Alternatives to 3He for Neutron Detection James Ely1 Edward Siciliano1, Richard Kouzes1, Martyn Swinhoe2 1. Pacific Northwest National Laboratory 2. Los Alamos National Laboratory IAEA Workshop March 22-24, 2011 Neutron detection is an essential aspect of interdiction of radiological threats for national security purposes since plutonium, a material used for nuclear weapons, is a significant source of fission neutrons. Radiation portal monitoring (RPM) systems, of which there are thousands deployed for national security and non-proliferation purposes, currently use 3He gas-filled proportional counters for detecting neutrons. The US supply of 3He comes almost entirely from the decay of tritium used in nuclear weapons in the US and Russia. Due to the large increase in use of 3He for national security, the supply has dwindled, and can no longer meet the demand. Consequently, a replacement technology for neutron detection is required in the very near future. In addition to alarming on the presence of actual neutron sources, national security applications also have a strict requirement for limiting neutron false alarms produced by a detector. This constrains any possible replacement neutron detection technology not to generate false neutron counts in the presence of a large gamma ray-only source. Of the currently available neutron detection technologies, BF3-filled proportional detectors, boron-lined proportional detectors, 6Li-loaded scintillating glass fiber, or non-scintillating coated plastic fiber detectors are the possible replacements for 3He detector technology—if they are proven to have appropriate capabilities. This presentation discusses these potential alternative neutron detection technologies for deployed RPM systems; and outlines what is needed for testing these alternatives and modeling to optimize their efficiency by varying moderator geometry. PNNL-SA-xxxxx

Research Project in Alternatives DOE NNSA Office of Non-Proliferation (NA-22) Project initiated in FY2009 Focus on commercially available technologies For use in portal monitor applications Provide same neutron detection capability as 3He-based Provide same level of gamma discrimination Fit in existing detector footprint Testing of commercial or near commercial modules Test neutron detection capability and gamma discrimination Several technologies appear viable Continue testing of longer term reliability and durability

Research Project in Alternatives Focus changed in FY2011 Research into safeguards applications; primarily multiplicity counters Research optimized configurations for existing materials Use available promising technologies Model and simulate to optimize moderator and detector Maximize detection of coincidence events Minimize die-away time Current multiplicity designs uses 3He at high pressure; significant challenge to identify suitable replacement

Example Multiplicity Counter Canberra Large Epi-Thermal Multiplicity Counter (LEMC) 126 3He tubes at 10 atm (1 inch dia. By 30 inches long)

Cross-sections of Neutron Detector Material Place holder to describe what we’re after Cross-section inversely proportional to neutron energy – need moderator to slow neutrons to thermal energies 5

Cross-sections of Neutron Detector Material Place holder to describe what we’re after Relatively small cross-sections for fast neutron detection via elastic scattering 6

Alternative Neutron Technology Commercially available technologies tested BF3 filled proportional tubes Boron-lined proportional tubes Scintillating glass fibers loaded with 6Li Non-scintillating fibers coated with scintillator and 6Li Multiplicity Counters Most promising alternatives Boron-10 based Lithium-6 based Less attractive Gadolinium-based: reaction products harder to detect and discriminate from other gammas Fast neutron detection: small cross sections Fission reactions: requires fissionable material

Neutron-Capture Kinematics for 3He &10B Assuming Thermal Neutrons: the Lab ~ Center of Mass, and the final-state total KE in Lab ~ Q value. Equating momenta gives values below. n + 3He  p + 3H (triton “t”) sT (thermal) = 5330 b, sT ~ 1/KEn, Q = 0.764 MeV Using KEp + KEt = Q, => KEp = 573 keV & KEt = 191 keV n + 10B  4He (alpha “a”) + 7Li sT (thermal) = 3840 b, sT ~ 1/KEn ~ 6% to g.s. with Q = 2.792 MeV => KE a = 1.777 MeV & KELi = 1.015 MeV ~ 94% to 7Li* with Q = 2.310 MeV => KE a = 1.470 MeV & KELi = 0.840 MeV 8

Evaluation Method used for 3He & BF3 Modeling and Simulation using MCNP “Reaction Rate” Method Defined as MCNP5 or MCNPX Tally Type 4 (Cell-Averaged Flux) with the Tally Multiplier Option for Reactions 9

Accuracy of Reaction-Rate Method for Simulating Total Counts in 3He Tubes 10

Considerations for BF3 Proportional Tubes Thermal cross-section is 72% of 3He Reaction products are higher energy than for 3He Better gamma discrimination High voltage requirements for BF3 proportional tubes Increases rapidly as pressure increases Max pressure ~ 1 atm to keep HV below 2-3 kV → to replace 3 atm 3He tube, will need ~ 3 tubes of BF3 at ~ 1atm (same size)

Accuracy of Reaction-Rate Method for Simulating Total Counts in BF3 Tubes 12

Evaluation Methods for Boron-Lined Tube “Surface Current” Method: Available Only with MCNPX Beta 2.7b or newer Defined as Tally Type 1 (Surface-Averaged Current) with the Neutron Capture Ion Algorithm (NCIA) on for the Physics options “Pulse-Height” Method: Also available Only with MCNPX Beta 2.7b or newer Defined as Tally Type 8 (w/out special treatment FT8 PHL “anti-coincidence” option) Also must have the NCIA on for the Physics options 13

Currents Vs. Pulse-Heights for B-Lined Tube Reaction Products 14

Measured Response of GE Reuter Stokes Prototype Multi-Tube Detector System 15

Efficiency of B-Lined Tube Vs. Lining Thickness 16

Considerations for the Boron-Lined Tube Use regular proportional gas and pressure P-10 or similar, less than 1 atm, HV < 1000V Increase surface area to increase efficiency About ½ as efficient (best case) as BF3 for same size tube For portal applications, needed 3 BF3 tubes to be equivalent to a single 3He tube at 3 atm, therefore, would need ~ 6 boron-lined tubes for equivalent capability But not enough room in current footprint, vendors went to smaller (and more) tubes to increase the surface area Straw tubes is one approach to maximize surface area 17

Neutron-Capture Kinematics for 6Li Assuming Thermal Neutrons: the Lab ~ Center of Mass, and the final-state total KE in Lab ~ Q value. Equating momenta gives values below. n + 6Li  4He (alpha “a”) + 3H (triton “t”) sT (thermal) = 940 b, sT ~ 1/Ken, Q = 4.78 MeV => KE a = 2.05 MeV & KEt = 2.73 MeV 18

Lithium-6 Zinc Sulfide (Ag) Coated Material Reaction products from 6Li generate scintillation light in the ZnS(Ag) Matrix of 6LiF crystals, ZnS and binder ZnS is opaque to scintillation light (thin layers only) Light transferred in wavelength shifting material Fibers – wavelength shifted light moves down fibers using total internal reflection Wavelength shifting light guides Collect light with photomultiplier tube Complicated mechanism allows for gamma-insensitivity via pulse shape discrimination 19

Lithium-6 Zinc Sulfide (Ag) Coated Material Pulses from gammas significantly different than from neutrons Plot from LANL paper (2000 INMM conference proceedings) 20

Lithium-6 Zinc Sulfide (Ag) Coated Material Lithium in ZnS matrix Thicker layers than boron lining (100-500 µm) Limited by ZnS opaqueness Estimate of amount of 6Li needed Use layers of 6Li matrix, with wavelength shifting material Perhaps 10x thicker per layer than optimal boron But cross section is 4x less than 10B → Need multiple layers, perhaps 5-10 to be equivalent to a single 3He tube in portal application 21

Considerations for Multiplicity Counter Canberra Large Epi-Thermal Multiplicity Counter (LEMC) 126 3He tubes at 10 atm (1 inch dia. By 30 inches long) BF3 estimate from portal work Efficiency -- will need ~ 10 for each 3He or 1260 tubes Die-away time considerations? New concept for boron – layered wire chambers? Lithium coated material estimate Will need ~ 10 layers for each 3He row – 30 layers

Lithium Coated Fibers LANL system Neutron Capture Counter for Residues (NCCR) 3 detectors shown (12 total) with 20 layers of LiF/ZnS and wavelength shifting fibers Good die away time (<5 µsec)

Multiplicity Counter Application Currently building up MCNP models to characterize technologies BF3 and boron-lined proportional tubes and 6Li coated wavelength shifting materials Starting from the LANL MNCP model of the Epi-thermal Neutron Multiplicity Counter (ENMC) Challenging to replace high pressure 3He Boron Straw tubes or other approach to increase surface area But still need to minimize die-away time Lithium Will need many layers 24

Initial Model: ENMC with 3He at 10 atm Efficiency 0.66; die-away time 23 µsec Consistent to LANL values (0.65 and 22) 25

Initial Model: ENMC with 3He at 1 atm Efficiency 0.44; die-away time 90 µsec Not huge drop in efficiency, but significant in die-away time 26

Initial Model: ENMC with BF3 at 1 atm Efficiency 0.38; die-away time 120 µsec Efficiency ~2 less than 3He, but die-away time 6x longer 27

Acknowledgements Support from: DOE NA-22 Office of Non-Proliferation and Verification, Research and Development 28