Thermal neutron scintillators using unenriched boron nitride and zinc sulphide
John McMillan
ANSRI Jan 2015
Neutron detectors 1970-2008 3He
proportional tubes were the industry standard
Greatest efficiency achieved by slowing neutrons to thermal energy where interaction cross-sections are highest.
Helium-3
The shortage of He-3 is an international crisis.
Due to the shortage of Helium-3, the US Dept of Homeland Security has put on hold all installations of radiation portal monitors at ports and borders as of Nov 2009.
He-3 is used in virtually all portal monitors in thermal neutron detectors. The (US) annual demand is estimated at 65000 litres, There is essentially no source that can meet this demand.
Supply is dwindling due to reduced use of tritium. Price has risen from $100 to $2000/litre in recent years.
– see "The 3He Supply Problem", R.L.Kouzes, PNNL-18388
Thermal neutron detectors which don't use Helium-3 my PhD on the "Barton detectors" (Polytechnic of North London - University of Leeds) Layered ZnS-6LiF scintillators with wavelength shifter readout Pulse-counting neutron discrimination
Detector design
detector
Detector design
detector
Pulse counting discrimination in ZnS
Pulses in time gate counted
Caines P.J., M Phil Thesis, University of London, 1972
Davidson P.L. Rutherford Laboratory Report RL-77-106A, 1977
Features of the PNL-Leeds detectors Active volume 90 x 14.4 x 14.4cm 37% efficient for 252Cf fission neutrons (8 detectors surrounding source) Totally insensitive to gammas and muons Robust, stable operation over many years over a range of temperatures in harsh environments Woodhead Railway Tunnel, Yorkshire Holborn Underground station, London Boulby Potash mine, Yorkshire (1km depth)
Thermal Neutron Detectors for Portal Applications large-area, square metres needed for portals unambiguous, good signal-to-noise ratio, high efficiency, low background real-time signal discrimination (not compute-intensive post processed) deployable reasonably robust stable over many years in harsh environments transportable minimal health & safety implications must use easily available materials
Improvements to existing design Choice of capture material 6LiF
is a controlled material and increasingly expensive
Can we make worse (but very much cheaper) detectors using boron compounds? Capture cross-section higher - but releases less energy Can probably use natural rather than isotopically enriched material.
Usable thermal neutron capture reactions
abundances
Boron Nitride Need inert boron compound
needs to be white or colourless
easily available with controlled grain size
Hexagonal boron nitride is available in ~5um platelets for cosmetic applications
Cubic boron nitride is available in controlled sizes as an abrasive
Geometric improvements PNL-Leeds detectors were optimized for volume configuration (maximum efficiency, lowest background…) Portal applications need to optimize effective area per unit cost
efficiency × area price
Smaller or less efficient detectors can still win if they are very much cheaper!
Geometric optimization MCNPX simulations
Four best layers
Contribute ~76%
of the efficiency
Can re-deploy the
other four to double
the area.
Optical optimization Redesign optical configuration for planar detector New waveshifting materials and techniques
Choice of thickness
Average pulse height (mV)
800
600
400
200
0 0
50
100 150 Thickness of layer (µm)
200
250
Pulsed LED light, 460nm, shone right through layer
Neutron capture efficiency cumulative capture efficiency
0.4
0.3
0.2
0.1
0.0 0
100
200 300 layers, each 1µm thick
400
500
Capture efficiency of BN-ZnS(Ag) screens. MCNPX simulations.
Improved production of layers Capture compound + Scintillator + Binder ~ 220 ± 10 microns Minimize wastage, avoid aggressive solvents… Original detectors used spreading technique Considered spray painting, powder coating, serigraphy, ink-jet systems – but went back to spreading.
Binder / solvent Kraton G1652; linear triblock copolymer based on styrene and ethylene/butylene (SEBS)
– dissolves in light mineral oils forming a gel
Solvent – "white spirit"
500
400
BN 241
Am
counts
300
200 6
LiF
100
0 0
50
100
150
200
bin number
Pulse height distributions in ZnS(Ag) screens
Light output from ZnS(Ag) screens 100
241 6
Light output (A.U.)
Am
LiF
80
60 BN 40
20
0 0
1
2
3 Energy (MeV)
4
5
6
BN gives ~0.43 of the light produced by 6LiFmaterial.
BN costs ~1000 less than 6LiF !!
Wavelength shifting lightguides Low cost technique using BBQ dye disperse dyed into surface of clear acrylic sheet
Tests with small samples suggest that this 1.2 times better than original Plexiglas GS2025 material
Neutron discrimination Original pulse counting system used hard-wired TTL Depends on choice of capture compound, scintillator, waveshifter and optical collection New computer based monitoring system controlling hardware decisions
Current status Working with ET-Enterprises, Ludlum, Eljen to produce a commercial version. Paper on layer production almost ready to go. Paper on detectors started Paper on discrimination started.
Research funded by UK Home Office Scientific Development Branch and STFC
Questions? e-mail
[email protected]