Development of dual MCP x-ray imager for 40 ~ 200 kev region National ICF Diagnostics Working Group Meeting - October 6-8, 2015 N. Izumi, G. N. Hall, A. C. Carpenter, F. V. Allen, J. G. Cruz, B. Felker, J. Holder, J. D. Kilkenny, A. Lumbard, R. Montesanti, N. E. Palmer, K. Piston, G. Stone, M. Thao, R. Vern, R. Zacharias, O. L. Landen, R. Tommasini, D. Hargrove, D. K. Bradley and P. M. Bel LLNL-PRES-661074 This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract DE-AC52-07NA27344. Lawrence Livermore National Security, LLC
Summary Using a novel dual MCP, we have demonstrated an imaging detector with 3.2x higher DQE @ 59 kev (1) Compton radiography requires gated x-ray imager with DQE > 4% (2) Single MCP camera becomes noisy over 30 kev due to depth dependent electron gain of MCP (3) With using dual MCP, it is possible to suppress the noise due to depth dependent gain (4) With using with dual MCP, DQE = 4.5% was achieved (5) The 1 st dual MCP camera AXIS is in its production phase 2 2
MOTIVATION We need a gated imager for 40~200 kev x rays 3 3
Compton backlight Radiography based on Compton scattering can reveal the DT fuel assembly near bang time Tommasini, et al., PoP 18, 2011 Short pulse laser (ARC) 10 µm Au wire Imploded core Gated x-ray Detector Gating is essential to reduce background from hohlraum Expected contrast Photoelectric dominated High pass filter >40 kev Ideal contrast (OD ~1) for high density DT fuel near maximum compression time Flat cross section allows broadband radiography (40~ 200 kev) 10x-20x more signal than K-alpha High-pass filtration suppress the background from core emission Compton radiography requires gated imager which works well in 40~200 kev region 4 4
Goal of the project We need a DQE of 4% to achieve a SNR of 20 to measure fuel R with 5% rms accuracy SNR 35 30 25 20 15 10 Single MCP Dual MCP 3% 5% 10% Obtainable accuracy of R measurement 5 0 0 1 2 3 4 5 6 20% Perfect Radiograph DQE (%) 300µm 5
Why do we need a new detector? Conventional MCP cameras becomes noisy in high energy regime 6 6
Standard MCP based camera becomes noisy for x ray energy over 30 kev Why does MCP image become noisy over 30keV? (1) Stochastic behavior of avalanche stream Multiplication has event-by-event variability (2) Avalanche starts from few electrons (3) Depth dependent MCP gain High energy x-rays penetrate deep into MCP NEXT slide 7
MCP gain depends on the depth of the detection event Depth effect is significant for high energy x rays MFP ( micron) 1 10 5 1 10 4 1 10 3 100 10 1 X-ray MFP in MCP glass (4gm/cm3) Detection events on top surface experience high gain Deep events experience small gain 0.1 1 10 100 1 10 3 x-ray energy (kev) 8 8
Due to this depth dependent gain, width of the pulse height response depends on x ray energy Narrow PHD for 5 kev photons Broad PHD for 59 kev photons 9 9
What is DQE? High DQE = low NF 10 1
Detective Quantum Efficiency (DQE) is a useful metric to evaluate SNR of noisy detectors (1) Quantum efficiency (QE) is a good metric for detectors which have narrow pulse height distribution (e.g. CCD, HPGe) QE detected_events (2) For detectors which have avalanche amplification incident noise, _ photons DQE is the quantity of interest ( e.g. Proportional counter, MCP) N N SNR QE N incident _ photons DQE SNR SNR SNR out in 2 DQE N out incident _ photons 11 1
Noise Factor (NF) is another useful metric which represents the relation of QE and DQE (1) Quantum Efficiency QE N N DQE where detected_events incident _ photons (2) Detective Quantum Efficiency QE NF NF Narrower pulse height response 2 cl 1 2 Cluster noise Standard deviation of the pulse height distribution Average of the pulse height distribution Lower NF Higher DQE 12 1
To achieve required SNR, we need a detector which has high DQE (small Noise Factor) Comparison of two detectors with the same QE Detected events per resolution element High DQE detector QE : 10% NF : 1 DQE : 10% Low DQE detector QE : 10% NF : 8 DQE : 1.25% Signal-to-Noise Ratio is determined by DQE 13 1
CONCEPT We can reduce noise using a dual MCP configuration 14
Proposed solution We are increasing DQE using a dual MCP configuration Dual MCP (AXIS) MCP Low gain MCP High gain phosphor FILM electron photon x X-ray The 1 st MCP (800 μm thick) is operated in low gain and works as thick photocathode The 2 nd MCP (460 μm thick) is operated in high gain and works as main amplifier Because of low gain of the 1 st MCP, the output is less sensitive to the depth of the detection high DQE 15 1
The dual MCP provides good performance regardless of gain setting Expected performance for 60 kev x-rays (Gain ~1000x) Dual MCP 800 + 460 μm Single MCP 460 μm QE 13% 8 % DQE 4 ~ 4.5% Independent to total gain 1.4 % Depends on MCP gain Dynamic range Gate width Expected to be wide because PHD is narrower 0.5 ~ 1ns Limited by electron transit time High-end tail of the PHD saturates first (Events near entrance) 70 ~ 100 ps Limited by electron transit time 16 1
EXPERIMENT We measured DQE of MCPs 17 1
We measured QE and DQE of the dual MCP configuration camera using 59 kev RI source MCP Low gain MCP High gain Phosphor FOP electron PMT x Am 241 X-ray Charge sensitive amplifier Multi Channel Analyzer 18 1
Dual MCP The pulse height response changes as a function of bias voltage given to the MCP1 The maximum DQE = 4.5% was observed when MCP 1 bias = 625V When When Beyond 1 st the MCP 1 st this MCP turned point, is not DQE on, biased, x-rays goes detected all down the because events on are pulse height 1 st coming MCP response start from contributing the is expanding 2 MCP the (depth output dependent Both gain effect) QE and DQE go up 10 4 Pulse height response 1.0 Counts / pc 10 3 10 2 10 1 QE and DQE 0.1 QE V2:600V 10 0 1 10 100 1000 0.01 DQE PMT output (pc) 100 1000 10 4 Total MCP gain 19
Dual MCP We optimized the bias voltage of 1 st MCP DQE = 4.5% was achieved successfully QE and DQE 0.1 10-1 MCP1: 800 μm MCP2: 460 μm QE Requirement DQE 2 nd MCP only 0.01 10-2 100 1000 10 4 *Nominal MCP gain *Total MCP gain was changed by adjusting the bias voltage to MCP1 20 2
DQE of the dual MCP set us was measured at several energies using the same method 100.00% QE and DQE 10.00% 1.00% Monte Carlo L: 800 um Measured QE Measured DQE 0.10% 0 50 100 150 X-ray energy 21
Current status of of AXIS 22
AXIS is DIM based x ray framing camera with two 40 mm x 40 mm frames 23 2
To use in harsh neutron background, KODAK TMAX 400 film is used for recording Dark Slide Film Pack Clamps 24 2
Assemble of MCP head is completed Phosphor HV supply Detector Transformers 9:1 Gain Transformers 25:1 Phosphor Slow Down Box Launch boards Bias Box Lemo 6 pin dc bias connector 50Ω Monitor Trace HV feed thrus Detector MCPs Grid Receiver boards HV contacts Gain MCPs 25 2
The AXIS head was assembled and currently being tested on HEX source at NSTec 26
The first AXIS image was obtained with using Manson source 27
Summary Using a novel dual MCP, we have demonstrated an imaging detector with 3.2x higher DQE @ 59 kev (1) Compton radiography requires gated x-ray imager with DQE > 4% (2) Single MCP camera becomes noisy over 30 kev due to depth dependent electron gain of MCP (3) With using dual MCP, it is possible to suppress the noise due to depth dependent gain (4) With using with dual MCP, DQE = 4.5% was achieved (5) The 1 st dual MCP camera AXIS is in its production phase 28 2