Advanced semiconductor detectors of neutrons Institute of Experimental and Applied Physics Czech Technical University in Prague Josef Uher Seminar
Outline Neutron detection principle Limitations of single planar structure 3D detectors Simulation results Converter filling technology Experimental tests seminar
The standard planar detector silicon detector Semiconductor itself can not detect neutrons directly! The bias (V) needed to deplete a detector is proportional to the square of the distance between the electrodes. For a planar detector, this is equal to the thickness of the detector. Similarly, the collection time (Tc) is proportional to the distance between the electrodes and, assuming a single trap level with a lifetime τ, the trapping is related to exp(-tc/τ). seminar
Planar detector + neutron converter Conversion of thermal neutrons to heavy charged particles in 6Li or 10B converter layer. 10B reaction (Cross section 3840 barns at 0.0253 eV): 10B+n a (1.47 MeV) + 7Li (0.84 MeV) + g (0.48MeV) (93.7%) 10B+n a (1.78 MeV) + 7Li (1.01 MeV) (6.3%) 6Li reaction (cross section 940 barns at 0.0253 eV) : 6Li + n a (2.05 MeV) + 3H (2.72 MeV) Converter (LiF) Detector a T seminar
Planar geometry – comparison of efficiency (amorphous B vs. LiF) 6LiF, enrichment 90% Amorphous 10B, enrichment 80% LiF: ranges in Si are RT=44.1um, Ra=8.6um, in LiF: 48um / 8um (1.8g/cm3) B: ranges in Si are RLi=3um / 2.7um, Ra=5.4um / 5.2um, in B: 5um / 8um (1 g/cm3) Threshold 50 keV Efficiencies are comparable. Higher cross section of 10B does not spawn a significant increase of efficiency. seminar
2D neutron array modification Neutron beam converter T back side contact n+ “Standard” 2D type n a p+ sensitive volume “Egg plate” 2D type (with enlarged surface to increase the detector efficiency) seminar
Neutron array modification Neutron beam back side contact grid “Channel” 2D type (maximized filling) low n+ n p+ “3D inverse” structure (there are pillars instead of pores) bottom view seminar
Examples of created structures Photo-electrochemical etching (KTH, Stockholm) An example of structures created in KTH Stockholm and University of Glasgow. (just examples) Laser ablation (University of Glasgow) seminar
“3D inverse” structure Processed using saw for chip separation 300x300mm 60mm deep 100x100mm 60mm deep 200x200mm 60mm deep seminar
3D geometry arrays - comparison of cylindrical vs. cubic 6LiF converter Fixed wall thickness – variance in the converter / cell size Cube Cylinder Ratio of captured neutrons: LiF: Box … 77.8%, cylinder … 61.0%, planar … 28.8% B: Box … 85.8%, cylinder … 65.9%, planar … 25.8% Maximal efficiency: ~32% Maximal efficiency: ~27% seminar
3D geometry arrays - comparison of cylindrical vs. cubic 10B converter Fixed wall thickness – variance in the converter / cell size Cube Cylinder Maximal efficiency: ~36% Maximal efficiency: ~31% seminar
3D geometry arrays - comparison of LiF vs. amorphous B converter Fixed wall thickness – variance in the converter / cell size LiF B Hustotou je ovlivněn makroskopický účinný průřez Threshold 50 keV LiF – with increasing converter density is increasing efficiency (max ~32%, density 2.64 g/cm3 (!)) B – with increasing converter density is decreasing efficiency (max ~36%, density 1.0 g/cm3 (!)) seminar
Deposited Energy Distribution The edge between the converter and the semiconductor LiF, r=1.2 gcm-3 40mm diameter LiF, r=2.64 gcm-3 40mm diameter The thermal neutron beam diameter is 2mm and it is penetrating the LiF converter in the center. seminar
Deposited Energy Distribution The edge between the converter and the semiconductor LiF, r=1.4 gcm-3 100mm diameter LiF, r=2.64 gcm-3 100mm diameter The thermal neutron beam diameter is 2mm and it is displaced 45mm from the LiF converter center. seminar
Planar and 3D geometry spectra comparison LiF density: 2.0 g/cm3 Surface density 2 mg/cm2 Layer thickness 10 mm Diameter 58 mm seminar
Pores filling using pressure Chip Empty pores Lead hob Metal pad Final preparation ready for pressing Poured powder BaSO4 = barium sulfide Lis = presser force = 80kN Pressing Covered by foil seminar
Pores filling using pressure BaSO4 BaSO4 BaSO4 LiF seminar
Pores filling using pressure Roentgenogram of filled structures BaSO4 LiF In case of BaSO4 are pores uniformly filled In case of LiF there is a higher noise due to lower Z of converter Average attenuation => estimation of filling depth. Pores depth is 150um Estimated average filling depth is 150mm seminar
2D stuffed detector A next step in the development would be 2D detector diode with etched pores filled with a neutron converter. seminar
Experimental samples Charge collection tests Detection efficiency tests seminar
Pyramidal dips Characterization of the structure using alphas from 241Am source. Spectrum - angle 0 deg Spectrum - angle 70 deg Position of the first peak Interpretation of such integral measurement is difficult. Further measurements are necessary. seminar
The Medipix 2 imaging detector Pixel array: 256x256 Pixel size: 55x55 mm2 Total sensitive area: ~2 cm2 Electronics for each pixel: preamplifier, two discriminators (energy window), 13-bit counter Read out: serial - 9 ms, parallel - 266 ms (clock 100MHz) Serial readout speed: 6 fps Integrated source of variable detector bias (5 - 105V) 4kB EEPROM for configuration Temperature monitoring Cables with length up to 5m Ability to flash a firmware seminar
Measurements Parameters of the Thermal Neutron Beam: Horizontal channel (neutron guide) of the LVR-15 nuclear research reactor at Nuclear Physics Institute of the Czech Academy of Sciences at Rez near Prague. Intensity about 107 neutrons/(cm2s) at reactor power of 8MW Beam Cross section: 4 mm (height) x 60 mm (width) The divergence of the neutron beam is < 0.5° Spallation neutron source in Paul Sherrer Institut at Villigen in Switzerland Intensity about 3·106 neutrons/(cm2s) at proton accelerator current of 1mA and proton energy of 590 MeV Beam Cross section: 40 cm in diameter seminar
Comparison of Medipix-2 and other neutron imaging detectors Tested: CCD camera with scintilator mixed with 6Li (pixel size 0.139 mm) Imaging plate (excitation by neutrons, deexcitation by laser scanner followed by light emission, pixel size 50mm) Medipix-2 s=0.83 pixel =46 mm CCD camera s=2.5 pixel =350 mm s=1.06 pixel =53 mm Imaging plate s=0.93 pixel =158 mm Medipix-1 The resolution the best in the World. Efficiency – 3D structures Area – quads (8cm^2) Medipix-2 has the highest dynamic range and resolution. The only disadvantages are lower efficiency (2-3%) and sensitive area. seminar
Sample objects – blank cartridge Medipix-2 Roentgenography Photograph Medipix-1 Neutronography Imaging plate CCD Medipix-1 Medipix-2 seminar
Sample objects – fishing line Fishing line diameter 100 mm Medipix-1 Imaging plate Medipix-2 seminar
Conclusion Simulation software More geometries has been simulated Optimal structure parameters have been found Filling using pressure has been tested Testing devices have been proposed Neutron imaging using Medipix detector has been successfully tested seminar