- Email: [email protected]
unshielded uranium by a 4 MeV linac
Applied Radiation and Isotopes 69 (2011) 1251–1254
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Neutron interrogation of shielded/unshielded uranium by a 4 MeV linac L. Lakosi n, C. Tam Nguyen, E. Serf Institute of Isotopes, Hungarian Academy of Sciences, PO Box 77, H-1525 Budapest, Hungary
a r t i c l e i n f o
abstract
Article history: Received 17 December 2010 Accepted 4 April 2011 Available online 13 April 2011
A non-destructive active assay method was developed for revealing illicit trafficking of uranium. Photoneutrons produced in beryllium or heavy water by bremsstrahlung from a linac induced fission in the samples. Delayed fission neutrons were detected by a neutron collar built up of 3He counters embedded in polyethylene moderator. High-enriched uranium samples shielded and unshielded by lead up to 14 mm thickness were detected, with a performance practically unaltered. 25 mg 235U can be revealed in a 1 min interrogation time. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Non-destructive assay Illicit trafficking Be/D2O converter Delayed neutron counting
1. Introduction As is well-known, active methods (‘‘interrogation’’) are widely used for revealing illicit nuclear materials (NM) by irradiation and detection of the induced response. Neutrons can penetrate high-Z shielding material, induce fission in the NM, and fission neutrons can more effectively be detected than passive g-rays. The active method does not suffer from difficulties due to shielding encountered in passive g-spectroscopy. This involves an external source that irradiates the material whether or not it is shielded, and the subsequently emitted radiation is detected and used to characterize the material being present. In order to accomplish such an analysis, active radiation portal monitors are to be deployed at customs and border checkpoints. If the induced radiation comprises neutrons, they may penetrate the shielding as well. In such cases it may be advisable not to take the material out of its holder during the irradiation and counting. Fission neutrons can be distinguished from irradiating ones via utilizing the time correlation among the former (coincidence counting), or via separating delayed neutrons from the primary and prompt fission ones by their time sequence. In the fields of safeguards and nuclear forensics, pulsed D-T neutron generators represent a sensitive and versatile variant of active interrogation systems, by counting delayed fission neutrons or gammas (see references in Lakosi et al., 2008; Lakosi and Nguyen, 2008). Larger systems designed for inspecting sea cargo relying on D-D (Hall et al., 2007; Slaughter et al., 2007; Church et al., 2007) or photoneutron interrogation by linacs are also common as pulsed neutron sources (references in Lakosi et al.,
n
Corresponding author. E-mail addresses: [email protected], [email protected] (L. Lakosi).
0969-8043/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2011.04.003
2008; Lakosi and Nguyen, 2008). Assay of hidden or shielded NM via delayed neutrons and gammas from photo- and neutroninduced fissions, by the use of pulsed 10–20 MeV linacs, has also been reported (Jones et al., 2005; Norman et al., 2005; Kinlaw and Hunt, 2006; Jones et al., 2006,2007; Norman et al., 2007; Sterbentz et al., 2007). Uranium mass of samples embedded in large concrete packages was assessed by photon interrogation and photofission, using 15–30 MeV linac (Gmar et al., 2005). Shielded HEU was interrogated with a 60 keV neutron beam produced by a 2 MeV proton linac (Kerr et al., 2007). A photoneutron interrogation project has been carried out by applying 4 MeV electron linac of the Institute of Isotopes as a neutron source, to induce fission in low- (LEU) (Lakosi et al., 2008) and high-enriched uranium (HEU) samples (Lakosi and Nguyen, 2008). The electron energy has been converted into bremsstrahlung by a platinum foil, whereas neutrons for interrogation have been produced either in heavy water or beryllium. Delayed neutrons produced in the fissile material have been detected, distinguished from interrogating neutrons by using time discrimination. Having performed the assay of bare uranium previously, results of the interrogation of HEU samples behind lead shielding are presently reported. Typically about 1% of fission neutrons only are delayed, emitted by fission products. It is necessary therefore to count delayed neutrons for long enough time, in order to achieve good statistics. It means that pulsing neutron sources are necessary, and delayed neutrons can be counted in the interval between pulses. As a result of the continuously pulsing irradiation, the intensity of delayed neutrons goes into saturation with an amplitude depending on the pulse repetition rate. By halving the rate, saturation intensity halves as well, in parallel to the mean intensity of the electron current. It has been shown (Lakosi and Nguyen, 2008) that 20 s irradiation allows a saturation degree
1252
L. Lakosi et al. / Applied Radiation and Isotopes 69 (2011) 1251–1254
of about 80%, therefore initial irradiations of 20 s at least were applied before starting irradiation-measurement cycles.
2. Experimental Electron pulses, duration of 2.6 ms, have been fired with a repetition rate of 25 Hz. Selected from 50, 25, 12.5, or 6.25 Hz, this was established to be the optimum setting in previous experiments (Lakosi et al., 2008; Lakosi and Nguyen, 2008). Bremsstrahlung was generated on a 20mm 30 mm size, 0.9 mm thick platinum converter positioned at 3 cm distance from the exit window of the linac. The diameter of the electron beam was about 2 cm at converter distance. Neutron production was due to (e,gamma) and (gamma,n) double conversion. Beryllium and heavy water was applied alternatively as photoneutron converter. Their (gamma,n) reaction thresholds are 1.67 and 2.23 MeV, respectively. The neutron energy available from the 9Be(gamma,n)8Be and D(gamma,n)H reaction is up to 2.33 and 1.77 MeV, respectively, at 4 MeV endpoint energy bremsstrahlung. However, the yields abruptly vanish above around 0.9 MeV neutron energy, whereas maximum intensity of the spectrum of evaporated
1.E+05 1.E+04 1.E+03
Time constant: ~ 2 ms
1.E+02 1.E+01 1.E+00
0
5
10
15
20
25
30
35
40
Time [ms] Fig. 1. Time spectrum of the D2O converter alone, without U sample. 25 Hz, 1.9 mA, 1000 cycles.
neutrons is at about 0.5 MeV, even at much higher (15–50 MeV) linac energies (Sa´fa´r and Lakosi, 1994, Facure et al., 2005). A home-made neutron collar, as reported earlier (Lakosi et al., 2006; Lakosi and Nguyen, 2008), consisted of concentric polyethylene cylinders of an outer size of 300 mm diameter and 470 mm total length. The innermost ring (of 20 mm wall thickness) forms a measurement cavity for the material to be assayed. In between the outermost and inner cylinders, 12 proportional counters (type SNM-28) of diameter 32mm 308 mm length each, filled with 3He gas to a pressure of 4.0E5 Pa (4 atm) served as neutron detectors. Based on previous experiences, Cd foil was not used. Heavy water or beryllium photoneutron converters of a mass of 105 and 170 g, respectively, were applied at the top of the collar. The signal processing electronics, consisting of a 400 channel analyzer in multiscaler mode of operation as a time analyzer, was almost the same as reported previously (Lakosi and Nguyen, 2008). The multiscaler receives commands from a PC through a micro-controller. Triggering the analyzer was synchronized with the linac control command pulse. A channel width (dwell time) of 100 ms was selected for the present measurements. Irradiation-measurement cycles of up to 1000 were carried out at 25 Hz. As experienced previously (Lakosi et al., 2008; Lakosi and Nguyen, 2008), time spectra acquired at 25 Hz show that the pulse length of prompt (interrogating and fission) neutrons cover a half of the 40 ms interval between two pulses. Exponential decay with a time constant of about 2 ms was observed, with a total pulse length of 20 ms. Thus, the effective time remaining for delayed neutron measurement is about 20 ms, no matter if using heavy water or beryllium converter, as observed previously. An irradiation-measurement cycle lasts for 40 ms at 25 Hz. Thus, 1000 cycles last for 40 s. In Fig. 1 a time spectrum of the D2O converter alone is displayed, taken during 1000 cycles at a mean electron current of 1.9 mA, while no U sample was in the measurement cavity. It was practically the same using the beryllium converter as well. Time spectra of HEU oxide powder samples of 10.5, 5.5, and 2.3 g mass of 36% enrichment, as well as of a 0.53 g sample of 90% enrichment are seen in Fig. 2. All the spectra were taken by the D2O converter, during 1000 irradiation-measurement cycles. The mean electron current was uniformly 1.9 mA at 25 Hz. In order to reach a sufficient degree of saturation, 20 s irradiations were
1.E+05
1.E+05
1.E+04
1.E+04 10.5g (36% enr.)
1.E+03
MEASUREMENT
1.E+02
1.E+02
1.E+01
1.E+01
1.E+00
1.E+00
1.E+05
1.E+05 1.E+04
1.E+04 2.33g (36% enr.)
1.E+03
1.E+02
1.E+01
1.E+01 0
10
20 Time [ms]
30
0.53g (90% enr.)
1.E+03
1.E+02
1.E+00
5.55g (36% enr.)
1.E+03
40
1.E+00
0
10
20
30
40
Time [ms]
Fig. 2. Time spectra of 10.5, 5.5, 2.3 g (36% enr.), and 0.53 g UO2 (90% enr.) samples without and with Pb holders.
L. Lakosi et al. / Applied Radiation and Isotopes 69 (2011) 1251–1254
1.5x104
1.0x104
5.0x103
0.0
3. Results and discussion The number of counts was summed up in the time interval for counting delayed neutrons, i.e. from channel number 200 to 400, corresponding to an effective measurement time of 20 s during 1000 cycles. Total interrogation time thus corresponds to 1 min, including an additional 20 s initial irradiation and 40 s counting time, out of which 20 s is effective. The results obtained by using heavy water converter are seen in Table 1. Indicated are the sample masses, U-235 contents, number of counts, the count rates related to the effective measurement time, and the ratios of the results for shielded to those for unshielded samples. It is seen that the results of the measurements of shielded samples agree well with results obtained without shielding. Deviations are within 710% in the most cases. They can be attributed to the limited reproducibility of the position of the converters and samples, and of the pulse amplitudes, because the linac was started again at every interrogation run (of 1000 cycles). The statistical uncertainty is 1–2% only, negligible. In Fig. 3 the data as a function of 235U mass are plotted, because the samples were not of equal isotopic composition. The signal (number of counts) is not fully linear. This issue was examined in more detail previously for LEU samples of different enrichment (Lakosi et al., 2008), and a good linear response was found up to 10g 235U mass. In that case the mass of samples was uniformly 400 g, the (low) enrichment of samples was only varied. In the present case, however, sample masses are different. Deviation from the linearity may be due to the flux depression of thermal neutrons in the interior of the bigger (and of higher enrichment) samples. The method is sensitive to the 238U content as well, because of the imperfect thermalization and fast fission, although 50–60 times weaker than to 235U (Lakosi et al., 2008). This effect is, therefore, vanishing compared with the former reason. Table 1 Results for unshielded and shielded UO2 samples by using heavy water converter at a mean electron current of 1.9 mA and 1000 cycles. Sample mass (g)
U-235 mass (g)
Leadholder
No. of counts
Eff. count rate (cps)
Ratio shielded/ unshielded
10.49 (Enr. 36%)
3.20
No Smaller Bigger
1730 17,979 19,500
893 896 972
1.003 1.088
5.55 (Enr. 36%)
1.69
No Smaller Bigger
12,611 12,086 13,300
628 602 663
0.958 1.055
2.33 (Enr. 36%)
0.71
No Smaller Bigger
6836 6423 5819
341 320 290
0.940 0.851
0.53 (Enr. 90%)
0.40
No Smaller Bigger
2317 2039 2251
115 102 112
0.880 0.972
Sample in thin Pb-container Sample in thick Pb-container Sample without Pb-container
2.0x104
Counts
uniformly carried out before starting cyclic delayed neutron measurements. The samples were interrogated both unshielded and shielded, being inserted in two lead holders of outer diameters 44 and 63 mm by heights 87 and 106 mm, respectively. Holder mass was 0.60 and 2.70 kg, with wall thickness of 4 and 14 mm, respectively. So each diagram contains a time spectrum of the sample unshielded and two spectra of the sample being placed in the lead holders. It can be seen that there are no significant differences between the spectra of the samples being shielded or not, so the actual interrogation of an unknown material can well be performed with the material left in its holder (up to 14 mm lead shielding at least).
1253
0
1
2 235
3
U content, g
Fig. 3. Results without and with Pb holders.
Results for the two converters are very similar. Assuming a detection limit of 6.8 count/s (cps) corresponding to twice the background level at 1000 cycles, this means 136 counts during the effective measurement time (20 s). Taking into account a sensitivity of about 5500 counts/g 235U content (Table 1, Fig. 3), this corresponds to a lower detection limit of about 25 mg 235U.
4. Summary No influence of a shielding container was observed (up to 14 mm lead thickness at least). Delayed neutron signal is in a good approximation linearly related to the 235U content. Minor deviation from linearity may be due to the flux depression of thermal neutrons in the interior of bigger samples. A sensitivity of about 5500 counts/g 235U was achieved. This corresponds to a lower detection limit of 25 mg 235U at a mean electron current intensity of 1.9 mA in a 1 min total interrogation time (1000 cycles). By increasing the electron current or interrogation time (number of cycles), the response can be enhanced. Half of the 40 ms time interval between pulses can only be exploited, i.e. 20 ms effective time is available for counting at 25 Hz frequency. Summing up, an efficient laboratory method and an equipment have been developed for a quantitative assay of unknown U-containing material, without opening its holder. Experience gathered and results may form a basic knowledge for designing a larger system of an active radiation portal monitor.
Acknowledgments Thanks are due to Mr. P. Hargittai for operating the linac. This work was supported by the Hungarian Atomic Energy Authority. References Church, J.A., Slaughter, D.R., Asztalos, S., Biltoft, P., Descalle, M.-A., Hall, J., Luu, T., Manatt, D., Mauger, J., Norman, E.B., Petersen, D., Prussin, S., 2007. Signals and interferences in the nuclear car wash. Nucl. Instr. Meth. B 261, 351–355. Facure, A., Falcao, R.C., Silva, A.X., Crispim, V.R., Vitorelli, J.C., 2005. A study of neutron spectra from medical linear accelerators. Appl. Radiat. Isot. 62, 69–72. Gmar, M., Jeanneau, F., Laine´, F., Makil, H., Poumarede, B., Tola, F., 2005. Assessment of actinide mass embedded in large concrete waste packages by photon interrogation and photofission. Appl. Radiat. Isot. 63, 613–619.
1254
L. Lakosi et al. / Applied Radiation and Isotopes 69 (2011) 1251–1254
Hall, J.M., Asztalos, S., Biltoft, P., Church, J., Descalle, M.-A., Luu, T., Manatt, D., Mauger, G., Norman, E., Petersen, D., Pruet, J., Prussin, S., Slaughter, D., 2007. The nuclear car wash: neutron interrogation of cargo containers to detect hidden SNM. Nucl. Instr. Meth. B 261, 337–340. Jones, J.L., Yoon, W.Y., Norman, D.R., Haskell, K.J., Zabriskie, J.M., Watson, S.M., Sterbentz, J.W., 2005. Photonuclear–based nuclear material detection system for cargo containers. Nucl. Instr. Meth. B 241, 770–776. Jones, J.L., Norman, D.R., Haskell, K.J., Sterbentz, J.W., Yoon, W.Y., Watson, S.M., Johnson, J.T., Zabriskie, J.M., Bennett, B.D., Watson, R.W., Moss, C.E., Harmon, J.F., 2006. Detection of shielded nuclear material in a cargo container. Nucl. Instr. Meth. A 562, 1085–1088. Jones, J.L., Blackburn, B.W., Watson, S.M., Norman, D.R., Hunt, A.W., 2007. Highenergy photon interrogation for nonproliferation applications. Nucl. Instr. Meth. B 261, 326–330. Kerr, P., Rowland, M., Dietrich, D., Stoeffl, W., Wheeler, B., Nakae, L., Howard, D., Hagmann, C., Newby, J., Porter, R., 2007. Active detection of small quantities of shielded highly-enriched uranium using low-dose 60 keV neutron interrogation. Nucl. Instr. Meth. B 261, 347–350. Kinlaw, M.T., Hunt, A.W., 2006. Fissionable isotope identification using the time dependence of delayed neutron emission. Nucl. Instr. Meth. A 562, 1081–1084.
Lakosi, L., Nguyen, C.T., 2008. Neutron interrogation of high-enriched uranium by a 4 MeV linac. Nucl. Instr. Meth. B 266, 3295–3301. Lakosi, L., Bagi, J., Nguyen, C.T., 2006. Quantitative non-destructive assay of PuBe neutron sources. Nucl. Instr. Meth. B 243, 385–391. Lakosi, L., Nguyen, C.T., Bagi, J., 2008. Photoneutron interrogation of low-enriched uranium induced by bremsstrahlung from a 4 MeV linac. Nucl. Instr. Meth. B 266, 295–300. Norman, D.R., Jones, J.L., Yoon, W.Y., Haskell, K.J., Sterbentz, J.W., Zabriskie, J.M., Hunt, A.W., Harmon, F., Kinlaw, M.T., 2005. Inspection applications with higher electron beam energies. Nucl. Instr. Meth. B 241, 787–792. Norman, D.R., Jones, J.L., Blackburn, B.W., Haskell, K.J., Johnson, J.T., Watson, S.M., Hunt, A.W., Spaulding, R., Harmon, F., 2007. Time-dependent delayed signatures from energetic photon interrogations. Nucl. Instr. Meth. B 261, 316–320. Sa´fa´r, J., Lakosi, L., 1994. Photoneutron production by bremsstrahlung in the target and the converter. Appl. Radiat. Isot. 45, 1057–1059. Slaughter, D.R., et al., 2007. The nuclear car wash: a system to detect nuclear weapons in commercial cargo shipments. Nucl. Instr. Meth. A 579, 349–352. Sterbentz, J.W., Jones, J.L., Yoon, W.Y., Norman, D.R., Haskell, K.J., 2007. Benchmark validation comparisons of measured and calculated delayed neutron detector responses for a pulsed photonuclear assessment technique. Nucl. Instr. Meth. B 261, 373–377.
Contents lists available at ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Neutron interrogation of shielded/unshielded uranium by a 4 MeV linac L. Lakosi n, C. Tam Nguyen, E. Serf Institute of Isotopes, Hungarian Academy of Sciences, PO Box 77, H-1525 Budapest, Hungary
a r t i c l e i n f o
abstract
Article history: Received 17 December 2010 Accepted 4 April 2011 Available online 13 April 2011
A non-destructive active assay method was developed for revealing illicit trafficking of uranium. Photoneutrons produced in beryllium or heavy water by bremsstrahlung from a linac induced fission in the samples. Delayed fission neutrons were detected by a neutron collar built up of 3He counters embedded in polyethylene moderator. High-enriched uranium samples shielded and unshielded by lead up to 14 mm thickness were detected, with a performance practically unaltered. 25 mg 235U can be revealed in a 1 min interrogation time. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Non-destructive assay Illicit trafficking Be/D2O converter Delayed neutron counting
1. Introduction As is well-known, active methods (‘‘interrogation’’) are widely used for revealing illicit nuclear materials (NM) by irradiation and detection of the induced response. Neutrons can penetrate high-Z shielding material, induce fission in the NM, and fission neutrons can more effectively be detected than passive g-rays. The active method does not suffer from difficulties due to shielding encountered in passive g-spectroscopy. This involves an external source that irradiates the material whether or not it is shielded, and the subsequently emitted radiation is detected and used to characterize the material being present. In order to accomplish such an analysis, active radiation portal monitors are to be deployed at customs and border checkpoints. If the induced radiation comprises neutrons, they may penetrate the shielding as well. In such cases it may be advisable not to take the material out of its holder during the irradiation and counting. Fission neutrons can be distinguished from irradiating ones via utilizing the time correlation among the former (coincidence counting), or via separating delayed neutrons from the primary and prompt fission ones by their time sequence. In the fields of safeguards and nuclear forensics, pulsed D-T neutron generators represent a sensitive and versatile variant of active interrogation systems, by counting delayed fission neutrons or gammas (see references in Lakosi et al., 2008; Lakosi and Nguyen, 2008). Larger systems designed for inspecting sea cargo relying on D-D (Hall et al., 2007; Slaughter et al., 2007; Church et al., 2007) or photoneutron interrogation by linacs are also common as pulsed neutron sources (references in Lakosi et al.,
n
Corresponding author. E-mail addresses: [email protected], [email protected] (L. Lakosi).
0969-8043/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2011.04.003
2008; Lakosi and Nguyen, 2008). Assay of hidden or shielded NM via delayed neutrons and gammas from photo- and neutroninduced fissions, by the use of pulsed 10–20 MeV linacs, has also been reported (Jones et al., 2005; Norman et al., 2005; Kinlaw and Hunt, 2006; Jones et al., 2006,2007; Norman et al., 2007; Sterbentz et al., 2007). Uranium mass of samples embedded in large concrete packages was assessed by photon interrogation and photofission, using 15–30 MeV linac (Gmar et al., 2005). Shielded HEU was interrogated with a 60 keV neutron beam produced by a 2 MeV proton linac (Kerr et al., 2007). A photoneutron interrogation project has been carried out by applying 4 MeV electron linac of the Institute of Isotopes as a neutron source, to induce fission in low- (LEU) (Lakosi et al., 2008) and high-enriched uranium (HEU) samples (Lakosi and Nguyen, 2008). The electron energy has been converted into bremsstrahlung by a platinum foil, whereas neutrons for interrogation have been produced either in heavy water or beryllium. Delayed neutrons produced in the fissile material have been detected, distinguished from interrogating neutrons by using time discrimination. Having performed the assay of bare uranium previously, results of the interrogation of HEU samples behind lead shielding are presently reported. Typically about 1% of fission neutrons only are delayed, emitted by fission products. It is necessary therefore to count delayed neutrons for long enough time, in order to achieve good statistics. It means that pulsing neutron sources are necessary, and delayed neutrons can be counted in the interval between pulses. As a result of the continuously pulsing irradiation, the intensity of delayed neutrons goes into saturation with an amplitude depending on the pulse repetition rate. By halving the rate, saturation intensity halves as well, in parallel to the mean intensity of the electron current. It has been shown (Lakosi and Nguyen, 2008) that 20 s irradiation allows a saturation degree
1252
L. Lakosi et al. / Applied Radiation and Isotopes 69 (2011) 1251–1254
of about 80%, therefore initial irradiations of 20 s at least were applied before starting irradiation-measurement cycles.
2. Experimental Electron pulses, duration of 2.6 ms, have been fired with a repetition rate of 25 Hz. Selected from 50, 25, 12.5, or 6.25 Hz, this was established to be the optimum setting in previous experiments (Lakosi et al., 2008; Lakosi and Nguyen, 2008). Bremsstrahlung was generated on a 20mm 30 mm size, 0.9 mm thick platinum converter positioned at 3 cm distance from the exit window of the linac. The diameter of the electron beam was about 2 cm at converter distance. Neutron production was due to (e,gamma) and (gamma,n) double conversion. Beryllium and heavy water was applied alternatively as photoneutron converter. Their (gamma,n) reaction thresholds are 1.67 and 2.23 MeV, respectively. The neutron energy available from the 9Be(gamma,n)8Be and D(gamma,n)H reaction is up to 2.33 and 1.77 MeV, respectively, at 4 MeV endpoint energy bremsstrahlung. However, the yields abruptly vanish above around 0.9 MeV neutron energy, whereas maximum intensity of the spectrum of evaporated
1.E+05 1.E+04 1.E+03
Time constant: ~ 2 ms
1.E+02 1.E+01 1.E+00
0
5
10
15
20
25
30
35
40
Time [ms] Fig. 1. Time spectrum of the D2O converter alone, without U sample. 25 Hz, 1.9 mA, 1000 cycles.
neutrons is at about 0.5 MeV, even at much higher (15–50 MeV) linac energies (Sa´fa´r and Lakosi, 1994, Facure et al., 2005). A home-made neutron collar, as reported earlier (Lakosi et al., 2006; Lakosi and Nguyen, 2008), consisted of concentric polyethylene cylinders of an outer size of 300 mm diameter and 470 mm total length. The innermost ring (of 20 mm wall thickness) forms a measurement cavity for the material to be assayed. In between the outermost and inner cylinders, 12 proportional counters (type SNM-28) of diameter 32mm 308 mm length each, filled with 3He gas to a pressure of 4.0E5 Pa (4 atm) served as neutron detectors. Based on previous experiences, Cd foil was not used. Heavy water or beryllium photoneutron converters of a mass of 105 and 170 g, respectively, were applied at the top of the collar. The signal processing electronics, consisting of a 400 channel analyzer in multiscaler mode of operation as a time analyzer, was almost the same as reported previously (Lakosi and Nguyen, 2008). The multiscaler receives commands from a PC through a micro-controller. Triggering the analyzer was synchronized with the linac control command pulse. A channel width (dwell time) of 100 ms was selected for the present measurements. Irradiation-measurement cycles of up to 1000 were carried out at 25 Hz. As experienced previously (Lakosi et al., 2008; Lakosi and Nguyen, 2008), time spectra acquired at 25 Hz show that the pulse length of prompt (interrogating and fission) neutrons cover a half of the 40 ms interval between two pulses. Exponential decay with a time constant of about 2 ms was observed, with a total pulse length of 20 ms. Thus, the effective time remaining for delayed neutron measurement is about 20 ms, no matter if using heavy water or beryllium converter, as observed previously. An irradiation-measurement cycle lasts for 40 ms at 25 Hz. Thus, 1000 cycles last for 40 s. In Fig. 1 a time spectrum of the D2O converter alone is displayed, taken during 1000 cycles at a mean electron current of 1.9 mA, while no U sample was in the measurement cavity. It was practically the same using the beryllium converter as well. Time spectra of HEU oxide powder samples of 10.5, 5.5, and 2.3 g mass of 36% enrichment, as well as of a 0.53 g sample of 90% enrichment are seen in Fig. 2. All the spectra were taken by the D2O converter, during 1000 irradiation-measurement cycles. The mean electron current was uniformly 1.9 mA at 25 Hz. In order to reach a sufficient degree of saturation, 20 s irradiations were
1.E+05
1.E+05
1.E+04
1.E+04 10.5g (36% enr.)
1.E+03
MEASUREMENT
1.E+02
1.E+02
1.E+01
1.E+01
1.E+00
1.E+00
1.E+05
1.E+05 1.E+04
1.E+04 2.33g (36% enr.)
1.E+03
1.E+02
1.E+01
1.E+01 0
10
20 Time [ms]
30
0.53g (90% enr.)
1.E+03
1.E+02
1.E+00
5.55g (36% enr.)
1.E+03
40
1.E+00
0
10
20
30
40
Time [ms]
Fig. 2. Time spectra of 10.5, 5.5, 2.3 g (36% enr.), and 0.53 g UO2 (90% enr.) samples without and with Pb holders.
L. Lakosi et al. / Applied Radiation and Isotopes 69 (2011) 1251–1254
1.5x104
1.0x104
5.0x103
0.0
3. Results and discussion The number of counts was summed up in the time interval for counting delayed neutrons, i.e. from channel number 200 to 400, corresponding to an effective measurement time of 20 s during 1000 cycles. Total interrogation time thus corresponds to 1 min, including an additional 20 s initial irradiation and 40 s counting time, out of which 20 s is effective. The results obtained by using heavy water converter are seen in Table 1. Indicated are the sample masses, U-235 contents, number of counts, the count rates related to the effective measurement time, and the ratios of the results for shielded to those for unshielded samples. It is seen that the results of the measurements of shielded samples agree well with results obtained without shielding. Deviations are within 710% in the most cases. They can be attributed to the limited reproducibility of the position of the converters and samples, and of the pulse amplitudes, because the linac was started again at every interrogation run (of 1000 cycles). The statistical uncertainty is 1–2% only, negligible. In Fig. 3 the data as a function of 235U mass are plotted, because the samples were not of equal isotopic composition. The signal (number of counts) is not fully linear. This issue was examined in more detail previously for LEU samples of different enrichment (Lakosi et al., 2008), and a good linear response was found up to 10g 235U mass. In that case the mass of samples was uniformly 400 g, the (low) enrichment of samples was only varied. In the present case, however, sample masses are different. Deviation from the linearity may be due to the flux depression of thermal neutrons in the interior of the bigger (and of higher enrichment) samples. The method is sensitive to the 238U content as well, because of the imperfect thermalization and fast fission, although 50–60 times weaker than to 235U (Lakosi et al., 2008). This effect is, therefore, vanishing compared with the former reason. Table 1 Results for unshielded and shielded UO2 samples by using heavy water converter at a mean electron current of 1.9 mA and 1000 cycles. Sample mass (g)
U-235 mass (g)
Leadholder
No. of counts
Eff. count rate (cps)
Ratio shielded/ unshielded
10.49 (Enr. 36%)
3.20
No Smaller Bigger
1730 17,979 19,500
893 896 972
1.003 1.088
5.55 (Enr. 36%)
1.69
No Smaller Bigger
12,611 12,086 13,300
628 602 663
0.958 1.055
2.33 (Enr. 36%)
0.71
No Smaller Bigger
6836 6423 5819
341 320 290
0.940 0.851
0.53 (Enr. 90%)
0.40
No Smaller Bigger
2317 2039 2251
115 102 112
0.880 0.972
Sample in thin Pb-container Sample in thick Pb-container Sample without Pb-container
2.0x104
Counts
uniformly carried out before starting cyclic delayed neutron measurements. The samples were interrogated both unshielded and shielded, being inserted in two lead holders of outer diameters 44 and 63 mm by heights 87 and 106 mm, respectively. Holder mass was 0.60 and 2.70 kg, with wall thickness of 4 and 14 mm, respectively. So each diagram contains a time spectrum of the sample unshielded and two spectra of the sample being placed in the lead holders. It can be seen that there are no significant differences between the spectra of the samples being shielded or not, so the actual interrogation of an unknown material can well be performed with the material left in its holder (up to 14 mm lead shielding at least).
1253
0
1
2 235
3
U content, g
Fig. 3. Results without and with Pb holders.
Results for the two converters are very similar. Assuming a detection limit of 6.8 count/s (cps) corresponding to twice the background level at 1000 cycles, this means 136 counts during the effective measurement time (20 s). Taking into account a sensitivity of about 5500 counts/g 235U content (Table 1, Fig. 3), this corresponds to a lower detection limit of about 25 mg 235U.
4. Summary No influence of a shielding container was observed (up to 14 mm lead thickness at least). Delayed neutron signal is in a good approximation linearly related to the 235U content. Minor deviation from linearity may be due to the flux depression of thermal neutrons in the interior of bigger samples. A sensitivity of about 5500 counts/g 235U was achieved. This corresponds to a lower detection limit of 25 mg 235U at a mean electron current intensity of 1.9 mA in a 1 min total interrogation time (1000 cycles). By increasing the electron current or interrogation time (number of cycles), the response can be enhanced. Half of the 40 ms time interval between pulses can only be exploited, i.e. 20 ms effective time is available for counting at 25 Hz frequency. Summing up, an efficient laboratory method and an equipment have been developed for a quantitative assay of unknown U-containing material, without opening its holder. Experience gathered and results may form a basic knowledge for designing a larger system of an active radiation portal monitor.
Acknowledgments Thanks are due to Mr. P. Hargittai for operating the linac. This work was supported by the Hungarian Atomic Energy Authority. References Church, J.A., Slaughter, D.R., Asztalos, S., Biltoft, P., Descalle, M.-A., Hall, J., Luu, T., Manatt, D., Mauger, J., Norman, E.B., Petersen, D., Prussin, S., 2007. Signals and interferences in the nuclear car wash. Nucl. Instr. Meth. B 261, 351–355. Facure, A., Falcao, R.C., Silva, A.X., Crispim, V.R., Vitorelli, J.C., 2005. A study of neutron spectra from medical linear accelerators. Appl. Radiat. Isot. 62, 69–72. Gmar, M., Jeanneau, F., Laine´, F., Makil, H., Poumarede, B., Tola, F., 2005. Assessment of actinide mass embedded in large concrete waste packages by photon interrogation and photofission. Appl. Radiat. Isot. 63, 613–619.
1254
L. Lakosi et al. / Applied Radiation and Isotopes 69 (2011) 1251–1254
Hall, J.M., Asztalos, S., Biltoft, P., Church, J., Descalle, M.-A., Luu, T., Manatt, D., Mauger, G., Norman, E., Petersen, D., Pruet, J., Prussin, S., Slaughter, D., 2007. The nuclear car wash: neutron interrogation of cargo containers to detect hidden SNM. Nucl. Instr. Meth. B 261, 337–340. Jones, J.L., Yoon, W.Y., Norman, D.R., Haskell, K.J., Zabriskie, J.M., Watson, S.M., Sterbentz, J.W., 2005. Photonuclear–based nuclear material detection system for cargo containers. Nucl. Instr. Meth. B 241, 770–776. Jones, J.L., Norman, D.R., Haskell, K.J., Sterbentz, J.W., Yoon, W.Y., Watson, S.M., Johnson, J.T., Zabriskie, J.M., Bennett, B.D., Watson, R.W., Moss, C.E., Harmon, J.F., 2006. Detection of shielded nuclear material in a cargo container. Nucl. Instr. Meth. A 562, 1085–1088. Jones, J.L., Blackburn, B.W., Watson, S.M., Norman, D.R., Hunt, A.W., 2007. Highenergy photon interrogation for nonproliferation applications. Nucl. Instr. Meth. B 261, 326–330. Kerr, P., Rowland, M., Dietrich, D., Stoeffl, W., Wheeler, B., Nakae, L., Howard, D., Hagmann, C., Newby, J., Porter, R., 2007. Active detection of small quantities of shielded highly-enriched uranium using low-dose 60 keV neutron interrogation. Nucl. Instr. Meth. B 261, 347–350. Kinlaw, M.T., Hunt, A.W., 2006. Fissionable isotope identification using the time dependence of delayed neutron emission. Nucl. Instr. Meth. A 562, 1081–1084.
Lakosi, L., Nguyen, C.T., 2008. Neutron interrogation of high-enriched uranium by a 4 MeV linac. Nucl. Instr. Meth. B 266, 3295–3301. Lakosi, L., Bagi, J., Nguyen, C.T., 2006. Quantitative non-destructive assay of PuBe neutron sources. Nucl. Instr. Meth. B 243, 385–391. Lakosi, L., Nguyen, C.T., Bagi, J., 2008. Photoneutron interrogation of low-enriched uranium induced by bremsstrahlung from a 4 MeV linac. Nucl. Instr. Meth. B 266, 295–300. Norman, D.R., Jones, J.L., Yoon, W.Y., Haskell, K.J., Sterbentz, J.W., Zabriskie, J.M., Hunt, A.W., Harmon, F., Kinlaw, M.T., 2005. Inspection applications with higher electron beam energies. Nucl. Instr. Meth. B 241, 787–792. Norman, D.R., Jones, J.L., Blackburn, B.W., Haskell, K.J., Johnson, J.T., Watson, S.M., Hunt, A.W., Spaulding, R., Harmon, F., 2007. Time-dependent delayed signatures from energetic photon interrogations. Nucl. Instr. Meth. B 261, 316–320. Sa´fa´r, J., Lakosi, L., 1994. Photoneutron production by bremsstrahlung in the target and the converter. Appl. Radiat. Isot. 45, 1057–1059. Slaughter, D.R., et al., 2007. The nuclear car wash: a system to detect nuclear weapons in commercial cargo shipments. Nucl. Instr. Meth. A 579, 349–352. Sterbentz, J.W., Jones, J.L., Yoon, W.Y., Norman, D.R., Haskell, K.J., 2007. Benchmark validation comparisons of measured and calculated delayed neutron detector responses for a pulsed photonuclear assessment technique. Nucl. Instr. Meth. B 261, 373–377.