- Email: [email protected]
Simulation study of a “fission electron-collection” neutron detector
Radiation Measurements 73 (2015) 46e50
Contents lists available at ScienceDirect
Radiation Measurements journal homepage: www.elsevier.com/locate/radmeas
Simulation study of a “fission electron-collection” neutron detector Dong Wang a, *, Chuanfei Zhang b, Jianhua Zhang a a b
Institute of Nuclear Physics and Chemistry, Mianyang 621900, China China Academy of Engineering Physics, Mianyang 621900, China
h i g h l i g h t s We present a full simulation of physical processes of detection for a FECND. Giving the general properties of escaping electrons. Confirming the impact of escaping fragments. Giving a design to decrease the impact of incoming gamma flux.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 June 2014 Received in revised form 4 January 2015 Accepted 12 January 2015 Available online 13 January 2015
This work studied a “fission electron-collection” neutron detector via the Monte Carlo method. The detector consists of two metal electrodes, labeled the coated and collection electrodes, mounted in a vacuum. The first electrode is coated with triuranium octoxide. The detector uses the “fission electroncollection” technique, which does not require an intermediate material and directly collects electrons from the coating. This detector can achieve rapid, flat-energy responses, which are important for measuring pulsed neutron sources. This paper presents the physical detection processes and Monte Carlo simulation studies using the Geant4 toolkit. The results indicate that the detector sensitivity is approximately 1.5 1021 [C/(n/cm2)] and the FWHM of response function is 2.5 ns. Additionally, the escaping electrons are characterized, and the detector sensitivity is determined for various coating thicknesses. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Neutron detector Fission Electron collection Geant4
1. Introduction Neutron detectors are important for studies on high-flux neutron sources, such as Z-pinches (Ruggles et al., 2004) and nuclear reactors (Randolph and Sessions, 1989; Filliatre et al., 2010). The efficiency of a neutron detector generally depends on the incident neutron energy. Measuring the absolute neutron fluence for an unknown spectrum requires a detector with a flat-energy response. There are two common options for such measurements: a low-energy neutron detector may be used after moderating the neutrons (Hanson and McKibben, 1947), or the flat characteristic of some cross-section may be utilized in a neutron detector. Because of the favorable effects of energetic fission fragments (FFs) resulting from fission reactions, many neutron fluence
* Corresponding author. E-mail address: [email protected] (D. Wang). http://dx.doi.org/10.1016/j.radmeas.2015.01.008 1350-4487/© 2015 Elsevier Ltd. All rights reserved.
detectors, such as the fission chamber (FC) (Rossi and Staub, 1949), fission track detector (Hashemi-Nezhad et al., 2006), and fission diamond detector (Pietropaolo et al., 2011), are based on FF detection. Fission is particularly useful because of its relatively flat cross-section over several energy regions (e.g., the cross-section for 235 U only varies 17% from 1 MeV to 6 MeV). These detectors detect FFs that escape from the fissile material surface. However, numerous secondary electrons are induced by FF interactions with the fissile material because of the former's high kinetic energy. After experiencing a sufficient impulse, these electrons can travel a short distance from their initial position and even escape from the surface if they possess sufficient energy. Effectively collecting the escaping electrons can provide signals to a detector. The schematic for a “fission electron-collection” neutron detector (FECND) is shown in Fig. 1. The detector, also known as a vacuum FC (Chuklyaev and Pepelyshev, 2003), was named based on its working principle. In a FECND, the shell provides a vacuum for the electrodes. The coated electrode is plated with a fissile material on the side facing the collection electrode, and the collection
D. Wang et al. / Radiation Measurements 73 (2015) 46e50
Fig. 1. Schematic for a “fission electron-collection” neutron detector. The detector consists of a coated and collection electrode. A triuranium octoxide layer is coated on the side of the coated electrode facing the collection electrode. The detector shell maintains a vacuum.
electrode is used to collect escaped electrons. Electrical signals are generated when charged particles move between the two electrodes, and the detector operates in the current mode. This detector does not require working gases and a DC voltage, which avoids adverse effects such as ion pair recombination and electric field distortions (Poujade and Lebrun, 1999). The detector achieves a flat-energy response because the signal intensity is proportional to the fission rate, which is proportional to the fission cross section in turn. Furthermore, the small distance between the electrodes (several millimeters) and vacuum environment for electron movement give the detector a good response time, which is defined as the drift time for an electron in the gap between the electrodes (Calviani et al., 2008). The fluence evolution must be determined as a function of time for pulsed neutron sources, such as the China Fast Burst Reactor II, which requires a rapid-response detector operating in the current mode, and FECND satisfies this requirement. This paper simulates a FECND using the Geant4 toolkit.
2. Physical detection processes When the FECND is irradiated with neutrons, the coating material generates several fission products, including FFs, neutrons, and gamma rays; here, the FFs carry most of the fission energy (more than 85%). During a fission reaction, two FFs are generated and move in opposite directions. The FFs pass through the fissile material with a straight trajectory because of their large mass relative to extra-nuclear electrons. One FF may move towards the coated substrate and be deposited in it or the fissile material. Another FF may move towards the collection electrode and either escape the fissile material or be deposited in it. This effect is known as self-absorption and is considered negligible in FCs (Jammes et al., 2012). According to Bohr theory (Bohr, 1913), ions are slowed by two electric fields when passing through matter. These fields are generated by the atomic electrons and nuclei. The energy loss rate,
47
S(E) ¼ dE/dx, defines the stopping power and has two components, namely, the electronic and nuclear stopping powers. Electronic stopping may generate numerous ionized electrons. As a heavy ion, an FF has a high initial kinetic energy. Electronic stopping is the primary factor and bases on the effective charge of the FF (Ziegler et al., 1985), which is a function of the relative velocity between the FF and electrons. FFs are generated with high charge states, and their electronic stopping power decreases as they lose kinetic energy while slowing, because their effective charge decreases with the velocity. This phenomenon differs from that observed for lighter particles such as a-particles or protons. When the kinetic energy of an FF decreases below several tens of kiloelectron-volts, nuclear stopping becomes dominant. Compared to the primary energy for the FF, which is often tens of megaelectron-volts, nuclear stopping only slightly influences the FECND detection process and can be neglected. The detection efficiency for a FECND depends on the number of electrons collected by the collection electrode. This number is determined by two factors, namely, the number of electrons generated by the FFs and the probability they escape. The first factor depends on the fission reaction probability and FF energy losses, which both relate to the fissile layer thickness. The second factor also relates to the fissile layer thickness. Therefore, the FECND efficiency can be adjusted by modifying the fissile layer thickness. FFs may also escape and reduce the detector output because of their positive charge. This effect is proportional to the product of the number and average effective charge for these fragments. 3. Simulation description Simulations were performed using the Geant4 code (Version 4.9.6 p02), which has been successfully used to simulate FC €gler et al., 2013). Geant4 is an object-oriented toolkit behavior (Ko for radiation detector simulations that can be used to provide detailed information on the process for particles passing through matter. The information provided by the toolkit includes the detector geometry, materials involved, incident particle generation, particles tracking through the matter, the physical processes controlling particle interactions, and event and track storage. The geometric structure used for this simulation is shown in Fig. 1. A cylindrical aluminum shell with a diameter of 106.2 mm composed the FECND. This shell ensured the electrodes remain in a vacuum. The electrodes were two coaxial circular plates 100 mm in diameter and were located in the center of the shell. The shell, coated electrode, and collection electrode were 0.1 mm, 5 mm, and 5 mm thick, respectively. The gap between the shell and electrode was 2 mm. The coated and collection electrodes were made of aluminum. The coating material was triuranium octoxide containing enriched uranium (90% 235U and 10% 238U). The coating thickness was a few mg/cm2. The most important part of the simulation was selecting the Physics List, which contains the interactions available to the simulation. Geant4 provides several predefined Physics Lists (known as the Reference Physics Lists) that were constructed from experience and validated via past applications and experiments. Our simulation used the QGSP_BERT_HP Reference Physics List. Except for the FF generation, this Physics List considers all of the physical processes involved in this study. The G4ParaFissionModel class was introduced to simulate the FF generation. The EM Physics in the QGSP_BERT_HP Reference Physics Lists was replaced with G4EmStandardPhysics_option4 to ensure a high accuracy level (Geant4 Collaboration, 2012). To simulate lowenergy electron generation, the lower limit for the default energy range was set to 250 eV, which is the lower limit for low-energy
48
D. Wang et al. / Radiation Measurements 73 (2015) 46e50
electromagnetic processes (Agostinelli et al., 2003) in the Geant4 program. In Geant4, gamma rays, electrons, and positrons have a production threshold defined as the range cut. Using the specified range cut Geant4 calculates the corresponding energy for all materials. Even when the range cut is specified to a lower energy equivalent than the default lower limit, the energy cut is still set to the default value (Geant4 Collaboration, 2012). Therefore, the range cut for this study was set to a small value (10 nm) to ensure that the default value, 250 eV, operated properly. The installed neutron-data library was G4NDL version 4.3, which is generated from the ENDF/B-VI (Report BNL-NCS-17541, 1991) neutron cross-sectional library. The G4EMLOW version 6.32 package was installed to address photons, electrons, and positrons. The General Particle Source (GPS) tool was used to define the collimated neutron source. The detector was irradiated using mono-energetic neutrons parallel to the cylinder axis, and the neutron source diameter was 94.87 mm. The event processing was handled via the Geant4 methods UserSteppingAction, BeginOfEventAction/EndOfEventAction, and BeginOfRunAction/EndOfRunAction. During the stepping action, the steps, interactions, and particle energies involved were saved as files. The statistical uncertainty of the results was no greater than a few percent. The computation time required to reach 1% statistical uncertainty was typically 10 days when the collected particles were tallied. To shorten the computation time, batch files containing several runs were distributed to different CPU threads. 4. Results and discussion The physical processes associated with the FECND detection indicate that the escaping electrons originating from the FFs and fission gamma rays constitute a “positive” detector output. Determining the general properties for these electrons is important to characterize the detector. Table 1 shows the mechanisms for generating the escaping electrons and their corresponding proportions, and nearly 97% were generated via ion ionization. Given that some electrons or positrons generated by gamma rays are energetic (called delta rays), these particles can create further electrons. This phenomenon indicates that approximately 2.7% of the escaping electrons were generated by electron or positron ionization. The detected signal primarily originates from escaping electrons ionized by ions. The time characteristics depend on the energy distribution for these electrons. Given the significant mass difference between the FFs and electrons, their interactions can be estimated using classic collision theory. According to this theory, the maximum energy transferred from an ion with mass m and kinetic energy E to an electron with mass m0 in one collision is 4Em0/m, or approximately 1/500 of the energy per nucleon. Because the mass of an electron is small relative to an FF, only a small fraction of the kinetic energy is transferred from the FF to the electron. Therefore, the kinetic energy for the electron is approximately a few kiloelectron-volts. Fig. 2 shows the energy
Fig. 2. Energy distribution for escaping electrons ionized by ions. The incident neutron energy is 1 MeV, and the triuranium octoxide coating is 1 mg/cm2 thick.
distribution for escaping electrons ionized by ions. The electron energy is below 2.5 keV, which agrees with the results predicted by classic collision theory. During the Geant4 simulation, the G4Step class stores the transient information for a step and includes its two endpoints, namely, the PreStepPoint and PostStepPoint, which contains coordinates and volumes for these points (Geant4 Collaboration, 2012). The elapsed flight time during the step is the time increment between the two endpoints and can be obtained using the GetDeltaTime method from the G4Step class. Only one step is required to connect the collection to the coated electrodes because a vacuum exists between them. Thus, the DeltaTime is the flight time between the two electrodes. Fig. 3 shows the DeltaTime distribution for escaping electrons ionized by ions. The DeltaTime represents the charge collection time for escaping electrons, and its distribution is the sensitive volume response function for the FECND, g0(t). The final FECND response, g(t), results from the convolution of g0(t) with the response from the measuring circuit, h(t). Neglecting the distributed capacitances, the capacitance of the FECND is 34.8 pF. A distributed inductance and internal resistance for the detector of 2.5 108 H and 50 U, respectively, yields a 2.5 ns FWHM for the final response function, which is similar to semiconductor
Table 1 Production mechanisms for escaping electrons and their corresponding electron proportion. The incident neutron energy is 1 MeV, and the triuranium octoxide coating thickness is 1 mg/cm2. Geant4 electromagnetic interaction type
Proportion
G4ionionization G4eionization G4Comptonscattering G4photoelectriceffect
97.17% 2.73% 0.04% 0.06%
Fig. 3. DeltaTime distribution for escaping electrons ionized by ions.
D. Wang et al. / Radiation Measurements 73 (2015) 46e50
detectors (Knoll, 2010). This characteristic allows pulsed neutron radiation bursts approximately 1 ms in duration to be monitored. In addition to electrons, FFs may also potentially escape. Escaping FFs yield a negative detector output. Despite their lower quantity compared to escaping electrons, FFs significantly affect the detector because they may retain higher charge states after escaping the fissile layer. Table 2 provides statistical information for FFs that escape the fissile layer when the incident neutron energy is 1 MeV and the triuranium octoxide thickness coating is 1 mg/cm2 Table 3 shows the number of electrons collected and total charge for the collected FFs using different fissile coating thicknesses. The difference between these two values defines the net collecting electric charge. The maximum ratio for the two parts is approximately 16%, which indicates that the escaping FFs lead to a maximum signal reduction of 16%. Table 3 indicates the number of collected FFs increases with increasing layer thickness up to 7 mg/cm2, which is approximately equal to the extreme FF range. Conservation of momentum means light fragments receive more energy from a single fission event and consequently exhibit a larger range than heavy fragments. Calculations using the SRIM (Ziegler et al., 1985) Monte Carlo code indicate that a typical light fragment (95Mo, 90 MeV) from a triuranium octoxide coating is 7.4 mg/cm2, which agrees well with the Geant4 simulation results. The FECND sensitivity [C/(n/cm2)] is defined as the ratio between the output electricity and number of neutrons that reach the detector surface per unit area. Fig. 4 shows sensitivities obtained from different fissile coating thicknesses. The detector sensitivity was approximately 1.5 1021 [C/(n/cm2)]. The detector achieves a flat-energy response because the signal is proportional to the fission cross section. Fig. 5 compares the flatness for the fission cross section and net collecting electric charge provided by the simulation. The results meet expectations. The vacuum quality must be determined to ensure proper FECND operation. The calculations indicate that the device sensitivity changes less than 3% when the internal pressure is 10 Pa. This result indicates that residual gas ionization can be neglected if the internal pressure is below 10 Pa. Background signals from other sources must be differentiated because the FECND operates in the current mode (Filliatre et al., 2011). These other signal sources include incoming gamma fluxes, alpha particles from coating decay, and beta minus from activating the detector structures. For incoming gamma fluxes, signals can be suppressed relatively by adding a layer (to the other surface of the coated electrode) and a collection electrode (to the other side of the coated electrode) and adjusting the electrode thicknesses and distances. Adding another layer to the coated electrode can increase the neutron sensitivity, while adjusting the thickness and distance can reverse the electrons impact from gamma flux, both enhance the signal-to-noise (SNR) ratio of the detector. Table 4 details the design and corresponding sensitivities for neutrons and gamma particles. For alpha particles, the background current was approximately 0.02 pA, which is equivalent to the current from a 107 n/ cm2/s neutron flux, which is the lower measurement limit for the FECND. The simulations for beta minus indicate that the impact is negligible for pulsed neutron only a few microseconds in duration. For neutron measurements in the presence of a high gamma
Table 2 Information on escaping FFs. The simulated number of incident neutrons is 1 109. Average primary energy (MeV)
Number of escaping FFs
Sum of charge Average charge (elementary charge) (elementary charge)
57.75
2182
20,817
9.54
49
Table 3 Signal constitution and sensitivity for layers with different thicknesses. The simulated number of incident neutrons is 1 109. Layer thickness (mg/cm2)
Number of collected electrons
Proportion Number of Total charge for the collected collected FFs (elementary (%) FFs charge)
1.0 2.0 3.0 4.0 5.0 6.0 7.0
163,436 210,490 237,646 242,407 243,165 248,192 249,621
20,396 30,348 36,773 39,027 40,023 41,210 41,272
12.48 14.42 15.47 16.10 16.46 16.60 16.53
2094 3092 4941 5847 6295 6605 6680
Fig. 4. Sensitivities for different fissile coating thicknesses. The incident neutron energy is 1 MeV.
Fig. 5. Fission cross section and net collecting electric charge for different neutron energies. The fission cross section data were obtained using the Geant4 database, and the electric output was determined from the simulated results. These values were normalized to 1 MeV. The triuranium octoxide was 1 mg/cm2 thick for the simulation.
flux, comparing the FECND with two related concepts, namely, the FC and the self-powered neutron detector (SPND), is beneficial. An SPND is inappropriate in this situation because it cannot distinguish signals from gamma events (Filliatre et al., 2010). An FC operating in the pulse or Campbelling mode is best when the neutron source operates in the steady-state mode because the contribution from gamma events can be rejected or significantly
50
D. Wang et al. / Radiation Measurements 73 (2015) 46e50
Table 4 Detailed design for enhancing the SNR. Parameter
Parameter value
Collection electrode material Coated electrode material Shell material Layer material Diameter of shell Diameter of electrodes Thickness of shell Thickness of collection electrodes Thickness of coated electrode Thickness of layers Distance between the electrodes Neutron (1 MeV) sensitivity Gamma (1e1.75 MeV) sensitivity
Aluminum Aluminum Aluminum Triuranium octoxide 106.2 mm 100 mm 0.1 mm 400 mm 15 mm 4.2 mg/cm2 7 mm ~3.0 1021 [C/(n/cm2)] <6.0 1022 [C/(n/cm2)]
suppressed (Filliatre et al., 2011). The FECND is best-suited for detecting short-pulsed neutron sources because it operates in the current mode, responds quickly, and reflects fluence evolution as a function of time. 5. Conclusions This study detailed Monte Carlo simulations of a FECND. The detector sensitivity is approximately 1.5 1021 [C/(n/cm2)]. The FWHM of the FECND response function is 2.5 ns, which allows pulsed neutron radiation bursts ~1 ms in duration to be monitored. The simulation results indicate that over 97% of the escaping electrons originate from ionization by ions and that escaping FFs decrease the detector output approximately 16%. The energy response flatness for the FECND exhibits a trend similar to the fission cross section of a fissile material. The FECND provides several advantages because of its simple design, freedom from a working gas and DC voltage, and good time and flat-energy response characteristics. Acknowledgments This work was supported by the National Natural Science
Foundation of China (Grant No. 11205141) and the Science and Technology Foundation of China Academy of Engineering Physics (Grant No. 2012B0103003).
References Agostinelli, S., Allison, J., Amako, K., et al., 2003. Geant4da simulation toolkit. Nucl. Instrum. Methods. Phys. Res. A 506, 250e303. Bohr, N., 1913. On the theory of the decrease of velocity of moving electrified particles on passing through matter. Philos. Mag. 25, 10e31. Calviani, M., Cennini, P., Karadimos, D., et al., 2008. A fast ionization chamber for fission cross-section measurements at n_TOF. Nucl. Instrum. Methods. Phys. Res. A 594, 220e227. Chuklyaev, S.V., Pepelyshev, Yu.N., 2003. The time resolution of a vacuum fission chamber. Instrum. Exp. Tech. 46, 748e752. Filliatre, P., Jammes, C., Geslot, B., et al., 2010. In vessel neutron instrumentation for sodium-cooled fast reactors: type, lifetime and location. Ann. Nucl. Energy 37, 1435e1442. Filliatre, P., Vermeeren, L., Jammes, C., et al., 2011. Estimating the g-ray contribution to the signal of fission chambers with Monte Carlo simulations. Nucl. Instrum. Methods. Phys. Res. A 648, 228e237. Geant4 Collaboration, 2012. Geant4 User's Guide for Application Developers. Hanson, A.O., McKibben, M.L., 1947. A neutron detector having uniform sensitivity from 10 keV to 3MeV. Phys. Rev. 72, 673e677. Hashemi-Nezhad, S.R., Zhuk, I.V., Potapenko, A.S., et al., 2006. Calibration of track detectors for fission rate determination: an experimental and theoretical study. Nucl. Instrum. Methods. Phys. Res. A 568, 816e825. Jammes, C., Filliatre, P., Loiseau, P., et al., 2012. On the impact of the fissile coating on the fission chamber signal. Nucl. Instrum. Methods. Phys. Res. A 681, 101e109. Knoll, G.F., 2010. Radiation Detection and Measurement. John Wiley & Sons, New York. €gler, T., Beyer, R., Hannaske, R., et al., 2013. Fission Experiments at nELBE. Phys. Ko Proc. 47, 178e183. Pietropaolo, A., Andreani, C., Rebai, M., et al., 2011. Fission diamond detectors for fast-neutron ToF spectroscopy. Europhys. Lett. 94, 62001. Poujade, O., Lebrun, A., 1999. Modeling of the saturation current of a fission chamber taking into account the distorsion of electric field due to space charge effects. Nucl. Instrum. Methods. Phys. Res. A 433, 673e682. Randolph, H.W., Sessions, H.T., 1989. Dynamic range and time response of a new hafnium neutron detector. Proc. Am. Power. Conf. 51, 722e724. Report BNL-NCS-17541, ENDF/B-VI, 1991. National Nuclear Data Center, Brookhaven National Laboratory, USA. Rossi, B.B., Staub, H.H., 1949. Ionization Chambers and Counters Experimental Techniques, vol. 2. McGraw-HillBook, New York. Ruggles, L.E., Porter Jr., J.L., Simpson, W.W., et al., 2004. High sensitivity neutron detector for Z. Rev. Sci. Instrum. 75, 3595e3597. Ziegler, J.F., Biersack, J.P., Littmark, U., 1985. The Stopping and Range of Ions in Solids. Pergamon Press, New York.
Contents lists available at ScienceDirect
Radiation Measurements journal homepage: www.elsevier.com/locate/radmeas
Simulation study of a “fission electron-collection” neutron detector Dong Wang a, *, Chuanfei Zhang b, Jianhua Zhang a a b
Institute of Nuclear Physics and Chemistry, Mianyang 621900, China China Academy of Engineering Physics, Mianyang 621900, China
h i g h l i g h t s We present a full simulation of physical processes of detection for a FECND. Giving the general properties of escaping electrons. Confirming the impact of escaping fragments. Giving a design to decrease the impact of incoming gamma flux.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 June 2014 Received in revised form 4 January 2015 Accepted 12 January 2015 Available online 13 January 2015
This work studied a “fission electron-collection” neutron detector via the Monte Carlo method. The detector consists of two metal electrodes, labeled the coated and collection electrodes, mounted in a vacuum. The first electrode is coated with triuranium octoxide. The detector uses the “fission electroncollection” technique, which does not require an intermediate material and directly collects electrons from the coating. This detector can achieve rapid, flat-energy responses, which are important for measuring pulsed neutron sources. This paper presents the physical detection processes and Monte Carlo simulation studies using the Geant4 toolkit. The results indicate that the detector sensitivity is approximately 1.5 1021 [C/(n/cm2)] and the FWHM of response function is 2.5 ns. Additionally, the escaping electrons are characterized, and the detector sensitivity is determined for various coating thicknesses. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Neutron detector Fission Electron collection Geant4
1. Introduction Neutron detectors are important for studies on high-flux neutron sources, such as Z-pinches (Ruggles et al., 2004) and nuclear reactors (Randolph and Sessions, 1989; Filliatre et al., 2010). The efficiency of a neutron detector generally depends on the incident neutron energy. Measuring the absolute neutron fluence for an unknown spectrum requires a detector with a flat-energy response. There are two common options for such measurements: a low-energy neutron detector may be used after moderating the neutrons (Hanson and McKibben, 1947), or the flat characteristic of some cross-section may be utilized in a neutron detector. Because of the favorable effects of energetic fission fragments (FFs) resulting from fission reactions, many neutron fluence
* Corresponding author. E-mail address: [email protected] (D. Wang). http://dx.doi.org/10.1016/j.radmeas.2015.01.008 1350-4487/© 2015 Elsevier Ltd. All rights reserved.
detectors, such as the fission chamber (FC) (Rossi and Staub, 1949), fission track detector (Hashemi-Nezhad et al., 2006), and fission diamond detector (Pietropaolo et al., 2011), are based on FF detection. Fission is particularly useful because of its relatively flat cross-section over several energy regions (e.g., the cross-section for 235 U only varies 17% from 1 MeV to 6 MeV). These detectors detect FFs that escape from the fissile material surface. However, numerous secondary electrons are induced by FF interactions with the fissile material because of the former's high kinetic energy. After experiencing a sufficient impulse, these electrons can travel a short distance from their initial position and even escape from the surface if they possess sufficient energy. Effectively collecting the escaping electrons can provide signals to a detector. The schematic for a “fission electron-collection” neutron detector (FECND) is shown in Fig. 1. The detector, also known as a vacuum FC (Chuklyaev and Pepelyshev, 2003), was named based on its working principle. In a FECND, the shell provides a vacuum for the electrodes. The coated electrode is plated with a fissile material on the side facing the collection electrode, and the collection
D. Wang et al. / Radiation Measurements 73 (2015) 46e50
Fig. 1. Schematic for a “fission electron-collection” neutron detector. The detector consists of a coated and collection electrode. A triuranium octoxide layer is coated on the side of the coated electrode facing the collection electrode. The detector shell maintains a vacuum.
electrode is used to collect escaped electrons. Electrical signals are generated when charged particles move between the two electrodes, and the detector operates in the current mode. This detector does not require working gases and a DC voltage, which avoids adverse effects such as ion pair recombination and electric field distortions (Poujade and Lebrun, 1999). The detector achieves a flat-energy response because the signal intensity is proportional to the fission rate, which is proportional to the fission cross section in turn. Furthermore, the small distance between the electrodes (several millimeters) and vacuum environment for electron movement give the detector a good response time, which is defined as the drift time for an electron in the gap between the electrodes (Calviani et al., 2008). The fluence evolution must be determined as a function of time for pulsed neutron sources, such as the China Fast Burst Reactor II, which requires a rapid-response detector operating in the current mode, and FECND satisfies this requirement. This paper simulates a FECND using the Geant4 toolkit.
2. Physical detection processes When the FECND is irradiated with neutrons, the coating material generates several fission products, including FFs, neutrons, and gamma rays; here, the FFs carry most of the fission energy (more than 85%). During a fission reaction, two FFs are generated and move in opposite directions. The FFs pass through the fissile material with a straight trajectory because of their large mass relative to extra-nuclear electrons. One FF may move towards the coated substrate and be deposited in it or the fissile material. Another FF may move towards the collection electrode and either escape the fissile material or be deposited in it. This effect is known as self-absorption and is considered negligible in FCs (Jammes et al., 2012). According to Bohr theory (Bohr, 1913), ions are slowed by two electric fields when passing through matter. These fields are generated by the atomic electrons and nuclei. The energy loss rate,
47
S(E) ¼ dE/dx, defines the stopping power and has two components, namely, the electronic and nuclear stopping powers. Electronic stopping may generate numerous ionized electrons. As a heavy ion, an FF has a high initial kinetic energy. Electronic stopping is the primary factor and bases on the effective charge of the FF (Ziegler et al., 1985), which is a function of the relative velocity between the FF and electrons. FFs are generated with high charge states, and their electronic stopping power decreases as they lose kinetic energy while slowing, because their effective charge decreases with the velocity. This phenomenon differs from that observed for lighter particles such as a-particles or protons. When the kinetic energy of an FF decreases below several tens of kiloelectron-volts, nuclear stopping becomes dominant. Compared to the primary energy for the FF, which is often tens of megaelectron-volts, nuclear stopping only slightly influences the FECND detection process and can be neglected. The detection efficiency for a FECND depends on the number of electrons collected by the collection electrode. This number is determined by two factors, namely, the number of electrons generated by the FFs and the probability they escape. The first factor depends on the fission reaction probability and FF energy losses, which both relate to the fissile layer thickness. The second factor also relates to the fissile layer thickness. Therefore, the FECND efficiency can be adjusted by modifying the fissile layer thickness. FFs may also escape and reduce the detector output because of their positive charge. This effect is proportional to the product of the number and average effective charge for these fragments. 3. Simulation description Simulations were performed using the Geant4 code (Version 4.9.6 p02), which has been successfully used to simulate FC €gler et al., 2013). Geant4 is an object-oriented toolkit behavior (Ko for radiation detector simulations that can be used to provide detailed information on the process for particles passing through matter. The information provided by the toolkit includes the detector geometry, materials involved, incident particle generation, particles tracking through the matter, the physical processes controlling particle interactions, and event and track storage. The geometric structure used for this simulation is shown in Fig. 1. A cylindrical aluminum shell with a diameter of 106.2 mm composed the FECND. This shell ensured the electrodes remain in a vacuum. The electrodes were two coaxial circular plates 100 mm in diameter and were located in the center of the shell. The shell, coated electrode, and collection electrode were 0.1 mm, 5 mm, and 5 mm thick, respectively. The gap between the shell and electrode was 2 mm. The coated and collection electrodes were made of aluminum. The coating material was triuranium octoxide containing enriched uranium (90% 235U and 10% 238U). The coating thickness was a few mg/cm2. The most important part of the simulation was selecting the Physics List, which contains the interactions available to the simulation. Geant4 provides several predefined Physics Lists (known as the Reference Physics Lists) that were constructed from experience and validated via past applications and experiments. Our simulation used the QGSP_BERT_HP Reference Physics List. Except for the FF generation, this Physics List considers all of the physical processes involved in this study. The G4ParaFissionModel class was introduced to simulate the FF generation. The EM Physics in the QGSP_BERT_HP Reference Physics Lists was replaced with G4EmStandardPhysics_option4 to ensure a high accuracy level (Geant4 Collaboration, 2012). To simulate lowenergy electron generation, the lower limit for the default energy range was set to 250 eV, which is the lower limit for low-energy
48
D. Wang et al. / Radiation Measurements 73 (2015) 46e50
electromagnetic processes (Agostinelli et al., 2003) in the Geant4 program. In Geant4, gamma rays, electrons, and positrons have a production threshold defined as the range cut. Using the specified range cut Geant4 calculates the corresponding energy for all materials. Even when the range cut is specified to a lower energy equivalent than the default lower limit, the energy cut is still set to the default value (Geant4 Collaboration, 2012). Therefore, the range cut for this study was set to a small value (10 nm) to ensure that the default value, 250 eV, operated properly. The installed neutron-data library was G4NDL version 4.3, which is generated from the ENDF/B-VI (Report BNL-NCS-17541, 1991) neutron cross-sectional library. The G4EMLOW version 6.32 package was installed to address photons, electrons, and positrons. The General Particle Source (GPS) tool was used to define the collimated neutron source. The detector was irradiated using mono-energetic neutrons parallel to the cylinder axis, and the neutron source diameter was 94.87 mm. The event processing was handled via the Geant4 methods UserSteppingAction, BeginOfEventAction/EndOfEventAction, and BeginOfRunAction/EndOfRunAction. During the stepping action, the steps, interactions, and particle energies involved were saved as files. The statistical uncertainty of the results was no greater than a few percent. The computation time required to reach 1% statistical uncertainty was typically 10 days when the collected particles were tallied. To shorten the computation time, batch files containing several runs were distributed to different CPU threads. 4. Results and discussion The physical processes associated with the FECND detection indicate that the escaping electrons originating from the FFs and fission gamma rays constitute a “positive” detector output. Determining the general properties for these electrons is important to characterize the detector. Table 1 shows the mechanisms for generating the escaping electrons and their corresponding proportions, and nearly 97% were generated via ion ionization. Given that some electrons or positrons generated by gamma rays are energetic (called delta rays), these particles can create further electrons. This phenomenon indicates that approximately 2.7% of the escaping electrons were generated by electron or positron ionization. The detected signal primarily originates from escaping electrons ionized by ions. The time characteristics depend on the energy distribution for these electrons. Given the significant mass difference between the FFs and electrons, their interactions can be estimated using classic collision theory. According to this theory, the maximum energy transferred from an ion with mass m and kinetic energy E to an electron with mass m0 in one collision is 4Em0/m, or approximately 1/500 of the energy per nucleon. Because the mass of an electron is small relative to an FF, only a small fraction of the kinetic energy is transferred from the FF to the electron. Therefore, the kinetic energy for the electron is approximately a few kiloelectron-volts. Fig. 2 shows the energy
Fig. 2. Energy distribution for escaping electrons ionized by ions. The incident neutron energy is 1 MeV, and the triuranium octoxide coating is 1 mg/cm2 thick.
distribution for escaping electrons ionized by ions. The electron energy is below 2.5 keV, which agrees with the results predicted by classic collision theory. During the Geant4 simulation, the G4Step class stores the transient information for a step and includes its two endpoints, namely, the PreStepPoint and PostStepPoint, which contains coordinates and volumes for these points (Geant4 Collaboration, 2012). The elapsed flight time during the step is the time increment between the two endpoints and can be obtained using the GetDeltaTime method from the G4Step class. Only one step is required to connect the collection to the coated electrodes because a vacuum exists between them. Thus, the DeltaTime is the flight time between the two electrodes. Fig. 3 shows the DeltaTime distribution for escaping electrons ionized by ions. The DeltaTime represents the charge collection time for escaping electrons, and its distribution is the sensitive volume response function for the FECND, g0(t). The final FECND response, g(t), results from the convolution of g0(t) with the response from the measuring circuit, h(t). Neglecting the distributed capacitances, the capacitance of the FECND is 34.8 pF. A distributed inductance and internal resistance for the detector of 2.5 108 H and 50 U, respectively, yields a 2.5 ns FWHM for the final response function, which is similar to semiconductor
Table 1 Production mechanisms for escaping electrons and their corresponding electron proportion. The incident neutron energy is 1 MeV, and the triuranium octoxide coating thickness is 1 mg/cm2. Geant4 electromagnetic interaction type
Proportion
G4ionionization G4eionization G4Comptonscattering G4photoelectriceffect
97.17% 2.73% 0.04% 0.06%
Fig. 3. DeltaTime distribution for escaping electrons ionized by ions.
D. Wang et al. / Radiation Measurements 73 (2015) 46e50
detectors (Knoll, 2010). This characteristic allows pulsed neutron radiation bursts approximately 1 ms in duration to be monitored. In addition to electrons, FFs may also potentially escape. Escaping FFs yield a negative detector output. Despite their lower quantity compared to escaping electrons, FFs significantly affect the detector because they may retain higher charge states after escaping the fissile layer. Table 2 provides statistical information for FFs that escape the fissile layer when the incident neutron energy is 1 MeV and the triuranium octoxide thickness coating is 1 mg/cm2 Table 3 shows the number of electrons collected and total charge for the collected FFs using different fissile coating thicknesses. The difference between these two values defines the net collecting electric charge. The maximum ratio for the two parts is approximately 16%, which indicates that the escaping FFs lead to a maximum signal reduction of 16%. Table 3 indicates the number of collected FFs increases with increasing layer thickness up to 7 mg/cm2, which is approximately equal to the extreme FF range. Conservation of momentum means light fragments receive more energy from a single fission event and consequently exhibit a larger range than heavy fragments. Calculations using the SRIM (Ziegler et al., 1985) Monte Carlo code indicate that a typical light fragment (95Mo, 90 MeV) from a triuranium octoxide coating is 7.4 mg/cm2, which agrees well with the Geant4 simulation results. The FECND sensitivity [C/(n/cm2)] is defined as the ratio between the output electricity and number of neutrons that reach the detector surface per unit area. Fig. 4 shows sensitivities obtained from different fissile coating thicknesses. The detector sensitivity was approximately 1.5 1021 [C/(n/cm2)]. The detector achieves a flat-energy response because the signal is proportional to the fission cross section. Fig. 5 compares the flatness for the fission cross section and net collecting electric charge provided by the simulation. The results meet expectations. The vacuum quality must be determined to ensure proper FECND operation. The calculations indicate that the device sensitivity changes less than 3% when the internal pressure is 10 Pa. This result indicates that residual gas ionization can be neglected if the internal pressure is below 10 Pa. Background signals from other sources must be differentiated because the FECND operates in the current mode (Filliatre et al., 2011). These other signal sources include incoming gamma fluxes, alpha particles from coating decay, and beta minus from activating the detector structures. For incoming gamma fluxes, signals can be suppressed relatively by adding a layer (to the other surface of the coated electrode) and a collection electrode (to the other side of the coated electrode) and adjusting the electrode thicknesses and distances. Adding another layer to the coated electrode can increase the neutron sensitivity, while adjusting the thickness and distance can reverse the electrons impact from gamma flux, both enhance the signal-to-noise (SNR) ratio of the detector. Table 4 details the design and corresponding sensitivities for neutrons and gamma particles. For alpha particles, the background current was approximately 0.02 pA, which is equivalent to the current from a 107 n/ cm2/s neutron flux, which is the lower measurement limit for the FECND. The simulations for beta minus indicate that the impact is negligible for pulsed neutron only a few microseconds in duration. For neutron measurements in the presence of a high gamma
Table 2 Information on escaping FFs. The simulated number of incident neutrons is 1 109. Average primary energy (MeV)
Number of escaping FFs
Sum of charge Average charge (elementary charge) (elementary charge)
57.75
2182
20,817
9.54
49
Table 3 Signal constitution and sensitivity for layers with different thicknesses. The simulated number of incident neutrons is 1 109. Layer thickness (mg/cm2)
Number of collected electrons
Proportion Number of Total charge for the collected collected FFs (elementary (%) FFs charge)
1.0 2.0 3.0 4.0 5.0 6.0 7.0
163,436 210,490 237,646 242,407 243,165 248,192 249,621
20,396 30,348 36,773 39,027 40,023 41,210 41,272
12.48 14.42 15.47 16.10 16.46 16.60 16.53
2094 3092 4941 5847 6295 6605 6680
Fig. 4. Sensitivities for different fissile coating thicknesses. The incident neutron energy is 1 MeV.
Fig. 5. Fission cross section and net collecting electric charge for different neutron energies. The fission cross section data were obtained using the Geant4 database, and the electric output was determined from the simulated results. These values were normalized to 1 MeV. The triuranium octoxide was 1 mg/cm2 thick for the simulation.
flux, comparing the FECND with two related concepts, namely, the FC and the self-powered neutron detector (SPND), is beneficial. An SPND is inappropriate in this situation because it cannot distinguish signals from gamma events (Filliatre et al., 2010). An FC operating in the pulse or Campbelling mode is best when the neutron source operates in the steady-state mode because the contribution from gamma events can be rejected or significantly
50
D. Wang et al. / Radiation Measurements 73 (2015) 46e50
Table 4 Detailed design for enhancing the SNR. Parameter
Parameter value
Collection electrode material Coated electrode material Shell material Layer material Diameter of shell Diameter of electrodes Thickness of shell Thickness of collection electrodes Thickness of coated electrode Thickness of layers Distance between the electrodes Neutron (1 MeV) sensitivity Gamma (1e1.75 MeV) sensitivity
Aluminum Aluminum Aluminum Triuranium octoxide 106.2 mm 100 mm 0.1 mm 400 mm 15 mm 4.2 mg/cm2 7 mm ~3.0 1021 [C/(n/cm2)] <6.0 1022 [C/(n/cm2)]
suppressed (Filliatre et al., 2011). The FECND is best-suited for detecting short-pulsed neutron sources because it operates in the current mode, responds quickly, and reflects fluence evolution as a function of time. 5. Conclusions This study detailed Monte Carlo simulations of a FECND. The detector sensitivity is approximately 1.5 1021 [C/(n/cm2)]. The FWHM of the FECND response function is 2.5 ns, which allows pulsed neutron radiation bursts ~1 ms in duration to be monitored. The simulation results indicate that over 97% of the escaping electrons originate from ionization by ions and that escaping FFs decrease the detector output approximately 16%. The energy response flatness for the FECND exhibits a trend similar to the fission cross section of a fissile material. The FECND provides several advantages because of its simple design, freedom from a working gas and DC voltage, and good time and flat-energy response characteristics. Acknowledgments This work was supported by the National Natural Science
Foundation of China (Grant No. 11205141) and the Science and Technology Foundation of China Academy of Engineering Physics (Grant No. 2012B0103003).
References Agostinelli, S., Allison, J., Amako, K., et al., 2003. Geant4da simulation toolkit. Nucl. Instrum. Methods. Phys. Res. A 506, 250e303. Bohr, N., 1913. On the theory of the decrease of velocity of moving electrified particles on passing through matter. Philos. Mag. 25, 10e31. Calviani, M., Cennini, P., Karadimos, D., et al., 2008. A fast ionization chamber for fission cross-section measurements at n_TOF. Nucl. Instrum. Methods. Phys. Res. A 594, 220e227. Chuklyaev, S.V., Pepelyshev, Yu.N., 2003. The time resolution of a vacuum fission chamber. Instrum. Exp. Tech. 46, 748e752. Filliatre, P., Jammes, C., Geslot, B., et al., 2010. In vessel neutron instrumentation for sodium-cooled fast reactors: type, lifetime and location. Ann. Nucl. Energy 37, 1435e1442. Filliatre, P., Vermeeren, L., Jammes, C., et al., 2011. Estimating the g-ray contribution to the signal of fission chambers with Monte Carlo simulations. Nucl. Instrum. Methods. Phys. Res. A 648, 228e237. Geant4 Collaboration, 2012. Geant4 User's Guide for Application Developers. Hanson, A.O., McKibben, M.L., 1947. A neutron detector having uniform sensitivity from 10 keV to 3MeV. Phys. Rev. 72, 673e677. Hashemi-Nezhad, S.R., Zhuk, I.V., Potapenko, A.S., et al., 2006. Calibration of track detectors for fission rate determination: an experimental and theoretical study. Nucl. Instrum. Methods. Phys. Res. A 568, 816e825. Jammes, C., Filliatre, P., Loiseau, P., et al., 2012. On the impact of the fissile coating on the fission chamber signal. Nucl. Instrum. Methods. Phys. Res. A 681, 101e109. Knoll, G.F., 2010. Radiation Detection and Measurement. John Wiley & Sons, New York. €gler, T., Beyer, R., Hannaske, R., et al., 2013. Fission Experiments at nELBE. Phys. Ko Proc. 47, 178e183. Pietropaolo, A., Andreani, C., Rebai, M., et al., 2011. Fission diamond detectors for fast-neutron ToF spectroscopy. Europhys. Lett. 94, 62001. Poujade, O., Lebrun, A., 1999. Modeling of the saturation current of a fission chamber taking into account the distorsion of electric field due to space charge effects. Nucl. Instrum. Methods. Phys. Res. A 433, 673e682. Randolph, H.W., Sessions, H.T., 1989. Dynamic range and time response of a new hafnium neutron detector. Proc. Am. Power. Conf. 51, 722e724. Report BNL-NCS-17541, ENDF/B-VI, 1991. National Nuclear Data Center, Brookhaven National Laboratory, USA. Rossi, B.B., Staub, H.H., 1949. Ionization Chambers and Counters Experimental Techniques, vol. 2. McGraw-HillBook, New York. Ruggles, L.E., Porter Jr., J.L., Simpson, W.W., et al., 2004. High sensitivity neutron detector for Z. Rev. Sci. Instrum. 75, 3595e3597. Ziegler, J.F., Biersack, J.P., Littmark, U., 1985. The Stopping and Range of Ions in Solids. Pergamon Press, New York.