonable intensities [2,3,4,5,6], they can be used for neutron detection. In Fig. 1 is presented an example of a low-background gamma-ray spectrum in which “neutron effects” are observed. The spectrum was collected with a low background gamma-ray spectrometer with active shield [7]. Neutron effects can be enhanced introducing materials with high cross sections for neutron induced gamma-ray emission. Even pure water can be used for this purpose (Fig. 2), due to the emission of 2223 keV line from neutron capture by hydrogen [3, 7]. There are, however, more efficient materials. One of these is cadmium [7], due to high cross section for thermal neutrons capture with emission of several gamma-ray lines, among them the 558.4 keV line (Fig. 3). The 558.4 keV gamma-ray line comes from the de-excitation of the first excited level of a stable 114Cd [8]. It is usually observed in decays of 114In or 114Ag, however, this level can be populated also in neutron induced processes like 114Cd(n,n′)114Cd, 113Cd(n,γ)114Cd or 113Cd(n,p)114Ag, followed by beta decay of the latter (Fig. 4).

Fig. 1
figure 1

Example of gamma-ray spectrum registered during more than a month (3.3 Ms) of measurement in shielded low background HPGe detector with composite shield (15 cm of lead, Cu lining and 70 × 70 cm multiwire Charpak chamber as active veto) [7]. The spectral lines produced in inelastic scattering of neutrons on germanium and other neutron effects are visible [2]

Full size image
Fig. 2
figure 2

The 2223 keV gamma line of neutron capture by hydrogen observed for water sample (0.5 L in Marinelli beaker) in low background gamma spectrometer with active shield [7] during 11 days lasting measurement

Full size image
Fig. 3
figure 3

The gamma line of neutron capture by cadmium observed for Marinelli beaker with cadmium inner lining filled with paraffin with water inside shield in role of additional moderator registered in low level gamma spectrometer equipped with active shield [7]

Full size image
Fig. 4
figure 4

The illustration for emission of 558 keV gamma line by exited 114Cd stable nuclei in chosen processes induced by neutrons [4, 13]

Full size image

Methods

The general idea of a modifying low background germanium spectrometer to enhance the signal from environmental background neutrons was proposed and tested many years ago [7]. The principle of modification is simple: some cadmium and moderator are placed inside the low background shield. In current, modified version a 2 mm thick cadmium foil (30 × 10 cm) was wrapped around (Fig. 5). Another 2 mm thick cadmium foil (10 × 10 cm) was placed on a top of end-cap. The HPGe detector was made [9] at the home Institute of Nuclear Physics (INP) workshop using a germanium crystal (purchased from UMICORE, Belgium, of ~ 50 mm diameter and ~ 50 mm height). The cryostat of the detector is made of ultrapure aluminum by Baltic Scientific Instruments (BSI), having U and Th impurities levels at 1 × 10−9 g/g range [9] and equipped with composite carbon fibre window. The detector was in standard lead shield of 10 cm thick bricks with 50 × 50 × 50 cm shielded space (mass 2.2 t, Fig. 6). Besides cadmium a moderator was placed inside the shield. Various configurations of moderator were tested. The final one consists of 60 L of water put in 12 plastic canisters of 5 L volume each (Fig. 5). In this configuration background spectra were collected in automatic mode, recording separately every hour a spectrum (namely 3590 s of counting and 10 s break). The acquisition system consists of standard NIM electronics including NIM crate (by FAST, Germany), an amplifier, model 2020 and HV power supply, model 3106 (Canberra, USA) and MCA Tukan-USB (by NCNR Świerk, Poland). Measurements were performed for 50.5 days from May to July 2014. Then spectra were summed together for each hour of day, reducing number of spectra from 1212 to 24. The aim was to check if there is any day cycle (day-night effect) of count rates in main background lines of spectrometer. Besides it, a total spectrum (with acquisition time of ~ 4.35 Ms) from all single hour spectra was created as well. All the summing was performed by a code in Fortran. Another set of measurements were conducted with a different moderator inside shield. Namely, there were five graphite cylinders (20 cm diameter, 32 cm height, about 10 L of volume each) with total mass of ~ 108 kg. They were placed inside the shield (Fig. 5) around the detector as moderator, four were standing in the corners of the shield and one was lying on them covering detector from top.

Fig. 5
figure 5

Details of arrangements of passive elements inside of gamma ray spectrometer shield: cadmium plate wrapped around detector and covering it from top, setup of twelve plastic containers filled with water around detector in two layers, setup of five graphite cylinders

Full size image
Fig. 6
figure 6

The general view of detector and lead shield

Full size image

Results and discussion

The results for the total-50-days gamma-ray spectrum are shown in Table 1 and in Fig. 7. All presented uncertainties are just counting statistics ones. The gamma-ray lines in the spectrum, presented in Table 1, were identified based on data given in literature [10,11,12]. Please note the absence of relatively intense lines from thorium series like 239 keV of 212Pb, 338 keV of 228Ac- or 583 keV of 208Tl, also the absence of some from uranium series like 1001 keV of 234mPa, present in typical low background gamma-ray spectra [12]. They were also not present in measurements performed underground (at depth of ~ 1 km inside a gallery within salt deposit of Sieroszowice copper mine) with the same detector [9]. This is the result of low level of Th and U traces in BSI low cryostat of detector. This confirms, that the high energy (2614 keV) line of 208Tl comes from outside of the shield. Apparently more difficult was removing traces of radon (222Rn) daughters from background, since the shield was not tight enough to prevent their diffusive penetration into shield, however that can be easily improved by introducing of LN2 vapors flush into the inner space of shield.

Table 1 Spectral gamma ray lines present in the combined spectrum collected during 4,351,080 s (~ 50.5 days) of the background investigation of the spectrometer
Full size table
Fig. 7
figure 7

Summed gamma ray spectrum for 50.5 days of collection with 60 L water moderator and cadmium elements localized inside shield (see text for details). The energy (in keV) of main visible gamma lines are depicted

Full size image

The results obtained for single hours in day are shown in the Figs. 8, 9 and 10. Figure 8 shows a diurnal changes in numbers count in the background for the main spectral lines from radon (222Rn) progeny: two lines of 214Pb (295 and 352 keV) and three lines of 214Bi (609, 1120 and 1764 keV). The drawings present also the average values obtained for the summary spectrum (as presented in Table 1). Shown are also residual count rates. One cannot observe any systematic changes of daily average values-of the intensity of all the lines since they show only random scatter around average values. This means that that no difference between working hours and rest of the day was noticed.

Fig. 8
figure 8

The hour-a-day summed spectra results for main radon daughter gamma lines. Each spectrum as effectively collected for about 180,000 s. Presented are counts rates (top) and their variation as rate residue (bottom)

Full size image
Fig. 9
figure 9

The hour-a-day summed spectra results for high energy gamma lines of terrestrial radionuclides like 40K and 208Tl. Each spectrum as effectively collected for ~ 180,000 s. Presented are counts rates (top) and their variation as rate residue (bottom)

Full size image
Fig. 10
figure 10

The hour-a-day summed spectra results for gamma lines occurring mainly due to interaction with secondary cosmic particles like muons or neutrons in modified gamma spectrometer. Each spectrum as effectively collected for about 180,000 s. Presented are counts rates (top) and their variation as rate residue (bottom)

Full size image

Figure 9 shows the daily changes in number of counts in the background for the main spectral lines from radionuclides permanently present in the surroundings of the spectrometer emitting photons at high energies, for which 10 cm lead shield is likely not sufficiently thick to eliminate complete external radiation: 40K (1461 keV) and 208Tl (2615 keV). The drawings present also the average values obtained for the summary spectrum (as presented in Table 1). Again, one cannot observe any systematic changes of daily average values-of the intensity of all the lines since they show only random scatter are around average values, but in this case this was an expected result. Also the residual count rates confirms statistical scattered pattern only.

Figure 10 shows a diurnal changes in number of counts in the background for the main gamma-ray spectral lines of photons emitted as a result of interactions with particles of the secondary cosmic rays, such as muons and neutrons. It should be noted, that the cosmic radiation is perhaps not the sole source of positrons and neutrons, however, it can be expected that this is the dominant source. The peak 511 keV comes from the annihilation of positrons with electrons. Positrons are produced in electromagnetic avalanches arising from interactions of cosmic muons with all surrounding matter (air, shield materials, detector itself), but they may also come from beta plus emitters present in traces (as activation products) in construction material of the detector and shield. Moreover they comes also from pair production processes from any high energy gamma quanta. Neutrons, seen in the background on Earth, besides cosmogenic origin are also produced for example in the nuclear reactions like (α, n) but neutron flux from these processes is about three orders of magnitude lower than that from cosmic rays at sea level. As in the previous two Figs. 8 and 9 are presented average values obtained from the total spectrum (as given in Table 1) together with residual count rates for different energies. For 511 keV like in the previous viewgraphs, one cannot see any systematic changes daily average values. However it is a differently for the less intense gamma-ray line of 2223 keV which is from deuterium formation. Here, the residual count rates suggest the existence of lower production rate in night and higher during the day time. Results for 558.4 keV peak do not show such variation, but also do not exclude possibility of existing of similar effect in future experiment done with better statistics.

The intensity of all the lines are around average values. The difficulty in observation of clear effect on counts rates in hour-a-day spectra could be the result of averaging of all effects by our procedure. It is known, that the neutron flux and well as the muon flux are functions of factors like humidity, air pressure or temperature [14, 15].

Results of comparison of water (~ 60 L) and graphite (~ 108 kg) moderators are shown in Table 2. The results for graphite were obtained in a single measurement lasting about 10 days (t = 825,008) in beginning of 2019, so 4 years later than the main experiment. One can notice that count rates in 511 keV is different by 34% in (higher for graphite), what can be explained by presence in graphite of traces of natural gamma-ray emitters, what is namely the case of Th resulting in additional counts in 511 keV peak coming from 510.8 keV gamma-ray line of 208Tl. The observed count rate for 558.4 keV gamma-ray line from de-excitation of 114Cd line diminished for graphite to 84% of that with water moderator. However the most important are the mentioned above traces of natural radioactivity isotopes in graphite, which increase the count rates in the main background lines by a factor ~ 2.5 (U series) or ~ 4 (40K) as well as added no detectable earlier peaks from traces of Th series.

Table 2 Comparison of intensity of chosen gamma-lines in spectrum for different moderator placed inside the spectrometer’s lead shield: 60 L of water (“H2O [CPS]”) or 108 kg of graphite (“C [CPS]”)
Full size table

Conclusions

Although the diurnal variation of count rate for 558.4 keV cadmium gamma-ray peak was not observed a kind of such effect is suggested for 2223 keV deuterium production gamma-ray peak. All other analyzed lined revealed only statistical scattering pattern. Proposed simple modification of gamma spectrometer is an useful tool for monitoring of neutron flux at sea-level together with the cosmic muons flux (the main component in 511 keV). Moreover, the changes of mass and geometry of the moderator could be used to the study of composition of neutron spectra, which was not done here. Any possible application, in which there is a need of parallel measurements of prompt gamma-rays emitters and neutron flux, could make use of the proposed method or its variations. The future development of our setup and method are planned to explore its properties more deeply. Further modification of hardware is needed to create optimal conditions. Computer simulations are planned are planned to get absolute calibration of neutron flux. One idea is use of digital output, which would enable to sort data in any desired way.